-------
REFERENCES
Allen, J. O.; Dookeran, N. M; Taghizadeh, K.; Lafleur, A. L.; Smith, K. A.; Sarofim, A. F. (1997) Measurement of
oxygenated polycyclic aromatic hydrocarbons associated with a size-segregated urban aerosol. Environ Sci
. Technol. 31:2064-2070. ' . ,
Allen, G. A.; Lawrence, J.; Koutrakis, P. (1999) Field validation of a semi-continuous method for aerosol black
carbon (aethalometer) and temporal patterns of summertime hourly black carbon measurements in
southwestern PA. Atmos. Environ. 33: 817-823.
Andrews, E.; Saxena, P.; Musarra, S.; Hildemann, L .M.; Koutrakis, P.; McMurry, P. H.; Olmez, I.; White, W. H.
(2000) Concentration and composition of atmospheric aerosols from the 1995 SEAVS experiment and a
review of the closure between chemical and gravimetric measurements. J. Air Waste Manage Assoc 50-
648-664.
Chow, J. C.; Watson, J. G.; Lu, Z.; Lowenthal, D. H.; Frazier, C. A.; Solomon, P. A.; Thuillier, R. H.; Magliano, K.
(1996) Descriptive analysis of PM25 and PM10 at regionally representative locations during
SJVAQS/AUSPEX. In: Parrish, D.; Trainer, M.; Rao, S. T.; Solomon, P. A., eds. A&WMA international
specialty conference on regional photochemical measurements and modeling, part 2; November 1993; San
Diego, CA. Atmos. Environ. 30: 2079-2112.
Christoforou, C. S.; Salmon, L. G.; Hannigan, M. P.; Solomon, P. A.; Cass, G. R. (2000) Trends in fine particle
concentration and chemical composition in southern California. J. Air Waste Manage. Assoc. 50: 43-53.
Cui, W.; Machir, J.; Lewis, L.; Eatough, D. J.; Eatough, N. L. (1997) Fine particulate organic material at Meadview
during the project MOHAVE summer intensive study. J. Air Waste Manage. Assoc. 47: 357-369.
Fraser, M. P.; Cass, G. R.; Simoneit, B. R. T.; Rasmussen, R. A. (1998) Air quality model evaluation data for
organics. 5. C6-C22 nonpolar and semipolar aromatic compounds. Environ. Sci. Technol. 32: 1760-1770.
Gertler, A. W.; Lowenthal, D. A.; Coulombe, W. G. (1995) PM10 source apportionment study in Bullhead City,
Arizona. J. Air Waste Manage. Assoc. 45: 75-82.
Hegg, D. A.; Livingston, J.; Hobbs, P. V.; Novakov, T.; Russell, P. (1997) Chemical apportionment of aerosol
column optical depth off the mid-Atlantic coast of the United States. J. Geophys. Res. 102: 25,293-25,303.
IMPROVE: interagency monitoring of protected visual environments [database]. (2000) [Data on particulate'organic
and elemental carbon concentrations after 1995]. Fort Collins, CO: National Park Service (NFS); Cooperative
Institute for Research in the Atmosphere (CIRA). Available at: http://vista.cira.colostate.edu/improve/ [2001,
January 26].
Khwaja, H. (1995) Atmospheric concentrations of carboxylic acids and related compounds at a semiurban site.
Atmos. Environ. 29: 127-139.
Klinedinst, D. B.; Currie, L. A. (1999) Direct quantification of PM25 fossil and biomass carbon within the northern
front range air quality study's domain. Environ. Sci. Technol. 33: 4146-4154.
Lewtas, J.; Pang, Y.; Booth, D.; Reimer, S.; Eatough, D. J.; Gundel, L. A. (2001) Comparison of sampling methods
for semi-volatile organic carbon associated with PM2 5. Aerosol. Sci. Technol. 34: 9-22.
Lioy, P. J.; Daisey, J. M. (1987) Toxic air pollution: a comprehensive study of non-criteria air pollutants. Chelsea,
MI: Lewis Publishers.
Malm, W. C.; Day, D. E. (2000) Optical properties of aerosols at Grand Canyon National Park. Atmos Environ 34-
3373-3391. ' '
Malm, W. C.; Gebhart, K. A. (1996) Source apportionment of organic and light-absorbing carbon using receptor
modeling techniques. Atmos. Environ. 30: 843-855.
Offenberg, J. H.; Baker, J. E. (2000) Aerosol size distributions of elemental and organic carbon in urban and
over-water atmospheres. Atmos. Environ. 34: 1509-1517.
Omar, A. H.; Biegalski, S.; Larson, S. M.; Landsberger, S. (1999) Particulate contributions to light extinction and
local forcing at a rural Illinois site. Atmos. Environ. 33: 2637-2646.
Pedersen, D. U.; Durant, J. L.; Penman, B. W.; Crespi, C. L.; Hemond, H. F.; Lafleur, A. L.; Cass, G. R. (1999)
Seasonal and spatial variations in human cell mutagenicity of respirable airborne particles in the northeastern
United States. Environ. Sci. Technol. 33: 4407-4415.
Rogge, W. F.; Mazurek, M. A.; Hildemann, L. M.; Cass, G. R.; Simoneit, B. R. T. (1993) Quantification of urban
organic aerosols at a molecular level: identification, abundance and seasonal variation. Atmos Environ Part A
27: 1309-1330.
Saxena, P.; Hildemann, L. M. (1996) Water-soluble organics in atmospheric particles: a critical review of the
literature and applications of thermodynamics to identify candidate compounds. J. Atmos. Chem. 24: 57-109.
March 2001
3 A-15 DRAFT-DO NOT QUOTE OR CITE
-------
Schauer, J. J.; Cass, G. R. (2000) Source apportionment of wintertime gas-phase and particle-phase air pollutants
using organic compounds as tracers. Environ. Sci. Technol. 34:1821-1832.
Turpin, B. J.; Huntzicker, J. J. (1995) Identification of secondary organic aerosol episodes and quantitation of
primary and secondary organic aerosol concentrations during SCAQS. Atmos. Environ. 29: 3527-3544.
U.S. Environmental Protection Agency. (1996) Air quality criteria for paniculate matter. Research Triangle Park,
NC: National Center for Environmental Assessment-RTF Office; report nos. EPA/600/P-95/001aF-cF. 3v.
Veltkamp, P. R.; Hansen, K. J.; Barkley, R. M.; Sievers, R. E. (1996) Principal component analysis of summertime
organic aerosols at Niwot Ridge, Colorado. J. Geophys. Res. [Atmos.] 101: 19,495-19,504.
March 2001
3 A-16
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
APPENDIX 3B
.Composition of Particulate Matter Source Emissions
This appendix includes discussions of the elemental composition of emissions from various
source categories discussed in Table 3-7. Discussions in this appendix incorporate material
dealing with the inorganic components of source emissions from Chapter 5 of the 1996 PM
AQCD (U. S. Environmental Protection Agency, 1996), updates to that material, and material
describing the composition of organic components in source emissions. The primary emphasis in
the discussions is on the composition of PM25 particle sources.
Soil and Fugitive Dust
The compositions of soils and average crustal material are shown in Table 3B-1 (adapted
from Warneck, 1988). Two entries are shown as representations of average crustal material.
Differences from the mean soil composition shown can result from local geology and climate
conditions. Major elements in both soil and crustal profiles are Si, Al, and Fe, which are found in
the form of various minerals. In addition, organic matter constitutes a few percent, on average, of
soils. In general, the soil profile is similar to the crustal profiles, except for the depletion of
soluble elements such as Ca, Mg, Na, and K. It should be noted that the composition of soils from
specific locations can vary considerably from these global averages, especially for elements like
Ca, Mg, Na, and K.
Fugitive dust emissions arise from paved and unpaved roads, building construction and
demolition, parking lots, mining operations, storage piles, and agricultural tilling in addition to
wind erosion. Figure 3B-1 shows examples of size distributions in dust from paved and unpaved
roads, agricultural soil, sand and gravel, and alkaline lake bed sediments, which were measured in
a laboratory resuspension chamber as part of a study in California (Chow et al., 1994). This figure
shows substantial variation in particle size among some of these fugitive dust sources. The PM, „
abundance (6.9%) in the total suspended PM (TSP) from alkaline lake bed dust is twice its
abundance in paved and unpaved road dust. Approximately 10% of the TSP is in the PM2 5
fraction and approximately 50% of TSP is in the PM10 fraction. The sand/gravel dust sample
shows that 65% of the mass is in particles larger than the PM10 fraction. The PM2 5 fraction of
March 2001
3B-1
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 3B-1. AVERAGE ABUNDANCES OF MAJOR ELEMENTS IN
SOIL AND CRUSTAL ROCK
Element
Si
Al
Fe
Ca
Mg
Na
K
Ti
Mn
Cr
V
Co
Elemental
Soil
0)
330,000
71,300
38,000
13,700
6,300
6,300
13,600
4,600
850
200
100
8
Abundances (ppmw)
(2)
277,200
81,300
50,000
36,300
20,900
28,300
25,900
4,400
950
100
135
25
Crustal Rock
(3)
311,000
77,400
34,300
25,700
33,000
31,900
29,500
4,400
670
48
98
12
Source: (1) Vinogradov (1959); (2) Mason (1966); (3) Turekian (1971), Model A; as quoted in Warneck (1988).
1 TSP is approximately 30 to 40% higher in alkaline lake beds and sand/gravel than in the other soil
2 types. The tests were performed after sieving and with a short (<1 min) waiting period prior to
3 sampling. It is expected that the fraction of PMj 0 and PM25 would increase with distance from a,
4 fugitive dust emitter as the larger particles deposit to the surface faster than do the smaller
5 particles.
6 The size distribution of samples of paved road dust obtained from a source characterization
7 study in California is shown in Figure 3B-2. As might be expected, most of the emissions are in
8 the coarse size mode. The chemical composition of paved road dust obtained in Denver, CO,
9 during the winter of 1987-1988 is shown in Figure 3B-3. The chemical composition of paved
10 road dust consists of a complex mixture of particulate matter from a wide variety of sources.
11 Hopke et al. (1980) found that the inorganic composition of urban roadway dust in samples from
March 2001
3B-2
DRAFT-DO NOT QUOTE OR CITE
-------
100
80
at
60
40
20
Paved Unpaved Agricultural Soil/Gravel
Road Dust Road Dust Soil
[Zl<1.0|jnn IHl<2.5|jm Bl<10|jm KDlSP
Alkaline
Lake Bed
Figure 3B-1. Size distribution of particles generated in a laboratory resuspension
chamber.
Source: Chow etal. (1994).
1 Urbana, IL, could be described in terms of contributions from natural soil, automobile exhaust,
2 rust, tire wear, and salt. Automobile contributions arose from exhaust emissions enriched in Pb;
3 from rust as Fe; tire wear particles enriched in Zn; brake linings enriched in Cr, Ba, and Mn; and
4 cement particles derived from roadways by abrasion. In addition to organic compounds from
5 combustion and secondary sources, road dust also contains biological material such as pollen and
6 fungal spores.
7 Very limited data exist for characterizing the composition in organic compounds
8 resuspended paved road dust and soil dust. The only reported measurements are from Rogge et al.
9 (1993a) and Schauer and Cass (2000), which consist of data for the fine particle fraction. The
10 resuspended road dust sample analyzed Rogge et al. (1993 a) was collected in Pasadena, California
during May of 1988. The sample analyzed by Schauer and Cass (2000) is a composite sample
March 2001
3B-3
DRAFT-DO NOT QUOTE OR CITE
-------
100
80
60
1
DL
40
20
95.8%
93.1%
(<2.5R
92.4%
99.2%
97.4%
(<2.5|J)
87.4%
34.9%
Road and Agricultural Residential Diesel Crude ON Construction
Soil Dust Burning Wood Truck Combustion Dust
Combustion Exhaust
Code:
2.5p-10M
1M-2.5|J
Figure 3B-2. Size distribution of California source emissions, 1986.
Source: Houck et al. (1989,1990).
March 2001 3B-4 DRAFT-DO NOT QUOTE OR CITE
-------
Chemical Species
Figure 3B-3. Chemical abundances for PM2 5 emissions from paved road dust in Denver,
CO. Solid bars represent fractional abundances, and the error bars
represent variability in species abundances. Error bars represent detection
limits when there are no solid bars.
Source: Watson and Chow (1994).
1
2
3
4
5
6
7.
collected at several sites in the Central Valley of California in 1995. In both cases, road dust
samples were resuspended in the laboratory. Samples were drawn through a PM2 0 cyclone
upstream of the collection substrate to remove particles with aerodynamic diameters greater than
2.0 /um. It is unclear if these samples are representative of road dust in other locations of the
United States. Table 3B-2 summarizes the organic compounds measured in these road dust
samples.
March 2001
3B-5
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 3B-2. SUMMARY OF PARTICLE-PHASE ORGANIC COMPOUNDS
PRESENT IN FINE PARTICLE ROAD DUST SAMPLE
Source
Pasadena Road Dust
(Roggeetal., 1993a)
San Joaquin Valley
Road Dust (Schauer
and Cass, 2000)
Contribution to Dominant Contributors to
Compound Class Particulate Mass (%) Emissions of Compound Class
n-Alkanes
n-Alkanoic acids
n-Alkenoic acids
Petroleum biomarkers
PAH
n-Alkanals
n-Alkanols
n-Alkanes
n-Alkanoic acids
n-Alkenoic acids
0.13
0.37
0.028
0.017
0.0059
0.046
0.021
0.023
0.23
0.095
C]7> Ci9, C2,
Palmitic acid and stearic acid
Oleic acid and linoleic Acid
Hopanes and steranes
No dominant compounds
Octacosanol and triacontanal
Hexacosanol and octacosanol
No dominant compounds
Palmitic acid and stearic acid
Oleic acid, linoleic acid, and
hexadecenoic acid
1 Stationary Sources
2 The elemental composition of primary particulate matter emitted in the fine fraction from a
3 variety of power plants and industries in the Philadelphia area is shown in Table 3B-3 as a
4 representative example of emissions from stationary fossil combustion sources (Olmez et al.,
5 1988). Entries for the coal fired power plant show that Si and Al followed by sulfate are the
6 major primary constituents produced by coal combustion, whereas fractional abundances of
7 elemental carbon were much lower and organic carbon species were not detected. Sulfate is the
8 major particulate constituent released by the oil fired power plants examined in this study, and,
9 again, elemental and organic carbon are not among the major species emitted. Olmez et al. (1988)
10 also compared their results to a number of similar studies and concluded that their data could have
11 much wider applicability to receptor model studies in other areas with some of the same source
12 types. The high temperature of combustion in power plants results in the almost complete
13 oxidation of the carbon in the fuel to CO2 and very small amounts of CO. Combustion conditions
14 in smaller boilers and furnaces allow the emission of unbumed carbon and sulfur in
March 2001
3B-6
DRAFT-DO NOT QUOTE OR CITE
-------
^H
51
fc
g
H
CO
0
>
pa
Q
CO
h-i
CO
u
o
c
f^
M"j
s
^4
g
g
§
0
H
^••4
WD
o
CU
M
o
u
s
a
3
H
3
1
ELPHIA
Q
j
a
PH
r 1
s
ffi
P""4
g
co
u
0
CO
"c
_j
ji
i
*
c
^
5
c
a
C
0
-2
c
5u
tn
1
£
•g
£
O
1
•5
c
U
u
c
c
1
^
w
$
1
CO
*
i o
* cd
* §
M S3
si
*
Cracker
^
r c
Is
\<
^
J^3
^
s
JS
o
00
*
(U
o
JT
T3
^
c
M
^*
U
o
s,
iS
"S
"o
-
NO
CN
0
41
r-
0
Q
2
CN
in
41
**
m
o
-H.
*n
O
m
CN
,__(
41
t —
CN
Q
£•'
V
U
-
CN
O
41
in
rn
en
in
0.16 ±0.0
CN
O
o
4)
oo
O
"*
^
^^
O-
41
CN
CN
O'
en
in
^_i
41
t-~
^
en
CN
0
41
ON
oo
O
?
U
|
U
-
r-
0
c>
41
NO
en
O
m
CN,
CN
0
41
O
CN
ON
41
CN
^
C--
i — i
41
f-.
m"
co
vO
. i— i
41
in
rn
m
ON
C3
41
ON
oo
"-•'
C?
5
2
en
in
en
41
NO
en
0.38 ±0.0
„
oo
O
41
en
NO
m
00
o
41
rn
rn
en
oo
o
41
p
en"
en
8
o
41
en
0
^N.
^a
^
en
O
0
41
in
CN
O
en
41-
00
NO
^_
S
C>
41
en
oo
O
o
41
Tf
ON
O
m
ON
0
O
41
in
"si*
o
m
41
2
^
^
<;
o
m
o
41
-
ON
O
o
CN
41
oo
ON
CN
CN
CN
41
en
~
^c
O
41
NO
CN"
ON
NO
0
41
ON
*™^
ON
NO
41
oq
g
K
0
CN
O
41
en
NO
C5
Q
CN
r-
CN
o
41
O
«
CN
C5
41
p
0\
O
41
""!
r~t
ON
0.62 ±0.10
^
a-
o
oo
«D
41
ON
CN
ON
NO
CN
41
CN
41
en
~
41
m
ON
CN
41
*"~*
ON
NO
O
41
en
g
-
en
CN
41
oo
NO
en
41
00
m
CN
CN
41
ON
*
41
^
cn
T}*
41
O
en
CN
41
ON
g>
cf
CO
en
in
41
ON
CN
Q
F— t
41
CN
Q
CN
ON
§
O
41
ON
O
O
m
0.022 ±0.11
g?
O
0
m
CN
41
NO
ON
O
O
O
41
m
o
o
CN
in
41
ON
O
~
8
o"
41
CN
O
ON
O
0
41
NO
O
ON
1.20 ±0.09
g
&
0
o
o
41
m
CN
0
ON
.030 ±0.004
o
CM
%
o
41
CN
m
p
41
en
CN"
en
p
_J
41
NO
en
en
0
41
rt
T— t
?
3
-
CN
0
O
41
O
en
O
41
CN
-
ON
en
o
.092 ±0.
o
en
8
o
41
S
0
en
s
o
41
f-
0
en
CN
41
a
u
GO
o
£
0
o
41
o
8
o"
m
0.38 ±0.1
CN
en
O
o
.024 ±0.
~
8
o
41
CN
O
ON
I
o
41
o
0
en
CN
O
41
g
P
CN
en
in
41
NO
00
en
o
>n
CN
-
41
NO
en
en
O
§
en
41
O
O
o
cT
CN
m
o
o
o
m
41
O
o
in
~
m
0
r—
41
O
>n
in
1
en
en
41
ON
' ON
en
oo
41
ON
m
—
O
CN
41
O
en
o
t-~
41
o
en
CN
m
O
*— *
41
m
m
CN
m
o
CN
41
m
I
5
en
0
^r
41
in
NO
en
en
41
-
m
41
CN
en
O
in
41
O
CN
en
^"
41
O
oo
m .
m
m
41
o
ON
CN
N§
J
m
o
o
41
CN
CN
0
ON
8
o
41
O
CN
o
-
CN
O
O
41
en
O
m
^
o
41
en
CN
O
41
NO
—•
m
C5
41
NO
g
U
tin
m
oo
c>
41
en
m
CN
41
in
~
CN
41
m
m
o
o
CN
41
O
O
m
o
»— i
41
O
?^
m
o
41
en
ON
I
o
u
O
0
TT
41
CN
ON
O
en
41
CN
CN
CN
O
41
§
en
~
i
CN
41
1
"
ON
g
O
in
41
o
o
o
in
ON
O
in
41
0
oo
en
|
March 2001
3B-7
DRAFT-DO NOT QUOTE OR CITE
-------
1ONARY
<
H
co
1
ps
•n t-~
Tt O
C5
«— CO CO
"-1
o o
^ 3 s
•H -H -H
g 00 0
2 <~: "">
•— • o
os co —
cX 2 *
41 -H •«
S 2 s
o\ *^
O\ CO CO
290 ±20
0.041 ±0.005
640 ±80
^ ~ t
a I a
a .g <
CO O
o
S §
•" «
9 1
CN
CO OS
oo
~H 1 1
Tl
t/-J
—• fN
0
co t--
-H -H
VO O
vo co
VO
CO -^
1-H
41 -H
CO OS
Os g
-H 41
VO o
^ §
CO CO
250 ±20
35 ±8
0. 0.
CO (S
CN
230 ±50
Q
_
00
CO
-H
Q
Q
Z
-
190 ±80
a.
a
£
o
00
Os ON
VO S
41 41
VO O
CO CO
Q Q
Z Z
.-. <— 1
*— 1 *— <
41 ^
OS Os
s §
il
OS OS
1290 ±60
490 ±190
It "s1
°- g;
* . fl
O CO
o .
CO V~l
41 41
O —
•^t r—
CN
0 ,Q
Z Z
ii
CO
1500 ±300
ND
CO
930 ±210
ND
fN
O
VO
-H Q
o Z
r-
1!
o on
S <
CO CO O
_ o
1 2 §
-H -H -H
| -3 |
Q Q Q
Z Z Z
fN
S
Q Q ^
Z Z 5
v>
in
^-
i— *
o
oo
§ i o
s
Os
O
CO
Q Q S
* 2 S
o co rN
41 41 41
O O 0
8 m vo
^ CO
fN
CO CO CO
00^
*~^ rn rN
o o •*-
S 2 S
CO fN
O O
^\ _LJ _|,l
1-1 -H -H
Zoo
CN OO
'e' 1- "i"
P. & OH
a a a
es BJ 1>
pa j c_>
Q Q
CO CO
S £
fN CO
t-;
i I
CO
CO CO
41 "H
g s
II
a a
Z on
March 2001
3B-8
DRAFT-DO NOT QUOTE OR CITE
-------
oa'
o
£
>"
03
Q |
|E
4> <
t>D
12
a 1
J J
&< -S
fe "»
o
CL,
E
O (o
I
•e
2
I
1 s,
Eddystone Cc
Fired Powe
Plant
§.!
oo 3
Q Q Q Q Q
en en en en en
T 0 ^ "* ~
O — — < O O
-H -H -H -H -H
^ f~ oo en ""!
O
Q Q Q Q Q
22222
en en
en ON
tN tN
.® Q ® Q Q
VO ON
G> 0
Q Q Q Q Q
en en «
>n m "1
S Q S ? Q
^, 2 m en 2
"> <^ 2
^-s ^v
s ft s 1. ^
3 T3 ^ JO 3
BO O H > ,3
en o . r~
tN tN
e Q Q ° ^ Q ^
^ ^ s
0
en en ON en ' r^-
fN
00 0 0
2 vo
"~
"- fN CN ON tN VO
r^ »—>
p ^ p vq >r> *}
ON S ^|- oo ON ^~l
£ p — — &
O
en — H ON en VO
^
00 §
O °° G> & <& & & & & g
c
CQ
(H
U
e
u
^S
*o
c
1
en
•o
sntage of the total emitte
V
S §3
1 s-
c! OJQ
1 >
C ni
1 $
i §5 'Sb
«2 f- . .8
1 cH o §
§ O £ E
O t-4 O ^
.0 ^ ^ S
1 If?
p ill
oo
00
•3
«
from Olmez i
1
u
o
w
March 2001
3B-9
DRAFT-DO NOT QUOTE OR CITE
-------
1 more reduced forms such as thiophenes and inorganic sulfides. A number of trace elements are
2 greatly enriched over crustal abundances in different fuels, such as Se in coal and V, Zn, and Ni in
3 oil. In fact, the higher V content of the fuel oil than in coal could help account for the higher
4 sulfate seen in the profiles from the oil-fired power plant compared to the coal-fired power plant
5 because V at combustion temperatures found in power plants is known to catalyze the oxidation
6 of reduced sulfur species. During combustion at lower temperatures, the emission of reduced
7 sulfur species also occurs. For example, Huffman et al. (2000) identified sulfur species emitted
8 by the combustion of several residual fuels oils (RFO) in a fire tube package boiler, which is
9 meant to simulate conditions in small institutional and industrial boilers. They found that sulfur
10 was emitted not only as sulfate (26 to 84%), but as thiophenes (13 to 39%) with smaller amounts
11 of sulfides and elemental S. They also found that Ni, V, Fe, Cu, Zn, and Pb are present mainly as
12 sulfates in emissions. Linak et al. (2000) found, when burning RFO, that the fire tube package
13 boiler produced particles with a bimodal size distribution in which about 0.2% of the mass was
14 associated with particles smaller than 0.1-jum AD, with the rest of the mass lying between 0.5 and
15 100 //m. Miller et al. (1998) found that larger particles consisted mainly of cenospheric carbon,
16 whereas trace metals and sulfates were found concentrated in the smaller particles in a fire tube
17 package boiler. In contrast, when RFO was burning in a refractory-lined combustor, which is
18 meant to simulate combustion conditions in a large utility residual oil fired boiler, Linak et al.
19 (2000) found that particles were distributed essentially unimodally, with a mean diameter of about,
20 0.1 //m.
21 Apart from emissions in the combustion of fossil fuels, trace elements are emitted as the
22 result of various industrial processes such as steel and iron manufacturing and nonferrous metal
23 production (e.g., for Pb, Cu, Ni, Zn, and Cd). As may be expected, emissions factors for the
24 various trace elements are highly source-specific (Nriagu and Pacyna, 1988). Inspection of
25 Table 3B-3 reveals that the emissions from the catalytic cracker and the oil-fired power plant are
26 greatly enriched in rare-earth elements such as La compared to other sources.
27 Emissions from municipal waste incinerators are heavily enriched in Cl arising mainly from
28 the combustion of plastics and metals that form volatile chlorides. The metals can originate from
29 cans or other metallic objects and some metals such as Zn and Cd are also additives in plastics or
30 rubber. Many elements such as S, Cl, Zn, Br, Ag, Cd, Sn, In, and Sb are enormously enriched
31 compared to then- crustal abundances. A comparison of the trace elemental composition of
March 2001
3B-10 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
incinerator emissions in Philadelphia, PA (shown in Table 3B-3), with the composition of
incinerator emissions in Washington DC, and Chicago, IL (Olmez et al., 1988), shows agreement
for most constituents to better than a factor of two.
Very limited data exist for characterizing the chemical composition of organic compounds
present in particulate emissions from industrial-scale stationary fuel combustion. Oros and
Simoneit (2000) have presented the abundance and distribution of organic constituents in coal
smokes that have been burned under laboratory conditions. This work provides the basis for
further investigation addressing the emissions of coal fired boilers.
Rogge et al. (1997a) measured the composition of the organic constituents in the particulate
matter emissions from a 50 billion kj/h boiler that was operating at 60% capacity and was burning
number 2 distillate fuel oil. The fine carbon particulate matter emissions from this boiler over
five tests were composed of an average of 14% organic carbon and 86% elemental carbon
(Hildemann et al., 1991). Significant variability in the distribution of organic compounds present
in the emissions from two separate tests was observed. Most of the identified organic mass
consisted of n-alkanonic acids, aromatic acids, n-alkanes, PAH, oxygeanted PAH, and chlorinated
compounds. It is unclear if these emissions are representative of typical fuel oil combustion units
in the United States. Rogge et al. (1997b) measured the composition of hot asphalt roofing tar
pots, and Rogge et al. (19.93b) measured the composition of emissions from home appliances that
use natural gas.
Motor Vehicles
Exhaust emissions of particulate matter from gasoline powered motor vehicles and diesel
powered vehicles have changed significantly over the past 25 years (Sawyer and Johnson, 1995;
Cadle et al., 1999). These changes have resulted from reformulation of fuels, the wide application
of exhaust gas treatment in gasoline-powered motor vehicles, and changes in engine design and
operation. Because of these evolving tailpipe emissions, along with the wide variability of
emissions between vehicles of the same class (Hildemann et al., 1991; Cadle et al., 1997; Sagebiel
et al., 1997; Yanowitz et al., 2000), well-defined average emissions profiles for the major classes
of motor vehicles have not been established. Two sampling strategies have been employed to
obtain motor vehicle emissions profiles: (1) the measurement of exhaust emissions from vehicles
operating on dynamometers and (2) the measurement of integrated emissions of motor vehicles
March 2001
3B-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 driving through roadway tunnels. Dynamometer testing can be used to measure vehicle emissions
2 operating over an integrated driving cycle and allows the measurement of emissions from
3 individual vehicles. However, dynamometer testing requires considerable resources and usually
4 precludes testing a very large number of vehicles. In contrast, a large number of vehicles can be
5 readily sampled in tunnels, but vehicles driving through tunnels operate over limited driving
6 conditions and the measurements represent contributions from a large number of vehicle types.
7 As a result, except in a few cases, runnel tests have not been effective at developing chemically
8 speciated particulate matter emissions profiles for individual motor vehicle classes. As a result,
9 several studies have measured the contribution of both organic and elemental carbon to the
10 particulate matter emissions from different classes of motor vehicles operating on chassis
11 dynamometers.
12 The principal components emitted by diesel and gasoline fueled vehicles are organic carbon
13 (OC) and elemental carbon (EC) as shown in Tables 3B-4a and 4b. As can be seen, the variability
14 among entries for an individual fuel type is large and overlaps that found between different fuel
15 types. On average, the abundance of elemental carbon is larger than that of organic carbon in the
16 exhaust of diesel vehicles, whereas organic carbon is the dominant species in the exhaust of
17 gasoline fueled vehicles. Per vehicle, total carbon emissions from light and heavy duty diesel
18 vehicles can range from 1 to 2 orders of magnitude higher than those from gasoline vehicles.
19 There appears to be a tendency for emissions of elemental carbon to increase relative to emissions
20 of organic carbon for gasoline fueled vehicles as simulated driving conditions are changed from a
21 steady 55 km/h to the various load conditions specified in the Federal Test Procedures (FTPs).
22 Also shown are the results of sampling from mixed vehicle types along roadsides and in tunnels.
23 As might be expected, most of the PM emitted by motor vehicles is in the PM2.5 size range.
24 Particles in diesel exhaust are typically trimodal consisting of a nuclei mode, an accumulation
25 mode and a coarse mode and are lognormal in form (Kittelson, 1998). More than 90% of the total
26 number of particles are in the nuclei mode, which contains only about 1 to 20% of the particle
27 mass with a mass median diameter of about 0.02 /an, whereas the accumulation mode (with a
28 mass median diameter of about 0.25 yum) contains most of the mass with a smaller fraction (5 to
29 20%) contained in the coarse mode. Kerminin et al. (1997), Bagley et al. (1998), and Kleeman
30 et al. (2000) also have shown that gasoline and diesel fueled vehicles produce particles that are
31 mostly less than 2.0 too. in diameter. Cadle et al. (1999) found that 91% of PM emitted by in-use
March 2001
3B-12 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 3B-4a. ORGANIC AND ELEMENTAL CARBON FRACTIONS OF DIESEL
AND GASOLINE ENGINE PARTICIPATE MATTER EXHAUST
Heavy-duty diesel engines3
Heavy-duty diesel engines (SPECIATE)b
Light-duty diesel engines0
Light-duty diesel engines (SPECIATE)"
Gasoline engines (hot stabilized)2
Gasoline engines ("smoker" and "high emitter")3'0
Gasoline engines (cold start)3
Organic Carbon
19 ±8%
21 - 36% .
30 ± 9%
22 - 43%
56 ±11%
76 ±10%
46 ± 14%
Elemental Carbon
75 ± 10%
52 - 54%
61 ± 16%
51 - 64%
25 ± 15%
7 ±6%
42 ± 14%
"Fujita et al. (1998) and Watson et al. (1998).
bU.S. EPA SPECIATE database.
"Norbeck et al. (1998).
Source: U.S. Environmental Protection Agency (1999).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
gasoline vehicles in the Denver area was in the PM2.5 size range, which increased to 97% for
"smokers" (i.e., light-duty vehicles with visible smoke emitted from their tailpipes) and 98% for
diesels. Durbin et al. (1999) found that about 92% of the PM was smaller than 2.5 ^m for
smokers and diesels. The mass median diameter of the PM emitted by the gasoline vehicles
sampled by Cadle et al. (1999) was about 0.12 ,um, which increased to 0.18 /^m for smokers and
diesels. Corresponding average emissions rates of PM25 found by Cadle et al. (1999) for diesels
were 552 mg/mi; for smokers they were 222 mg/mi; and, for gasoline vehicles, they were
38 mg/mi. The values for smokers and for diesels appear to be somewhat lower than those given
in Table 3B-5, whereas the value for gasoline vehicles falls in the range given for low and
medium gasoline vehicle emissions.
Examples of data for the trace elemental composition of the emissions from a number of
vehicle classes obtained as part of the North Frontal Range Air Quality Study (NFRAQS), which
took place in December 1997 in Colorado are shown in Table 3B-5. As can be seen from
Table 3B-5, emissions of total carbon (TC), which is equal to the sum of organic carbon (OC) and
elemental carbon (EC), from gasoline vehicles are highly variable. Gillies and Gertler (2000)
point out that there is greater variability in the concentrations of trace elements and ionic
March 2001
3B-13
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 3B-4b. CONTRIBUTION OF ORGANIC CARBON TO PARTICIPATE
MATTER CARBON EMISSIONS IN MOTOR VEHICLE EXHAUST COLLECTED
FROM VEHICLES OPERATED ON CHASSIS DYNAMOMETERS
GASOLINE POWERED VEHICLES
Light-duty vehicles
High-CO/VOC-emitting smokers
High-CO/VOC-emitting nonsmokers
Catalyst-equipped vehicles
Noncatalyst vehicles
DIESEL VEHICLES
Light-duty diesel vehicles
Medium-duty diesel vehicles
Heavy-duty diesel vehicles
Heavy-duty diesel vehicles
Year of Tests
1996-97
1994
1994
Mid-1980s
Mid-1980s
1996-1997
1996
1992
Mid-1980s
Test Cycle
FTP
IM-240
IM-240
FTP
FTP
FTP
FTP
c
c
Number of
Vehicles
195a
7
15
7
6
195a
2
6
2
OC%of
Total Carbon
70
91
76
69
89
40
50"
42
45
Notes
A
B
B
C
C
A
D
E
C
Notes:
A. From (Cadle et al., 1999). Average of summer and winter cold start emissions.
B. From (Sagebiel et al., 1997). Hot start testing of vehicles identified as either high emitters of carbon
monoxide or volatile organic compounds (VOCs).
C. From (Hildemann et al., 1991). Cold start tests.
D. From (Schauer et al., 1999). Hot start tests of medium duty vehicles operating on an FTP cycle.
E. From (Lowenthal et al., 1994). Only includes measurement of vehicles powered by diesel fuel operated
without an exhaust particulate trap.
"A total of 195 light duty vehicles were tested that include both gasoline powered vehicles and diesel powered
vehicles.
bFraction of particulate matter consisting of organic carbon was measured with and without an organics denuder
upstream of particulate filter. Results reported here represent measurement without an organics denuder for
consistency with other measurements. Using an organics denuder, the organic carbon comprised 39% of the
particulate matter carbon.
'Driving cycle comprised of multiple idle, steady acceleration, constant speed, deceleration steps (see reference
for more details).
1 species than for OC and EC among different source profiles (e.g., SPECIATE, Lawson and Smith
2 (1998), Norbeck et al. (1998)). They suggest that this may arise because their emissions are not
3 related only to the combustion process, but also to their abundances in different fuels and
March 2001
3B-14
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 3B-5. EMISSION RATES (mg/mi) FOR CONSTITUENTS OF PARTICIPATE
MATTER FROM GASOLINE AND DIESEL VEHICLES
Gasoline Vehicles
TC
OC
EC
N03"
SO4=
Na
Mg
Al
Si
P
S
Cl
K
Ca
Fe
Ni
Cu
Zn
Br
Ba
Pb
Low
9.07 ± 0.75
6.35 ± 0.54
2.72 ± 0.52
0.039 ± 0.027
0.1 58 ±0.036
0.060 ± 0.063
0.036 ± 0.022
0.083 ±0.016
0.066 ± 0.008
0.035 ± 0.004
0.085 ± 0.006
0.024 ±0.01 2
0.010 ±0.009
0.060 ±0.010
0.1 43 ±0.004
0.001 ±0.004
0.002 ± 0.004
0.048 ± 0.003
0.001 ± 0.002
0.01 3 ±0.1 36
0.007 ±0.006
Medium
41. 30 ±1.68
26.02 ±1.31
15.28 ±0.99
0.057 ± 0.028
0.51 8 ±0.043
0.023 ±0.1 11
0.068 ± 0.027
0.078 ±0.01 6
0.279 ± 0.01 1
0.152 ±0.007
0.442 ±0.009
0.038 ±0.012
0.01 9 ±0.009
0.212 ±0.011
0.756 ± 0.005
0.005 ± 0.004
0.01 6 ±0.003
0.251 ±0.004
0.01 6 ±0.002
0.009 ±0.1 38
0.085 ± 0.005
High
207.44 ± 7.29
95.25 ± 4.28
112.19 ±5.82
0.141 ±0.031
0.651 ±0.052
0.052 ± 0.092
0.041 ± 0.033
0.057 ±0.014
0.714 ±0.012
0.1 13 ±0.007
0.822 ± 0.022
0.081 ±0.020
0.031 ±0.035
0.210 ±0.030
1.047 ±0.010
0.01 1 ± 0.005
0.021 ±0.005
0.265 ± 0.023
0.079 ± 0.003
0.011 ±0.299
0.255 ± 0.008
Smoker
456.38 ± 16.80
350.24 ±15.27
106.14 ±5.42
0.964 ±0.051
2.160 ±0.137
0.000 ± 0.000
0.000 ± 0.000
0.000 ± 0.000
0.000 ± 0.000
0.000 ± 0.000
2.515 ±0.116
0.140 ±0.1 17
0.033 ± 0.386
0.362 ±0.250
2.438 ± 0.054
0.008 ±0.017
0.071 ±0.01 8
0.1 88 ±0.272
0.047 ±0.012
0.380 ±2.175
0.345 ± 0.032
Diesel
Light Duty
373.43 ±13.75
132.01 ± 5.82
241.42 ±12.11
1.474 ±0.071
2.902 ±0.1 65
0.000 ± 0.000
0.000 ± 0.000
0.000 ± 0.000
0.000 ± 0.000
0.000 ± 0.000
2.458 ±0.124
0.228 ±0.1 14
0.000 ± 0.426
0.1 50 ±0.304
0.515 ±0.057
0.014 ±0.018
0.024 ±0.021
0.000 ± 0.299
0.003 ±0.014
0.428 ± 2.390
0.1 53 ±0.033
Vehicles
Heavy Duty
1570.69 ±58.24
253.94 ±16.12
1316.75 ±55.33
1.833 ±1.285
3.830 ±1.286
1.288 ±2. 160
1.061 ±0.729
0.321 ±0.543
8.01 8 ±0.221
0.407 ±0.1 36
3.717±0.111
0.881 ±0.221
0.064 ± 0.248
0.716 ±0.107
0.376 ± 0.055
0.002 ± 0.057
0.001 ±0.062
0.707 ± 0.032
0.01 2 ±0.050
0.493 ±3. 108
0.008 ±0.1 54
Source: Lawson and Smith (1998).
1 lubricants and also to wear and tear during vehicle operation. Emissions from smokers are
2 comparable to those from diesel vehicles. Thus, older, poorly maintained gasoline vehicles could
3 be significant sources of PM2 5 (Sagebiel et al., 1997; Lawson and Smith, 1998), in addition to
March 2001
3B-15
DRAFT-DO NOT QUOTE OR CITE
-------
1 being significant sources of gaseous pollutants (e.g., Calvert et al., 1993). Durbin et al. (1999)
2 point out that although smokers constitute only 1.1 to 1.7% of the light-duty fleet in the South
3 Coast Air Quality Management District in California, they contribute roughly 20% of the total PM
4 emissions from the light-duty fleet. In general, motor vehicles that are high emitters of
5 hydrocarbons and carbon monoxide also will tend to be high emitters of PM (Sagebiel et al.,1997;
6 Cadle et al., 1997). Particle emission rates also are correlated with vehicle acceleration and
7 emissions occur predominantly during periods of heavy acceleration, even in newer vehicles
8 (Maricq et al., 1999).
%.
9 Although the data shown in Table 3B-5 indicate that S (mainly in the form of sulfate) is a
10 minor component of PM2 s emissions, S may be the major component of the ultrafine particles that
11 are emitted by either diesel or internal combustion engines (Gertler et al., 2000). It is not clear
12 what the source of the small amount of Pb seen in the auto exhaust profile is. It is extremely
13 difficult to find suitable tracers for automotive exhaust because Pb has been removed from
14 gasoline. However, it also should be remembered that restrictions in the use of leaded gasoline
15 have resulted in a dramatic lowering of ambient Pb levels.
16 Several tunnel studies have measured the distribution of organic and elemental carbon in the
17 integrated exhaust of motor vehicle fleets comprising several classes of motor vehicles (Pierson
18 and Brachaczek, 1983; Weingartner et al., 1997a; Fraser.et al., 1998a). The study by Fraser et al.
19 (1998a) found that organic carbon constituted 46% of the carbonaceous particulate matter
20 emissions from the vehicles operating in the Van Nuys tunnel in Southern California in the
21 Summer of 1993. Although diesel vehicles constituted only 2.8% of the vehicles measured by
22 Fraser et al. (1998a), the contribution of the organic carbon to the total particulate carbon
23 emissions obtained in the Van Nuys tunnels is in reasonable agreement with the dynamometer
24 measurements shown in Table 3B-4b.
25 Very few studies have reported comprehensive analyses of the organic composition of motor
26 vehicle exhaust. The measurements by Rogge et al. (1993c) are the most comprehensive, but are
27 not expected to be the best representation of current motor vehicle emissions because these
28 measurements were made in the mid-1980s. Measurements reported by Fraser et al. (1999) were
29 made in a tunnel study conducted in 1993 and represent integrated diesel and gasoline powered
30 vehicle emissions. In addition, exhaust emissions from two medium-duty diesel vehicles
31 operating over an FTP cycle were analyzed by Schauer et al. (1999). A unique feature of both the
March 2001
3B-16
DRAFT-DO NOT QUOTE OR CITE
-------
1 measurements by Faser et al. (1999) and Schauer et al. (1999) is that they include the
2 quantification of unresolved complex mixture (UCM), which comprises aliphatic and cyclic
3 hydrocarbons that cannot be resolved by gas chromatography (Schauer et al., 1999). Schauer
4 etal. (1999) have shown that all of the organic compound mass in their diesel exhaust samples
5 could be extracted and eluted by CG/MS techniques, even though not all of the organic compound
6 mass can identified on a single compound basis. Table 3B-6 summarizes the composition of
7 motor vehicle exhaust
8
9
TABLE 3B-6.
Source
Gasoline and diesel-
powered vehicles
driving through the
Van Nuys Tunnel
(Fraser et al., 1999)a
Medium-duty diesel
vehicles operated over
an FTP Cycle
(Schauer et al., 1999)
measured by Fraser et al.
(1999) and Schauer
etal. (1999).
SUMMARY OF PARTICLE-PHASE ORGANIC COMPOUNDS
EMITTED FROM
Compound Class
n-Alkanes
Petroleum biomarkers
PAH
Aromatic acids
Aliphatic acids
Substituted aromatic
UCM"
n-Alkanes
Petroleum biomarkers
PAH
Aliphatic acids
Aromatic acids
Saturated cycloalkanes
UCMb
MOTOR VEHICLES
Contribution to
Particulate Mass (%)
0.009
0.078
0.38
0.29
0.21
0.042
23.0
0.22
0.027
0.54
0.24
0.014
0.037
22.2
Dominant Contributors to
Emissions of Compound Class
C2, through C29
Hopanes and steranes
No dominant compound
Benzenedicarboxylic acids
Palmitic and stearic acids
No dominant compound
C20 through C2g
Hopanes and steranes
No dominant compound
n-Octadecanoic acid
Methylbenzoic acid
C2i through C25
"Includes emissions of brake wear, tire wear, and resuspension of road dust associated with motor vehicle traffic.
bUnresolved complex mixture.
March 2001
3B-17
DRAFT-DO NOT QUOTE OR CITE
-------
1 Several studies have measured the distribution of polycyclic aromatic hydrocarbons (PAHs)
2 in motor vehicles exhaust from on-road vehicles (Westerholm et al., 1991; Lowenthal et al., 1994;
3 Venkataraman et al., 1994; Westerholm and Egeback, 1994; Reilly et al., 1998; Cadle et al., 1999,
4 Weingartner et al., 1997b; Marr et al., 1999). Cadle et al. (1999) found high molecular weight
5 PAHs (PAHs with molecular weights greater than or equal to 202 g/mole) to make up from 0.1 to
6 7.0% of the particulate matter emissions from gasoline powered and diesel powered light duty
7 vehicles. It is important to note, however, that PAHs with molecular weights of
8 202 (fluoranthene, acephenanthrylene, and pyrene), 226 (benzo[ghi]fluoranthene and
9 cyclopenta[cd]pyrene), and 228 (benz[a]anthracene, chrysene, and triphenylene) exist in both the
10 gas-phase and particle-phase at atmospheric conditions (Fraser et al., 1998b). Excluding these
11 semi-volatile PAHs, the contribution of nonvolatile PAHs to the particulate matter emitted from
12 the light-duty vehicles sampled by Cadle etal. (1999) ranges from 0.013 to 0.18%. These
13 measurements are in good agreement with the tunnel study conducted by Fraser et al. (1999) and
14 the heavy-duty diesel truck and bus exhaust measurements by Lowenthal et al. (1994), except that
15 the nonvolatile PAH emissions from the heavy duty diesel vehicles tested by Lowenthal et al.
16 (1994) were moderately higher, making up approximately 0.30% of the particulate matter mass
17 emissions.
18
19 Biomass Burning
20 In contrast to the mobile and stationary sources discussed earlier, emissions from biomass
21 burning in woodstoves and forest fires are strongly seasonal and can be highly episodic within
22 then: peak emissions seasons. The burning of fuelwood is confined mainly to the winter months
23 and is acknowledged to be a major source of ambient air particulate matter in the northwestern
24 United States during the heating season. Forest fires occur primarily during the driest seasons of
25 the year in different areas of the country and are especially prevalent during prolonged droughts.
26 PM produced by biomass burning outside the United States (e.g., in Central America during the
27 spring of 1988) also can affect ambient air quality in the United States.
28 An example of the composition of fine particles (PM2 5) produced by woodstoves is shown
29 in Figure 3B-4. These data were obtained in Denver during the winter of 1987-1988 (Watson and
30 Chow, 1994). As was the case for motor vehicle emissions, organic and elemental carbon are the
31 major components of particulate emissions from wood burning. It should be remembered
March 2001
3B-18
DRAFT-DO NOT QUOTE OR CITE
-------
Chemical Species
Figure 3B-4. Chemical abundances for PM2 5 emissions from wood burning in Denver, CO.
Solid bars represent fractional abundances, and the error bars represent
variability in species abundances. Error bars represent detection limits when
there are no solid bars.
Source: Watson and Chow (1994).
1 that the relative amounts shown for organic carbon and elemental carbon vary with the type of
2 stove, the stage of combustion and the type and condition of the fuelwood. Fine particles are
3 dominant in studies of wood burning emissions. For instance, the mass median diameter of wood-
4 smoke particles was found to be about 0.17 fj.m in a study of the emissions from burning
5 hardwood, softwood, and synthetic logs (Dasch, 1982).
6 Kleeman et al. (1999) showed that the particles emitted by the combustion of wood in
7 fireplaces are predominately less than 1.0 /^m in diameter, such that the composition of fine
8 particulate matter (PM2 5) emitted from fireplace combustion of wood is representative of the total
9 particulate matter emissions from this source. Hildemann et al. (1991) and McDonald et al.
10 (2000) reported that smoke from fireplace and wood stove combustion consists of 48% to
11 71% OC and 2.9% to 15% EC. Average elemental and organic carbon contents for these
12 measurements are shown in Table 3B-7. It should be noted that the two methods used for the
March 2001
3B-19
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 3B-7. MASS EMISSIONS, ORGANIC CARBON, AND ELEMENTAL
CARBON EMISSIONS FROM RESIDENTIAL COMBUSTION OF WOOD
Wood Type
Softwood
Softwood
Hardwood
Hardwood
Hardwood
Combustion
Type
Fireplace
Fireplace
Fireplace
Fireplace
Wood Stove
Average Mass
Emission Rate
(gkg"' of wood
burned)
13.0
5.14
5.28
5.66
3.96
Number
of Tests
2
5
3
5
8
Percent
Organic
Carbon3
48.4
58.5
. 48.4
63.2
71.2
Percent
Elemental
Carbon3
5.2
15.0
2,9
7.0
9.0
References
Hildemannetal. (1991)
McDonald et al. (2000)
Hildemann et al. (1991)
McDonald et al. (2000)
McDonald et al. (2000)
•Hildemann et al. (1991) used the method described by Birch and Cary (1996) to measure EC and McDonald
et al. (2000) used the method reported by Chow et al. (1993) to measure OC.
1 measurements shown in Table 3B-7 have been reported to produce different relative amounts of
2 OC and EC for wood smoke samples, but show good agreement for total carbon (OC + EC)
3 measurements (Chow et al., 1993).
4 Hawthorne et al. (1988) and Hawthorne et al. (1989) measured gas-phase and particle-phase
5 derivatives of guaiacol (2-methoxyphenol), syringol (2,6-dimethoxyphenol), phenol, and catechol
6 (1,2-benzenediol) in the downwind plume of 28 residential wood stoves and fireplaces. Rogge
7 et al. (1998) reported a broad range of particle-phase organic compounds in the wood smoke
8 samples collected by Hildemann et al. (1991), which include n-alkanes, n-alkanoic acids,
9 n-alkenoic acids, dicarboxylic acids, resin acids, phytosterols, polycyclic aromatic hydrocarbons
10 (PAH), and the compounds reported by Hawthorne et al. (1989). Supplementing these
11 measurements, McDonald et al. (2000) reported the combined gas-phase and particle-phase
12 emissions of PAH and the compounds quantified by Hawthorne et al. (1989). The measurements
13 by Rogge et al. (1998), which represent a comprehensive data set of the organic compounds
14 present in wood smoke aerosol, are summarized in Table 3B-8. It should be noted, however, that
15 these nearly 200 compounds account for only approximately 15 to 25% of the organic carbon
16 particle mass emitted from the residential combustion of wood. Simoneit et al. (1999) have
17 shown that levoglucosan constitutes a noticeable portion of the organic compound mass not
18 identified by Rogge et al. (1998). In addition, Elias et al. (1999) used high-temperature gas
March 2001
3B-20
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 3B-8. SUMMARY OF PARTICLE-PHASE ORGANIC COMPOUNDS
EMITTED FROM THE COMBUSTION OF WOOD IN FIREPLACES
(Measurements were made using a dilution sampler and no
semi-volatile organic compound sorbent.)
Biomass Type
Fireplace
combustion of
softwood
Fireplace
combustion of
hardwood
Contribution to Particulate Dominant Contributors to Emissions
Compound Class Mass (%) of Compound Class
n-Alkanes
n-Alkanoic acids
n-Alkenoic acids
Dicarboxylic acids
Resin acids
Substituted phenols
Phytosterols
PAH
Oxygenated PAH
n-Alkanes
n-Alkanoic acids
n-Alkenoic acids
Dicarboxylic acids
Resin acids
Substituted phenols
Phytosterols
PAH
Oxygenated PAH
0.039
0.45
0.12
0.36
1.28
3.30
0.37
0.092
0.019
0.044
1.33
0.049
0.42
0.11
8.23
0.21
0.13
0.020
C2) through C31
Cl6> Cl8> C20, C2], C22) C24
Oleic and linoleic acid
Malonic acid
Abietic, dehydroabietic, isopimaric,
pimaric, and sandaracopimaric acids
Benzenediols and guaiacols
P-Sitosterol
Fluoranthene and pyrene
IH-phenalen-l-one
C21 through C29
C|6> C22, C24, C26
Oleic and linoleic acid
Succinic acid
Dehydroabietic acid
Benzediols, guaiacols, and syringols
P-sitosterol
No dominant compounds
1 H-phenalen-1 -one
Source: Roggeetal. (1998).
1 chromatography/mass spectrometry (HTGC-MS) to measure high-molecular-weight organic
2 compounds in smoke from South American leaf and steam litter biomass burning. These
3 compounds cannot be measured by the analytical techniques employed by Rogge et al. (1998) and,
4 therefore, are strong candidates to make up some of the unidentified organic mass in the wood
March 2001
3B-21
DRAFT-DO NOT QUOTE OR CITE
-------
1 smoke samples analyzed by Rogge et al. (1998). These compounds, which include triterpenyl
2 fatty acid esters, wax esters, triglycerides, and high-molecular-weight n-alkan-2-ones, are
3 expected to be present in North American biomass smoke originating from agricultural burning,
4 forest fires, grassland fires, and wood smoke.
5 Measurements of aerosol composition, size distributions, and aerosol emissions factors have
6 been made'in biomass burning plumes either on towers (Susott et al., 1991) or aloft on fixed-wing
7 aircraft (e.g., Radke et al., 1991) or on helicopters (e.g., Gofer et al., 1988). As was found for
8 woodstove emissions, the composition of biomass burning emissions is strongly dependent on the
9 stage of combustion (i.e., flaming, smoldering, or mixed), and the type of vegetation (e.g., forest,
10 grassland, scrub). Over 90% of the dry mass in particulate biomass burning emissions is
11 composed of organic carbon (Mazurek et al., 1991). Ratios of organic carbon to elemental carbon
12 are highly variable, ranging from 10:1 to 95:1, with the highest ratio found for smoldering
13 conditions and the lowest for flaming conditions. Emissions factors for total particulate emissions
14 increase by factors of two to four in going from flaming to smoldering stages in the individual
15 fires studied by Susott et al. (1991).
16 Particles in biomass burning plumes from a number of different fires were found to have
17 three distinguishable size modes, (1) a nucleation mode, (2) an accumulation mode, and
18 (3) a coarse mode (Radke et al., 1991). Based on an average of 81 samples, approximately 70%
19- of the mass was found in particles <3.5 //m in aerodynamic diameter. The fine particle
20 composition was found to be dominated by tarlike, condensed hydrocarbons and the particles were
21 usually spherical in shape. Additional information for the size distribution of particles produced
22 by vegetation burning was shown in Figure 3B-2.
23 An example of ambient data for the composition of PM25 collected at a tropical site that was
24 heavily affected by biomass burning is shown in Table 3B-9. The samples were collected during
25 November of 1997 on the campus of Sriwijaya University, which is located in a rural setting on
26 the island of Sumatra in Indonesia (Pinto et al., 1998). The site was subjected routinely to levels
27 of PM2 5 well in excess of the U.S. NAAQS as a result of the Indonesian biomass fires from the
28 summer of 1997 through the spring of 1988. As can be seen from a comparison of the data shown
29 in Table 3B-9 with those shown in Figure 3B-4, there are a number of similarities and differences
30 (especially with regard to the heavy metal content) in the abundances of many species. The
March 2001
3B-22
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 3B-9. MEAN AEROSOL COMPOSITION AT TROPICAL SITE
(SRIWIJAYA UNIVERSITY, SUMATRA, INDONESIA) AFFECTED
HEAVILY BY BIOMASS BURNING EMISSIONS3
Component
OC
EC
S04=
Al
Si
Cl
K
Ca
Ti
v
Abundance (%)
76
1.2
11
BDb
9.3 x lO'2
4.4
0.7
4.5 x IQ-2
4.2 x IQ-3
BDb
Component
Cr
Mn
Fe
Ni
Cu
Zn
As
Se
Br
Pb
Abundance (%)
BDb
BDb
3.9 x lO'2
<3.8 x IQ-5
4.8 x 1Q-4
3.1 x IQ-3
6.4 x 1Q-4
2.8 x IQ-4
3.6 x lO'2
3.1 x 1Q-3
aThe mean PM2 5 concentration during the sampling period (November 5 through 11, 1997) was 264 //g/m3.
bBeneath detection limit.
Source: Pinto et al. (1998).
1
2
3
4
5
6
7
8
9
10
11
abundances of some crustal elements (e.g., Si, Fe) are higher in Table 3B-9 than in Figure 3B-4,
perhaps reflecting additional contributions of entrained soil dust.
Limited emissions data that includes organic compound speciation information have been
reported for agricultural burning (Jenkins et al., 1996), forest fires (Simoneit, 1985), and grassland
burning (Standley and Simoneit, 1987). Jenkins et al. (1996) present PAH emissions factors for
the combustion of cereals (barley, corn, rice, and wheat), along with PAH emissions factors for
wood burning. Profiles of organic compounds in emissions from meat cooking (Rogge et al.,
1991) and cigarette smoke (Rogge et al., 1994) have been obtained.
March 2001
3B-23
DRAFT-DO NOT QUOTE OR CITE
-------
1 Natural Sources
2 Although sea-salt aerosol production is confined to salt water bodies, it is included here
3 because many marine aerosols can exert a strong influence on the composition of the ambient
4 aerosol in coastal areas. In some respects, the production of sea-salt aerosols is like that of
5 windblown dust in that both are produced by wind agitation of the surface. The difference
6 between the two categories arises because sea-salt particles are produced from the bursting of air
7 bubbles rising to the sea surface. Air bubbles are formed by the entrainment of air into the water
8 by breaking waves. The surface energy of a collapsing bubble is converted to kinetic energy in
9 the form of a jef of water that can eject drops above the sea surface. The mean diameter of the jet
10 drops is about 15% of the bubble diameter (Wu, 1979). Bubbles in breaking waves range in size
11 from a few yum to several mm in diameter. Field measurements by Johnson and Cooke (1979) of
12 bubble size spectra show maxima in diameters at around 100 //m, with the bubble size distribution
13 varying as (d/do)'5 with d0 = 100 //m.
14 Because sea-salt particles receive water from the surface layer, which is enriched in organic
15 compounds, the aerosol drops are composed of this organic material in addition to sea salt (about
16 3.5% by weight in sea water). Na+ (30.7%), Cl" (55.0%), SO4= (7.7%), Mg2+ (3.6%), Ca2+ (1.2%),
17 K* (1.1%), HCO3" (0.4%), and Br" (0.2%) are the major ionic species by mass in sea water
18 (Wilson, 1975). The composition of the marine aerosol also reflects the occurrence of
19. displacement reactions that enrich sea-salt particles in SO4" and NO3", while depleting them of Cl"
20 and Br.
21 Seasalt is concentrated in the coarse size mode with a mass median diameter of about 7 //m
22 for samples collected in Florida, the Canary Islands, and Barbados (Savoie and Prospero, 1982).
23 The size distribution of sulfate is distinctly bimodal. Sulfate in the coarse mode is derived from
24 sea water but sulfate in the submicron aerosol arises from the oxidation of dimethyl sulfide
25 (CH3SCH3) or DMS. DMS is produced during the decomposition of marine micro-organisms.
26 DMS is oxidized to methane sulfonic acid (MSA), a large fraction of which is oxidized to sulfate
27 (e.g., Herteletal., 1994).
28 Apart from sea spray, other natural sources of particles include the suspension of organic
29 debris and volcanism. Profiles of organic compounds in vegetative detritus have been obtained by
30 Rogge et al. (1993d). Particles are released from plants in the form of seeds, pollen, spores, leaf
31 waxes, and resins, ranging in size from 1 to 250 yum (Warneck, 1988). Fungal spores and animal
March 2001
3B-24
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
debris, such as insect fragments, also are to be found in ambient aerosol samples in this size range.
Although material from all the foregoing categories may exist as individual particles, bacteria
usually are found attached to other dust particles (Warneck, 1988). Smaller bioaerosol particles
include viruses, individual bacteria, protozoa, and algae (Matthias-Maser and Jaenicke, 1994).
hi addition to natural sources, other sources of bioaerosol include industry (e.g., textile mills),
agriculture, and municipal waste disposal (Spendlove, 1974). The size distribution of bioaerosols
has not been characterized as well as it has for other categories.
Trace metals are emitted to the atmosphere from a variety of sources such as sea spray,
wind-blown dust, volcanoes, wildfires and biotic sources (Nriagu, 1989). Biologically mediated
volatilization processes (e.g., biomethylation) are estimated to account for 30 to 50% of the
worldwide total Hg, As, and Se emitted annually, whereas other metals are derived principally
from pollens, spores, waxes, plant fragments, fungi, and algae. It is not clear, however, how much
of the biomethylated species are remobilized from anthropogenic inputs. Median ratios of the
natural contribution to globally averaged total sources for trace metals are estimated to be
0.39 (As), 0.15 (Cd), 0.59 (Cr), 0.44 (Cu), 0.41 (Hg), 0.35 (Ni), 0.04 (Pb), 0.41 (Sb), 0.58 (Se),
0.25 (V), and 0.34 (Zn), suggesting a not insignificant natural source for many trace elements.
It should be noted though that these estimates are based on emissions estimates that have
uncertainty ranges of an order of magnitude.
March 2001
3B-25
DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Bagley, S. T.; Gratz, L. D.; Johnson, J. H.; McDonald J. F. (1998) Effects of an oxidation catalytic converter and a
3 biodiesel fuel on the chemical, mutagenic, and particle size characteristics of emissions from a diesel engine.
4 Environ. Sci. Technol. 32: 1183-1191.
5 Birch, M. E.; Gary, R. A. (1996) Elemental carbon-based method for occupational monitoring of particulate diesel
6 exhaust: methodology and exposure issues. Analyst 121: 1183-1190.
7 Cadle, S. H.; Mulawa, P. H.; Ball, J.; Donase, C.; Weibel, A.; Sagebiel, J. C.; Knapp, K. T.; Snow, R. (1997)
8 Particulate emission rates from in-use high-emitting vehicles recruited in Orange county, California. Environ.
9 . Sci. Technol. 31:3405-3412.
10 Cadle, S. H.; Mulawa, P.; Hunsanger, E. C.; Nelson, K.; Ragazzi, R. A.; Barrett, R.; Gallagher, G. L.; Lawson, D. R.;
11 Knapp, K. T.; Snow, R. (1999) Light-duty motor vehicle exhaust particulate matter measurement in the
12 Denver, Colorado, area. J. Air Waste Manage. Assoc. 49: PM-164-174.
13 Calvert, J. G.; Heywood, J. B.; Sawyer, R. F.; Seinfeld, J. H. (1993) Achieving acceptable air quality: some
14 reflections on controlling vehicle emissions. Science (Washington, DC) 261: 37-45.
15 Chow, J. C.; Watson, J. G.; Pritchett, L. C.; Pierson, W. R.; Frazier, C. A.; Purcell, R. G. (1993) The DRI
16 thermal/optical reflectance carbon analysis system: description, evaluation and applications in U.S. air quality
17 studies. In: Lodge, J. P., Jr. Fourth international conference on carbonaceous particles in the atmosphere;
18 April 1991; Vienna, Austria. Atmos. Environ. Part A 27: 1185-1201.
19 Chow, J. C.; Watson, J. G.; Houck, J. E.; Pritchett, L. C.; Rogers, C. F.; Frazier, C. A.; Egami, R. T.; Ball, B. M.
20 (1994) A laboratory resuspension chamber to measure fugitive dust size distributions and chemical
21 compositions. Atmos. Environ. 28: 3463-3481.
22 Gofer, W. R., Ill; Levine, J. S.; Sebacher, D. I.; Winstead, E. L.; Riggin, P. J.; Brass, J. A.; Ambrosia, V. G. (1988)
23 Particulate emissions from a mid-latitude prescribed chaparral fire. J. Geophys. Res. [Atmos.] 93: 5207-5212.
24 Dasch, J. M. (1982) Particulate and gaseous emissions from wood-burning fireplaces. Environ. Sci. Technol.
25 16:639-645.
26 Durbin, T. D.; Smith, M. R.; Norbeck, J. M.; Truex, T. J. (1999) Population density, particulate emission
27 characterization, and impact on the particulate inventory of smoking vehicles in the South Coast Air Quality
28 Management District. J. Air Waste Manage. Asssoc. 49: 28-38.
29 Elias, V. O.; Simoneit, B. R. T.; Pereira, A. S.; Cabral, J. A.; Cardoso, J. N. (1999) Detection of high molecular
30 weight organic tracers in vegetation smoke samples by high-temperature gas chromatography-mass
31 spectrometry. Environ. Sci. Technol. 33: 2369-2376.
32. Fraser, M. P.; Cass, G. R.; Simoneit, B. R. T. (1998a) Gas-phase and particle-phase organic compounds emitted from
33 motor vehicle traffic in a Los Angeles roadway tunnel. Environ. Sci. Technol. 32: 2051 -2060.
34 Fraser, M. P.; Cass, G. R.; Simoneit, B. R. T.; Rasmussen, R. A. (1998) Air quality model evaluation data for
35 organics. 5. C6-C22 nonpolar and semipolar aromatic compounds. Environ. Sci. Technol. 32: 1760-1770.
36 Fraser, M. P.; Cass, G. R.; Simoneit, B. R. T. (1999) Particulate organic compounds emitted from motor vehicle
37 exhaust and in the urban atmosphere. Atmos. Environ. 33: 2715-2724.
38 Fujita, E.; Watson, J. G.; Chow, J. C.; Robinson, N. F.; Richards, L. W.; Kumar, N. (1998) Northern front range air
39 quality study, volume C: source apportionment and simulation methods and evaluation. Reno, NV: Desert
40 Research Institute; prepared for Colorado State University, Cooperative Institute for Research in the
41 Atmosphere.
42 Gertler, A. W.; Abu-Allaban, M.; Coulombe, W.; Gilles, J. A.; Pierson, W. R.; Rogers, C. F.; Sagebiel, J. C.; Tarnay,
43 L. (2000) Preliminary results of a tunnel study to characterize mobile source particulate emissions. Presented
44 at: PM2000: particulate matter and health—the scientific basis for regulatory decision-making, specialty
45 conference & exhibition; January; Charleston, SC. Pittsburgh, PA: Air & Waste Management Association.
46 Gillies, J. A.; Gertler, A. W. (2000) Comparison and evaluation of chemically speciated mobile source PM2 5
47 particulate matter profiles. J. Air Waste Manage. Assoc. 50: 1459-1480.
48 Hawthorne, S. B.; Miller, D. J.; Barkley, R. M.; Krieger, M. S. (1988) Identification of methoxylated phenols as
49 candidate tracers for atmospheric wood smoke pollution. Environ. Sci. Technol. 22:1191-1196.
50 Hawthorne, S. B.; Krieger, M. S.; Miller, D. J.; Mathiason, M. B. (1989) Collection and quantitation of methoxylated
51 phenol tracers for atmospheric pollution from residential wood stoves. Environ. Sci. Technol. 23:470-475.
52 Hertel, O.; Christensen, J.; Hov, 0. (1994) Modelling of the end products of the chemical decomposition of DMS in
53 the marine boundary layer. Atmos. Environ. 28:2431 -2449.
March 2001
3B-26
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
Hildemann, L; M.; Markowski, G. R.; Jones, M. C.; Cass, G. R. (1991) Submicrometer aerosol mass distributions of
emissions from boilers, fireplaces, automobiles, diesel trucks, and meat-cooking operations Aerosol Sci
Technol. 14: 138-152.
Hopke, P. K.; Lamb, R. E.; Natusch, D. F. S. (1980) Multielemental characterization of urban roadway dust Environ
Sci. Technol. 14: 164-172.
Houck, J. E.; Chow, J. C.; Watson, J. G.; Simons, C. A.; Pritchett, L. C.; Goulet, J. M.; Frazier, C. A. (1989)
Determination of particle size distribution and chemical composition of particulate matter from selected
sources in California: volume I and executive summary (final report). Sacramento, CA: California Air
Resources Board.
Houck, J. E.; Goulet, J. M.; Chow, J. C.; Watson, J. G.; Pritchett, L. C. (1990) Chemical characterization of emission
sources contributing to light extinction. In: Mathai, C. V., ed. Visibility and fine particles: an A&WMA/EPA
international specialty conference; October 1989; Estes Park, CO. Pittsburgh, PA: Air & Waste Management
Association; pp. 437-446. (A&WMA transactions series no. TR-17).
Huffman, G. P.; Huggins, F. E.; Shah, N.; Huggins, R.; Linak, W. P.; Miller, C. A.; Pugmire, R. J.; Meuzelaar, H. L.
C.; Seehra, M. S.; Manivannan, A. (2000) Characterization of fine particulate matter produced by combustion
of residual fuel oil. J. Air Waste Manage. Assoc. 50: 1106-1114. ;
Jenkins, B. M.; Jones, A. D.; Turn, S. Q.; Williams, R. B. (1996) Particle concentrations, gas-particle partitioning,
and species intercorrelations for polycyclic aromatic hydrocarbons (PAH) emitted during biomass burning'
Atmos. Environ. 30: 3825-3835.
Johnson, B. D.; Cooke, R. C. (1979) Bubble populations and spectra in coastal waters: a photographic approach
J. Geophys. Res. C: Oceans Atmos. 84: 3761-3766.
Kerminen, V. M.; Makela, T. E.; Ojanen, C. H.; Hillamo, R. E.; Vilhunen, J. K.; Rantanen, L.; Havers, N.;
Von Bohlen, A.; Klockow, D. (1997) Characterization of the particulate phase in the exhaust from a diesel
car. Environ. Sci. Technol.'Sl: 1883-1889.
Kittelson, D. B. (1998) Engines and nanoparticles: a review. J. Aerosol Sci. 29: 575-588.
Kleeman, M. J.; Schauer, J. J.; Cass, G. R. (1999) Size and composition distribution of fine particulate matter emitted
from wood burning, meat charbroiling, and cigarettes. Environ. Sci. Technol. 33: 3516-3523.
Kleeman, M. J.; Schauer, J. J.; Cass, G. R. (2000) Size and composition distribution of fine particulate matter emitted
from motor vehicles. Environ. Sci. Technol. 34: 1132-1142.
Lawson, D. R.; Smith, R. E. (1998) The northern front range air quality study: a report to the Governor and General
Assembly. Fort Collins, CO: Colorado State University; December.
Linak, W. P.; Miller, C. A.; Wendt, J. O. L. (2000) Comparison of particle size distributions and elemental
partitioning from the combustion of pulverized coal and residual fuel oil. J. Air Waste Manage Assoc
50: 1532-1544. .
Lowenthal, D. H.; Zielinska, B.; Chow, J. C.; Watson, J. G.; Gautam, M.; Ferguson, D. H.; Neuroth, G. R.; Stevens,
K. D. (1994) Characterization of heavy-duty diesel vehicle emissions. Atmos. Environ. 28: 731-743!
Maricq, M. M.; Podsiadlik, D. H.; Chase, R. E. (1999) Gasoline vehicle particle size distributions: comparison of
steady state, FTP, and US06 measurements. Environ. Sci. Technol. 33: 2007-2015.
Marr, L. C.; Kirchstetter, T. W.; Harley, R. A.; Miguel, A. H.; Hering, S. V.; Hammond, S. K. (1999)
Characterization of polycyclic aromatic hydrocarbons in motor vehicle fuels and exhaust emissions Environ
Sci. Technol. 33: 3091-3099.
Mason, B. (1966) Principles of geochemistry. 3rd ed. New York, NY: John Wiley & Sons, Inc.
Matthias-Maser, S.; Jaenicke, R. (1994) Examination of atmospheric bioaerosol particles with radii > 0 2 urn
J. Aerosol Sci. 25: 1605-1613.
Mazurek, M. A.; Cofer, W. R., Ill; Levine, J. S. (1991) Carbonaceous aerosols from prescribed burning of a boreal
forest ecosystem. In: Levine, J. S., ed. Global biomass burning: atmospheric, climatic, and biospheric
implications. Cambridge, MA: MIT Press; pp. 258-263.
McDonald, J. D.; Zielinska, B.; Fujita, E. M.; Sagebiel, J. C.; Chow, J. C.; Watson, J. G. (2000) Fine particle and
gaseous emission rates from residential wood combustion. Environ. Sci. Technol. 34: 2080-2091
Miller, C. A.; Linak, W. P.; King, C.; Wendt, J. O. L. (1998) Fine particle emissions from heavy fuel oil combustion
in a firetube package boiler. Combust. Sci. Technol. 134: 477-502.
Norbeck, J. M.; Durbin, T. D.; Truex, T. J. (1998) Measurement of primary particulate matter emissions from
light-duty motor vehicles. Riverside, CA: University of California, College of Engineering, Center for
Environmental Research and Technology; prepared for Coordinating Research Council, Inc. and South Coast
Air Quality Management District.
March 2001
3B-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 Nriagu, J. O. (1989) A global assessment of natural sources of atmospheric trace metals. Nature (London)
2 338:47-49. .
3 Nriagu, J. O.; Pacyna, J. M. (1988) Quantitative assessment of worldwide contamination of air, water and soils by
4 trace metals. Nature (London) 333: 134-139.
5 Olmez, I.; Sheffield, A. E.; Gordon, G. E.; Houck, J. E.; Pritchett, L. C.; Cooper, J. A.; Dzubay, T. G.; Bennett, R. L.
6 (1988) Compositions of particles from selected sources in Philadelphia for receptor modeling applications.
7 JAPCA 38:1392-1402.
8 Oros, D. R.; Simoneit, B. R. T. (2000) Identification and emission rates of molecular tracers in coal smoke
9 paniculate matter. Fuel 79: 515-536.
10 Pierson, W. R.; Brachaczek, W. W. (1983) Paniculate matter associated with vehicles on the road. II. Aerosol Sci.
11 Technol'. 2:1-40.
12 Pinto, J. P.; Grant, L. D.; Hartlage, T. A. (1998) Report on U.S. EPA air monitoring of haze from S.E. Asia biomass
13 ' fires'. Research Triangle Park, NC: U.S. Environmental Protection Agency, National Center for Environmental
14 Assessment; report no. EPA/600/R-98/071.
15 Radke, L. F.; Hegg, D. A.; Hobbs, P. V.; Nance, J. D.; Lyons, J. H.; Laursen, K. K.; Weiss, R. E.; Riggan, P. J.;
16 Ward, D. E. (1991) Particulate and trace gas emissions from large biomass fires in North America. In: Levine,
17 j. s., ed. Global biomass burning: atmospheric, climatic, and biospheric implications. Cambridge, MA: MIT
18 Press; pp. 209-224.
19 Reilly, P. T. A.; Gieray, R. A.; Whitten, W. B.; Ramsey, J. M. (1998) Real-time characterization of the organic
20 'composition and size of individual diesel engine smoke particles. Environ. Sci. Technol. 32: 2672-2679.
21 Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. (1991) Sources of fine organic
22 aerosol. 1. Charbroilers and meat cooking operations. Environ. Sci. Technol. 25: 1112-1125.
23 Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. (1993a) Sources of fine organic
24 aerosol. 3. Road dust, tire debris, and organometallic brake lining dust: roads as sources and sinks. Environ.
25 Sci. Technol. 27:1892-1904.
26 Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. (1993b) Sources of fine organic
27 aerosol. 5. Natural gas home appliances. Environ. Sci. Technol. 27: 2736-2744.
28 Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. (1993c) Sources of fine organic
29 aerosol. 2. Noncatalyst and catalyst-equipped automobiles and heavy-duty diesel trucks. Environ. Sci.
30 Technol. 27: 636-651.
31 Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. (1993d) Sources of fine organic
32 aerosol. 4. Particulate abrasion products from leaf surfaces of urban plants. Environ. Sci. Technol.
33 27:2700-2711.
34' Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. (1994) Sources of fine organic
35 aerosol. 6. Cigarette smoke in the urban atmosphere. Environ. Sci. Technol. 28: 1375-1388.
36 Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. (1997a) Sources of fine organic
37 aerosol. 8. Boilers burning no. 2 distillate fuel oil. Environ. Sci. Technol. 31: 2731 -2737.
38 Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. (1997b) Sources of fine organic
39 aerosol. 7. Hot asphalt roofing tar pot fumes. Environ. Sci. Technol. 31: 2726-2730.
40 Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. (1998) Sources of fine organic
41 aerosol. 9. Pine, oak, and synthetic log combustion in residential fireplaces. Environ. Sci. Technol. 32:13-22.
42 Sagebiel, J. C.; Zielinska, B.; Walsh, P. A.; Chow, J. C.; Cadle, S. H.; Mulawa, P. A.; Knapp, K. T.; Zweidinger,
43 R. B.; Snow, R. (1997) PM-10 exhaust samples collected during IM-240 dynamometer tests of in-service
44 vehicles in Nevada. Environ. Sci. Technol. 31: 75-83.
45 Savoie, D. L.; Prospero, J. M. (1982) Particle size distribution of nitrate and sulfate in the marine atmosphere.
46 Geophys. Res. Lett. 9:1207-1210.
47 Sawyer, R. F.; Johnson, J. H. (1995) Diesel emissions and control technology. Diesel exhaust: a critical analysis of
48 emissions, exposure, and health effects. Cambridge, MA: Health Effects Institute.
49 Schauer, J. J.; Cass, G. R. (2000) Source apportionment of wintertime gas-phase and particle-phase air pollutants
50 using organic compounds as tracers. Environ. Sci. Technol. 34: 1821-1832.
51 Schauer, J. J.; Kleeman, M. J.; Cass, G. R.; Simoneit, B. R. T. (1999) Measurement of emissions from air pollution
52 sources. 2. C, through C30 organic compounds from medium duty diesel trucks. Environ. Sci. Technol.
53 33: 1578-1587.
54 Simoneit, B. R. T. (1985) Application of molecular marker analysis to vehicular exhaust for source reconciliations.
55 Int. J. Environ. Anal. Chem. 22:203-233.
March 2001
3B-28
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Simoneit, B. R. T.; Schauer, J. J.; Nolte, C. G.; Oros, D. R.; Elias, V. O.; Fraser, M. P.; Rogge, W. F.; Cass, G. R.
(1999) Levoglucosan, a tracer for cellulose in biomass burning and atmospheric particles. Atmos. Environ. 33-
173-182.
Spendlove, J. C. (1974) Industrial, agricultural, and municipal microbial aerosol problems. In: Current problems in
aerobiology: proceedings of the thirtieth general meeting of the Society for Industrial Microbiology; August
1973; Evanston, IL. Dev. Ind. Microbiol. 15: 20-27.
Standley, L. J.; Simoneit, B. R. T. (1987) Characterization of extractable plant wax, resin, and thermally matured
components in smoke particles from prescribed burns. Environ. Sci. Technol. 21: 163-169.
Susott, R. A.; Ward, D. E.; Babbitt, R. E.; Latham, D. J. (1991) The measurement of trace emissions and combustion
characteristics for a mass fire. In: Levine, J. S., ed. Global biomass burning: atmospheric, climatic, and
biospheric implications. Cambridge, MA: MIT Press; pp, 245-257.
Turekian, K. K. (1971) Elements, geochemical distribution of. In: Encyclopedia of science and technology, v. 4. New
York, NY: McGraw-Hill Book Company; pp. 627-630.
U.S. Environmental Protection Agency. (1996) Air quality criteria for particulate matter. Research Triangle Park,
NC: National Center for Environmental Assessment-RTF Office; report nos. EPA/600/P-95/001aF-cF. 3v.
U.S. Environmental Protection Agency. (1999) Health assessment document for diesel emissions, SAB review draft.
Washington, DC: Office of Research and Development; report no. EPA/600/8-90/057C.
Venkataraman, C.; Lyons, J. M.; Friedlander, S. K. (1994) Size distributions of polycyclic aromatic hydrocarbons
and elemental carbon. 1. Sampling, measurement methods, and source characterization. Environ. Sci. Technol
28: 555-562.
Vinogradov, A. P. (1959) The geochemistry of rare and dispersed chemical elements in soils. 2nd ed. New York,
NY: Consultants Bureau, Inc.
Warneck, P. (1988) Chemistry of the natural atmosphere. New York, NY: Academic Press, Inc.
Watson, J. G.; Chow, J. C. (1994) Clear sky visibility as a challenge for society. Annu. Rev. Energy Environ.
19: 241-266.
Watson, J. G.; Fujita, E. M.; Chow, J. C.; Zielinska, B.; Richards, L. W.; Neff, W.; Dietrich, D. (1998) Northern fron
range air quality study final report. Reno, NV: Desert Research Institute; prepared for Colorado State
University, Cooperative Institute for Research in the Atmosphere.
Weingartner, E.; Burtscher, H.; Baltensperger, U. (1997a) Hygroscopic properties of carbon and diesel soot particles.
Atmos. Environ. 31: 2311-2327.
Weingartner, E.; Keller, C.; Stahel, W. A.; Burtscher, H.; Baltensperger, U. (1997b) Aerosol emission in a road
tunnel. Atmos. Environ. 31: 451-462.
Westerholm, R.; Egeback, K.-E. (1994) Exhaust emissions from light- and heavy-duty vehicles: chemical
composition, impact of exhaust after treatment, and fuel parameters. In: Symposium of risk assessment of
urban air: emissions, risk identification, and risk quantitation; May-June 1992; Stockholm, Sweden. Environ.
Health Perspect. 102(suppl. 4) 13-23.
Westerholm, R. N.; Almen, J.; Li, H.; Rannug, J. U.; Egeback, K.-E.; Gragg, K. (1991) Chemical and biological
characterization of particulate-, semi volatile-, and gas-phase-associated compounds in diluted heavy-duty
diesel exhausts: a comparison of three different semivolatile-phase samplers. Environ. Sci. Technol
25: 332-338.
Wilson, T. R. S. (1975) Salinity and the major elements of sea water. In: Riley, J. P.; Skirrow, G., eds. Chemical
oceanography: v. 1. 2nd ed. London, United Kingdom: Academic Press Inc.; pp. 365-413.
Wu, J. (1979) Spray in the atmospheric surface layer: review and analysis of laboratory and oceanic results.
J. Geophys. Res. C: Oceans Atmos. 84: 1693-1704.
Yanowitz, J.; McCormick, R. L.; Grabowski, M. S. (2000) In-use emissions from heavy-duty diesel vehicles.
Environ. Sci. Technol. 34: 729-740.
March 2001
3B-29
DRAFT-DO NOT QUOTE OR CITE
-------
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
4. ENVIRONMENTAL EFFECTS OF
PARTICULATE MATTER
4.1 INTRODUCTION
Several later chapters (Chapters 5 through 8) of this document assess the latest available
information on determinants of human exposures to particulate matter (PM); dosimetry of
particle deposition, clearance, and retention in human respiratory tract; epidemiologic analyses of
health effects associated with human exposures to ambient PM; and toxicologic evaluations of
pathophysiologic effects of PM and underlying mechanisms of action. The human exposure and
health-related findings assessed in those chapters provide key elements of the scientific bases to
support upcoming decision making regarding potential retention or revision of the primary PM
National Ambient Air Quality Standards (NAAQS). This chapter, in contrast, assesses
information pertinent to decision making regarding secondary standards aimed at protecting
against welfare effects' of PM. More specifically, this chapter assesses environmental effects of
atmospheric PM, including discussion of PM effects on vegetation and ecosystems, PM effects
on visibility, PM effects on man-made materials, and relationships of ambient PM to global
climate change processes.
4.2 EFFECTS ON VEGETATION AND ECOSYSTEMS
The Particulate Matter National Ambient Air Quality Standards (PM NAAQS) set in 1971
were specified in terms of total suspended particulates (TSP), which included both fine and
coarse mode particles (the latter ranging up to 25 to 40 /u.m in size). The 1987 revision of the
TSP NAAQS to PM10 standards focused attention on those particles (< 10 //m mean aerometic
diameter) capable of being deposited in lower (thoracic) portions of the human respiratory tract.
The subsequent 1997 PM NAAQS revisions retained the PM10 standards and added fine particle
(PM2 5) standards (both specified in terms of mass concentrations of particles undifferentiated in
terms of their specific chemical composition). The effects of PM on vegetation and ecosystems
as a basis for a secondary standard were not considered as part of the 1997 PM NAAQS
March 2001
4-1
DRAFT-DO NOT QUOTE OR CITE
-------
1 revisions. Vegetation and ecosystem effects of ambient PM evaluated in this chapter are
2 dependent not so much just on PiVI size-related mass concentration, but rather on exposure of
3 plants to PM components differentiated by chemical composition as well.
4 Particulate matter deposition to vegetation is not well understood. Unlike gaseous dry
5 deposition, neither the solubility of the particles nor the physiological activity of the surface are
6 likely to be of first order importance in determining deposition velocity (Vd). Factors that
7 contribute to surface wetness and stickiness may be critical determinants of sticking efficiency.
8 Available tabulations of deposition velocities are highly variable and suspect. High-elevation
9 forests receive larger particulate deposition loadings than equivalent lower elevation sites,
10 because of higher wind speeds and enhanced rates of aerosol impaction; orographic effects on
11 rainfall intensity and composition; increased duration of occult deposition; and, in many areas,
12 the dominance of coniferous species with needle-shaped leaves (Lovett, 1984). Recent evidence
13 indicates that all three modes of deposition, (1) wet, (2) occult, and (3) dry, must be considered in
14 determining inputs to watersheds or ecosystems because each may dominate over specific
15 intervals of time or space.
16 Exposure to a given mass concentration of airborne PM may lead to widely differing
17 phytotoxic responses, depending on the particular mix of deposited particles. The most common
18 and useful subdivision of PM, derived from the typical bimodal distribution of atmospheric
19 • particles, is into fine and coarse particles (Wilson and Suh, 1997). The smallest particle at or
20 near 1.0 to 2.5 ,um generally is taken as the division between fine and coarse, although this is not
21 an absolute and is subject to some shift (e.g., with changing ambient humidity). However, the
22 typical the rule of thumb, as previously used in the 1996 PM Air Quality Criteria Document or
23 "PM AQCD" (U.S. Environmental Protection Agency, 1996a), is that fine PM nominally falls in
24 the range of 0 to 2.5 yum and coarse-mode PM, 2.5 to 10.0 /u.m.
25 In general, fine-mode PM is secondary in nature, having condensed from the vapor phase or
26 been formed by chemical reaction from gaseous precursors in the atmosphere. These particles
27 exist in a nucleation mode (having a mass median aerodynamic diameter or MMAD of about
28 0.06 /um) and may grow by coagulation of existing particles or by condensation of additional
29 gases onto existing particles into an accumulation mode (about 0.5 fj.ro). Sulfur and nitrogen
30 oxides (SOX and NOJ, as well as volatile organic gases, are common precursors for fine PM,
31 often neutralized with ammonium cations as particulate salts. Condensation of volatilized metals
March 2001
4-2
DRAFT-DO NOT QUOTE OR CITE
-------
1 and products of incomplete combustion also are common precursors. Reactions of many of these
2 materials with an oxidizing atmosphere lead to high secondary PM concentrations during
3 summer months in many parts of the United States.
4 In general, coarse-mode particles are primary in nature, having been produced and emitted
5 from a point or area source as a fully formed particle. They range in size from ca. 2.5 to 100 //m.
6 This material is created by abrasion and subsequent suspension by wind or mechanical means.
7 Suspended geologic material contains the chemical and, potentially, the biological signature of
8 the soil from which it derives (dominated by iron, silica, aluminum, and calcium). Additional
9 anthropogenically derived coarse-mode PM derives from fly ash, automobile tires and brake
10 linings, and industrial effluent associated with crushing and grinding operations. Coarse-mode
11 particles also include biogenically derived organic materials (e.g., fragments of plants and
12 insects, pollen, fungal spores, bacteria and viruses included in marine aerosols).
13 Atmospheric deposition of particles to ecosystems takes place via both wet and dry
14 processes through three major routes: (1) precipitation scavenging in which particles are
15 deposited in rain and snow; (2) fog, cloud-water, and mist interception; and (3) dry deposition,
16 a much slower, yet more continuous removal to surfaces (Hicks, 1986).
17 Precipitation scavenging includes rainout involving within-cloud nucleation phenomena
18 and washout involving below-cloud scavenging by impaction. Total inputs from wet deposition
19 to vegetative canopies can be significant (Table 4-1), although not all wet deposition involves
20 particle scavenging because gaseous pollutants also dissolve during precipitation.
21 Wet deposition is not affected by surface properties as much as is dry or occult deposition.
22 However, forested hillsides may receive much (four- to sixfold) greater precipitation than short
23 vegetation in nearby valleys because of a variety of orographic effects (Unsworth and Wilshaw,
24 1989). Additionally, closer aerodynamic coupling to the atmosphere of the tall forest canopy
25 than of the shorter canopies in the valleys leads to more rapid foliar drying, reduced residence
26 time of solubilized particulate materials available for foliar uptake, and, consequently, more rapid
27 and more extreme concentration of such materials on the cuticular surface. The results of direct
28 physical effects on leaves are not known.
29 Most of wet deposited particulate material passes through the plant canopy to the soil by
30 throughfall and stemflow, causing soil-mediated ecosystem-level responses. Rainfall also
31 removes much of the dry-deposited PM resident on foliar surfaces, reducing direct foliar effects
March 2001
4-3
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-1. RELATIVE IMPORTANCE OF WET, DRY, PARTICIPATE,
AND TOTAL DEPOSITION TO THREE FOREST SITES8
Deposition
Total Nitrogen1"
Site
Duke Forest
Gary Forest
Austin Forest
Wet
(%)
75
71
71
Dry
(%)
25
20
29
Particle
(%)
0.11
0.94
0.58
Total
(kg ha'1)
9.87
5.80
6.57
Wet
(%)
64
76
83
Total Sulfur*
Dry
(%)
33
20
13
Particle
(%)
2.7
4.2
4.3
Total
(kg ha1)
17.20
7.60
7.79
"Data from Allen et al. (1994). Sampling was by triple filter pack, so that fine-mode particles could be sampled
preferentially. An average particle deposition velocity of 0.9 cm s'1 was derived, as in Hicks et al. (1987).
bWet nitrogen consists of NO3' and NH4+, dry nitrogen consists of vapor phase HNO3 and NO2, and particulate
nitrogen consists of NO3".
"Wet sulfur consists of SO4", dry sulfur consists of vapor phase SO2, and particulate sulfur consists of pSO4~.
1 (Lovett and Lindberg, 1984). This washing effect, combined with differential foliar uptake and
2 foliar leaching (both of which depend on the physiological status of the vegetation), alters the
3 composition of rainwater that reaches the soil. Dry deposition onto foliage and subsequent wet
4 removal by runoff enhances soil-mediated effects of particulate deposition, both by enhancing
5 total dry deposition relative to unvegetated surfaces nearby and by accelerating passage of
6 deposited particles to the soil. The most significant effects of wet deposition occur through soil-
7 mediated processes involving biogeochemcial cycling of major and minor nutrients and trace
8 elements.
9 Dry deposition is more effective for coarse particles of natural origin and elements such as
10 iron and manganese, whereas wet deposition generally is more effective for fine PM of
11 atmospheric origin and elements such as cadmium, chromium, lead, nickel, and vanadium
12 (Smith, 1990a). The actual importance of wet versus dry deposition, however, is highly variable,
13 depending on ecosystem type, location, and elevation. For the Walker Branch Watershed, a
14 deciduous forest in rural eastern Tennessee, dry deposition constituted a major fraction of total
15 annual atmospheric input of cadmium and zinc (=20%), lead (=55%), and manganese (=90%),
16 but wet deposition rates for single precipitation events exceeded dry deposition rates by one to
17 four orders of magnitude (Lindberg and Harriss, 1981). Miller et al. (1993) emphasized that
18 immersion of high-elevation forests in cloudwater for 10% or more of the year can enhance
19 significantly overall efficiency of transfer of atmospheric particles and gases to a forest canopy.
March 2001 4-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 Dry deposition of particles occurs to all vegetational surfaces exposed to the atmosphere
2 (U.S. Environmental Protection Agency, 1982). The range of particle sizes, the diversity of
3 canopy surfaces, and the variety of chemical constituents in airborne PM have slowed progress in
4 both prediction and measurement of dry particulate deposition. Wet deposition generally is
5 confounded by fewer factors and has been easier to quantify (Chapter 2).
6 Emphasis in this and the next section is placed on discussion of PM effects on individual
7 plants in natural habitats and terrestrial ecosystems. Except for the deposition of nitrogen and
8 sulfur-containing compounds and their effects exerted via acidic precipitation, information
9 concerning the effects of deposition of other specific substances as PM on crops is not readily
10 available. The U.S. National Acid Precipitation Assessment Program (NAPAP) Biennial Report
11 to Congress: An Integrated Assessment presents an extensive overall discussion of the effects of
12 acidic deposition (National Science and Technology Council, 1998). The effects of gaseous
13 sulfur oxides and nitrogen oxides on crops are discussed in detail in EPA criteria documents for
14 those substances (U.S. Environmental Protection Agency, 1982, 1993). A detailed discussion of
15 the ecological effects of acidic precipitation and nitrate deposition on aquatic ecosystems also
16 can be found in the EPA Nitrogen Oxides Air Quality Criteria Document (U.S. Environmental
17 Protection Agency, 1993). Neither nitrate or sulfate deposition on crops is discussed in this
18 chapter, as they are added frequently in fertilizers. Also, the effects of lead on crops, vegetation,
19 and ecosystems are discussed in the EPA document, Air Quality Criteria for Lead (U.S.
20 Environmental Protection Agency, 1986).
21 The effects of deposited PM may be direct or indirect. Indirect effects are chiefly
22 nutritional responses mediated through the soil and result from the effects of PM components on
23 soil processes. In the following sections, the direct effects on individual plants are discussed
24 first, followed by effects on plant species and their interactions in ecosystems.
25 4.2.1 Direct Effects of Particulate Matter on Individual Plant Species
26 Particulate matter in the atmosphere may affect vegetation directly following deposition on
27 foliar surfaces, indirectly by changing the soil chemistry, or through changes in the amount of
28 radiation reaching the Earth's surface through PM-induced climate change processes. Indirect
29 impacts, however, are usually the most significant because they can alter nutrient cycling and
March 2001
4-5
DRAFT-DO NOT QUOTE OR CITE
-------
1 inhibit plant nutrient uptake. The possible direct responses to PM deposition are considered in
2 this section, and the indirect responses in the sections on ecosystems.
3 Particles transferred from the atmosphere to foliar surfaces may reside on the leaf, twig or
4 bark surface for extended periods; be taken up through the leaf surface; or be removed from the
5 plant via resuspension to the atmosphere, washing by rainfall, or litter-fall with subsequent
6 transfer to the soil. Any PM deposited on above-ground plant parts may exert physical or
7 chemical impacts. The effects of "inert" PM are mainly physical, whereas those of toxic particles
8 are both chemical and physical. The chemical effects of dust deposited on plant surfaces or soil
9 are more likely to be associated with their chemistry than simply with the mass of deposited
10 particles and may be more important than any physical effects (Farmer, 1993).
11 Studies of the direct effects of chemical additions to foliage in particulate deposition have
12 found little or no effects of PM on foliar processes unless exposure levels were significantly
13 higher than typically would be experienced in the ambient environment. Interpretation of the
14 effects of atmospheric chemical deposition at the level of individual plants and ecosystems is
15 difficult because of the complex interactions that exist among biological, physicochemical, and
16 climatic factors. The majority of the easily identifiable direct and indirect effects, other than
17 climate, occur in severely polluted areas around heavily industrialized point sources, such as
18 limestone quarries, cement kilns, and smelting facilities for iron, lead, or various other metals.
19" The diverse chemical nature and size characteristics of ambient airborne particles and the lack of
20 any clear distinction between effects attributed to phytotoxic particles and to other forms of air
21 pollutants confound the direct effects of PM on foliar surfaces. Most documented toxic effects of
22 particles on vegetation reflect their acidity, trace metal content, nutrient content, surfactant
23 properties, or salinity. These materials typically elicit similar biological effects, whether
24 deposited as coarse or fine particles, in wet, dry, or occult form, and, frequently, whether
25 deposited to foliage or to the soil. Studies of direct effects of particles on vegetation have not yet
26 advanced to the stage of reproducible exposure experiments. Experimental difficulties in
27 application of ambient particles to vegetation have been discussed by Olszyk et al. (1989).
28 4.2.1.1 Effects of Coarse Particles
29 Coarse-mode particles, ranging in size from 2.5 to 100 fjm, are chemically diverse, are
30 dominated by local sources, and are typically deposited near the source because of their
March 2001
4-6
DRAFT-DO NOT QUOTE OR CITE
-------
1 sedimentation velocities. Airborne coarse particles are derived from road, cement kiln, and
2 foundry dust; fly ash; tire particles and brake linings; soot and cooking oil droplets; biogenic
3 materials (e.g., plant pollen, fragments of plants, fungal spores, bacteria and viruses) and sea salt.
4 Most coarse particles in rural and some urban areas are composed of silicon, aluminum, calcium,
5 and iron, suggesting that their main source is fugitive dust from disturbed land, roadways,
6 agricultural tillage, or construction. Rapid sedimentation of these particles tends to restrict their
7 direct effects on vegetatioolargely to roadsides and forest edges. L
8 Physical Effects—Radiation. Dust can have both a physical and chemical impact.
9 Deposition of inert PM on above-ground plant organs may result in an increase in radiation
10 received, in leaf temperature and blockage of stomata. Increased leaf temperature, heat stress,
11 reduced net photosynthesis, and leaf chlorosis, necrosis, and abscission were reported by
12 Guderian (1986). Road dust decreased the leaf temperature on Rhododendron catavshiense by
13 ca. 4 °C (Eller, 1977), whereas foundry dust caused an 8.7 °C increase in leaf temperature of
14 black poplar (Populus nigrd) (Guderian, 1986) under the conditions of the experiment.
15 Broad-leaved plants exhibited greater temperature increases because of particle loading than did
16 the needle-like leaves of conifers. Deciduous (broad) leaves exhibited larger temperature
17 increases because of particle loading than did conifer (needle) leaves, a function of poorer
18 coupling to the atmosphere. Inert road dust caused a three- to fourfold increase in the absorption
19 coefficient of leaves ofHedera helix (Eller, 1977; Guderian, 1986) for near infrared radiation
20 (NIR; 750 to 1350 nm). Little change occurred in absorption for photosynthetically active
21 radiation (PAR; 400 to 700 nm). The increase in NIR absorption was equally at the expense of
22 reflectance and transmission in these wavelengths. The net energy budget increased by ca. 30%
23 in the dust-affected leaves. Deposition of coarse particles increased leaf temperature and
24 contributed to heat stress, reduced net photosynthesis, and caused leaf chlorosis, necrosis, and
25 abscission (Dassler et al., 1972; Parish, 1910; Guderian, 1986; Spinka, 1971).
26 Starch storage in dust-affected leaves increased with dust loading under high (possibly
27 excessive) radiation, but decreased following dust deposition when radiation was limiting. These
28 modifications of the radiation environment had a large impact on single-leaf utilization of light.
29 The boundary layer properties, determined by leaf morphology and environmental conditions,
March 2001
4-7
DRAFT-DO NOT QUOTE OR CITE
-------
1 strongly influenced the direct effects of particle deposition on radiation heating (Eller, 1977;
2 Guderian, 1986) and on gas exchange as well.
3 Brandt and Rhoades (1973) attributed the reduction in growth of trees because of crust
4 formation from limestone dust on the leaves. Crust formation reduced photosynthesis and
5 formation of carbohydrate needed for normal growth, induced premature leaf-fall, destruction of
6 leaf tissues, inhibited growth of new tissue,* and reduced storage. Dust may affect
7 photosynthesis, respiration, and transpiration, and it may allow penetration of phytotoxic gaseous
8 pollutants, thereby causing visible injury symptoms and decreased productivity. Permeability of
9 leaves to ammonia increased with increasing dust concentrations and decreasing particle size
10 (Farmer, 1993).
11 Dust also has been reported to physically block stomata (Krajickova and Mejstfik, 1984).
12 Stomatal clogging by particulate matter from automobiles, stone quarries, and cement plants was
13 also studied by Abdullah and Iqbal (1991). The percentage of clogging was low in young leaves
14 when compared with old and mature leaves and the amount of clogging varied with species and
15 locality. The maximum clogging of stomata observed was about 25%. The authors cited no
16 evidence that stomatal clogging inhibited plant functioning. The heaviest deposit of dust is
17 usually on the upper surface of broad-leaved plants, however, whereas the majority of the
18 stomata are on the lower surface where stomatal clogging would be less likely.
19 Chemical Effects. The chemical composition of PM is usually the key phytotoxic factor
20 leading to plant injury. Cement-kiln dust on hydration liberates calcium hydroxide, which can
21 penetrate the epidermis and enter the mesophyll, and, in some cases, the leaf surface alkalinity
22 may reach to pH 12. Lipid hydrolysis coagulation of the protein compounds and ultimately
23 plasmolysis of the leaf tissue result in reduction in growth and quality of plants (Guderian, 1986).
24 In experimental studies, application of cement kiln dust of known composition for 2 to 3 days
25 yielded dose-response curves between net photosynthetic inhibition or foliar injury and dust
26 application rate (Darley, 1966). Lerman and Darley (1975) determined that leaves must be
27 misted regularly to produce large effects. ^Alkalinity was probably the essential phytotoxic
28 property of the applied dusts.
29 Particulate matter in the form of sea salt enters the atmosphere from oceans following
30 mixing of air into the water and subsequent bursting of bubbles at the surface. This process can
March 2001
4-8
DRAFT-DO NOT QUOTE OR CITE
-------
1 be a significant source of sulfate, sodium chloride, and trace elements in the atmosphere over
2 coastal vegetation, resulting in the formation of the maritime forest, a specialized ecosystem.
3 Sea-salt particles can serve as nuclei for the adsorption and subsequent reaction of other gaseous
4 and particulate pollutants. Both nitrate and sulfate from the atmosphere have been found to be
5 associated with coarse and fine sea-salt particles (Wu and Okada, 1994). Direct effects on
6 vegetation reflect these inputs, as well as classical salt injury caused by the sodium and chloride
7 that constitute the bulk of these particles. Salt pruning is a common phenomenon near the ocean
8 (i.e., salt spray kills the buds on the windward side of trees and shrubs).
9 4.2.1.2 Effects of Fine Particles
10 Fine PM is generally secondary in nature, having condensed from the vapor phase or been
11 formed by chemical reaction from gaseous precursors in the atmosphere and is generally smaller
12 than 1 to 2.5 //m. Nitrogen and sulfur oxides, volatile organic gases, condensation of volatilized
13 metals, and products of incomplete combustion are common precursors for fine PM. Reactions
14 of many of these materials with an oxidizing atmosphere contribute to high secondary PM
15 concentrations during summer months in many U.S. areas. The conclusion reached in the 1982
16 PM AQCD (U.S. Environmental Protection Agency, 1982) that sufficient data were not available
17 for adequate quantification of dose-response functions for direct effects of fine aerosols on
18 vegetation continues to be true today. Only a few studies have been completed on the direct
19 effects of acid aerosols (U. S. Environmental Protection Agency, 1982). The major effects are
20 indirect and occur through the soil (Section 4.3).
21 Nitrogen. Nitrate is observed in both fine and coarse particles. Nitrates from atmospheric
22 deposition represent a substantial fraction of total nitrogen inputs to southeastern forests (e.g.,
23 Lovett and Lindberg, 1986). However, much of this is contributed by gaseous nitric acid vapor,
24 and a considerable amount of the particulate nitrate is taken up indirectly, through the soil.
25 Garner et al. (1989) estimated deposition of nitrogen to forested landscapes in eastern North
26 America at 10 to 55 kg/ha/year for nitrate and 2 to 10 kg/ha/year for ammonium. About half of
27 these values were ascribed to dry deposition.
28 Atmospheric additions of particulate nitrogen in excess of vegetation needs are lost from
29 the system, mostly as leachate from the soil as nitrate. Managed agricultural ecosystems may be
March 2001
4-9
DRAFT-DO NOT QUOTE OR CITE
-------
1 able to utilize deposited particulate nitrogen more efficiently than native ecosystems, although
2 many cultivated systems also lose considerable nitrogen as nitrate in runoff, deep drainage, or tail
3 water. It has proven difficult to quantify direct foliar fertilization by uptake of nitrogen from
4 ambient particles.
5 There is no doubt that foliar uptake of nitrate can occur, as clearly shown by the efficacy of
6 foliar fertilization in horticultural systems. Potassium nitrate was taken up by leave's of
7 deciduous fruit trees (Weinbaum and Neumann, 1977) and resulted in increased foliar nitrogen
8 concentrations. Not all forms of nitrogen are absorbed equally, nor are all equally benign.
9 Following foliar application of 2600 ppm of nitrogen as Ca(NO3)2, (NH4)2SO4, or (NH2)2CO to
10 apple canopies (Rodney, 1952; Norton and Childers, 1954), leaf nitrogen levels were observed to
11 increase to similar levels, but calcium nitrate and ammonium sulfate caused visible foliar
12 damage, whereas urea did not. Urea is generally the recommended horticultural foliar fertilizer.
13 The mechanism of uptake of foliarly deposited nitrate is not well established. Nitrate
14 reductase is generally a root-localized enzyme. It is generally not present in leaves, but is
15 inducible there. This typically occurs when the soil is heavily enriched in NO3". As the root
16 complement of nitrate reductase becomes overloaded, unreduced nitrate reaches the leaves
17 through the transpiration stream. Nitrate metabolism has been demonstrated in leaf tissue
18 (Weinbaum and Neumann, 1977) following foliar fertilization. Residual nitrate reductase
19 activity in leaves may be adequate to assimilate typical rates of particulate nitrate deposition.
20 Uptake of nitrate may be facilitated by codeposited sulfur (Karmoker et al., 1991; Turner and
21 Lambert, 1980).
22 Nitrate reductase is feedback-inhibited by its reaction product, NH4+. The common
23 atmospheric aerosol, NH4NO3, therefore may be metabolized in two distinct biochemical steps,
24 first the ammonium (probably leaving nitric acid) and then the nitrate. Volatilization losses of
25 nitric acid during this process, if they occur, have not been characterized.
26 Direct foliar effects of particulate nitrogen have not been documented. Application of a
27 variety of fine nitrogenous aerosol particles (0.25 ywm) ranging from 109 to 244 /ug/m3 nitrogen,
28 with or without 637 //g/m3 sulfur, caused no consistent short-term (2- to 5-h) effect on gas
29 exchange in oak, maize, or soybean leaves (Martin et al., 1992).
30 Although no evidence exists for direct transfer of nutrient particulate aerosols into foliage,
31 a few studies give insights into the potential for ammonium and nitrate transfer into leaves.
March 2001
4-10
DRAFT-DO NOT QUOTE OR CITE
_
-------
1 Fluxes of both NO3" and NH/, measured in wet deposition and in throughfall plus stemflow in
2 forests, commonly indicate higher fluxes of nitrogen above the canopy (Parker, 1983; Lindberg
3 et al., 1987; Sievering et al., 1996), indicating net foliar uptake. Lovett and Lindberg (1993)
4 reported a linear relationship between inorganic nitrogen fluxes in deposition and throughfall,
5 suggesting that uptake may be considered passive to some extent.
6 Garten and Hanson (1990) studied the movement of I5N-labeled nitrate and ammonium
7 across the cuticles of red maple (Acer rubrum) and white oak (Quercus alba) leaves when
8 applied as an artificial rain mixture. Brumme et al. (1992), Bowden et al. (1989), and Vose and
9 Swank (1990) have published similar data for conifers. These studies show the potential for
10 nitrate and ammonium to move into leaves, where it may contribute to normal physiological
11 processes (e.g., amino acid production; Wellburn, 1990). Garten (1988) showed that internally
12 translocated 35S was not leached readily from tree leaves of yellow poplar (Liriodendron
13 tulipifera) and red maple (Acer rubrum), suggesting that SO42" would not be as mobile as the
14 nitrogen-containing ions discussed by Garten and Hanson (1990). Further, when the foliar
15 extraction method is used it is not possible to distinguish sources of chemical deposited as gases
16 or particles (e.g., nitric acid [HNO3], nitrogen dioxides [NO2], nitrate [NO3"], or sources of
17 ammonium deposited as ammonia [NH3] or ammonium ion [NH4+]) (Garten and Hanson, 1990).
18 Particle deposition contributes only a portion of the total atmospheric nitrogen deposition
19 reaching vegetation but, when combined with gaseous and precipitation-derived sources, total
20 nitrogen deposition to ecosystems has been identified as a possible causal factor leading to
21 changes in natural ecosystems (See Section 4.3).
22 Sulfur. Anthropogenic sulfur emissions are >90% as SO2. Most of the remaining emission
23 of sulfur is directly as sulfate (U.S. Environmental Protection Agency, 1996a). Sulfur dioxide is
24 hydrophilic and is rapidly hydrated and oxidized to sulfite and bisulfite and then to sulfate, which
25 is approximately 30-fold less phytotoxic. The ratio of sulfate/SO2 increases with aging of the air
26 mass and, therefore, with distance from the source. Sulfate is sufficiently hygroscopic that, in
27 humid air, it may exist significantly in the coarse particulate fraction. As dilution of both SO2
28 and particulate SO42" occurs with distance from the source, it is unusual for damaging levels of
29 particulate sulfate to be deposited. Gas to particle conversion in this case is of benefit to
30 vegetation.
March 2001
4-11
DRAFT-DO NOT QUOTE OR CITE
-------
1 Sulfur is an essential plant nutrient. Low dosages of sulfur serve as a fertilizer, particularly
2 for plants growing in sulfur-deficient soil (Hogan et al., 1998). Current levels of sulfate
3 deposition reportedly exceed the capacity of most vegetative canopies to immobilize the sulfur
4 (Johnson, 1984). Nitrogen uptake in forests may be regulated loosely by sulfur availability, but
5 sulfate additions in excess of needs do not typically lead to injury (Turner and Lambert, 1980).
6 There are few field demonstrations of foliar sulfate uptake (Krupa and Legge, 1986, 1998).
7 Sulfate in throughfall is often enriched above levels in precipitation. The relative importance of
8 foliar leachate and prior dry-deposited sulfate particles remains difficult to quantify (Cape et al.,
9 1992). Leaching rates are not constant and may respond to levels of other pollutants, including
10 acids. Uptake and foliar retention of gaseous and particulate sulfur are confounded by variable
11 rates of translocation and accessibility of deposited materials to removal and quantification by
12 leaf washing. Following soil enrichment with 35SO42" in a Scots pine forest, the apparent
13 contribution of leachate to throughfall was only a few percent, following an initial burst of over
14 90% because of extreme disequilibrium in labeling of tissue sulfate pools (Cape et al., 1992).
15 Olszyk et al. (1989) provide information on the impacts of multiple pollutant exposures
16 including particles (NO3', 142 ^g/m3; NH/, 101 //g/m3; SO42', 107 //g/m3). They found that only
17 gaseous pollutants produced direct (harmful) effects on vegetation for the concentrations
18 documented, but the authors hypothesized that long-term accumulation of the nitrogen and sulfur
19 compounds contributed from particle deposition might have effects on plant nutrition over long
20 periods of time. Martin et al. (1992) exposed oak (Quercus macrocarpd), soybean (Glycine
21 max), and maize (Zea mays) plants to acute exposures (2 to 5 h) of aerosols (0.25 //m) containing
22 only nitrate (109 /ug/m3), ammonium and nitrate (244 and 199 //g/m3, respectively), or
23 ammonium and sulfate (179 and 637 //g/m3, respectively). They found that these exposures,
24 which exceeded the range of naturally occurring aerosol concentrations, had little effect on foliar
25 photosynthesis and conductance. Martin et al. (1992) concluded that future investigations should
26 focus on the effects of particles on physiological characteristics of plants following chronic
27 exposures.
28 Acidic Deposition. The effects of acidic deposition have been accorded wide attention in
29 the media and elsewhere (Altshuller and Linthurst, 1984; Hogan et al., 1998). Probably the most
30 extensive assessment of acidic deposition processes and effects is the NAPAP Biennial Report to
March 2001
4-12
DRAFT-DO NOT QUOTE OR CITE
-------
1 Congress: An Integrated Assessment (National Science and Technology Council, 1998).
2 Concern regarding the effects of acidic deposition on crops and forest trees has resulted in
3 extensive monitoring and research. Exposures to acidic rain or clouds can be divided into
4 "acute" exposures to higher ionic concentrations (several /^mol/L), and "chronic" long-term
5 repeated exposures to lower concentrations (Cape, 1993). Pollutant concentrations in rainfall
6 have been shown to have little capacity for producing direct effects on vegetation (Altshuller and
7 Linthurst, 1984; Hogan et al., 1998); however, fog and clouds, which may contain solute
8 concentrations up to 10 times those found in rain, have the potential for direct effects. More than
9 80% of the ionic composition of most cloud water is made up of four major pollutant ions: H+,
10 NH4+, NO3", and SO42". Ratios of hydrogen to ammonium and sulfate to nitrate vary from site to
11 site with all four ions usually present in approximately equal concentrations. Available data from
12 plant effect studies suggest that hydrogen and sulfate ions are more likely to cause injury than
13 ions containing nitrogen (Cape, 1993).
14 The possible direct effects of acidic precipitation on forest trees have been evaluated by
15 experiments on seedlings and young trees. The size of mature trees makes experimental
16 exposure difficult, therefore necessitating extrapolations from experiments on seedlings and
17 saplings; however, such extrapolations must be used with caution (Cape, 1993). Both conifers
18 and deciduous species have shown significant effects on leaf surface structures after exposure to
19 simulated acid rain or acid mist at pH 3.5. Some species have shown subtle effects at pH 4 and
20 above. Visible lesions have been observed on many species at pH 3 and on sensitive species at
21 pH 3.5 (Cape, 1993). The relative sensitivities of forest vegetation to acidic precipitation based
22 on macroscopic injury have been ranked as follows: herbaceous dicots > woody dicots >
23 monocots > conifers (Percy 1991).
24 Huttunen (1994) described the direct effects of acid rain or acidic mist on epicuticular
25 waxes whose ultrastructure is affected by plant genotype and phenotype. The effects of air
26 pollutants on epicuticular waxes of conifers have received greater study than the waxes of other
27 species. Leafage and shorter life span of broad-leaved trees make them less indicative of the
28 effects of acid precipitation. Many experimental studies indicate that epicuticular waxes that
29 function to prevent water loss from plant leaves can be destroyed by acid rain in a few weeks
30 (Huttunen, 1994). This function is crucial in conifers because of their longevity and evergreen
31 foliage. Microscopic observations of epicuticular wax structures have, for a long time, suggested
March 2001
4-13
DRAFT-DO NOT QUOTE OR CITE
-------
1 links between acidic deposition and aging. In Norway spruce (Picea abies), acid rain causes not
2 only the aging of needles (which in northern conditions normally last from 11 to 14 years) to be
3 shortened but also accelerates the erosion rate of the waxes as the needles age.
4 The effects of acidic precipitation and fog on red spruce (Picea rubens) have been studied
5 extensively (Schier and Jensen, 1992). Visible foliar injury of the needles in the form of a
6 reddish-brown discoloration has been observed on red spruce seedlings experimentally exposed
7 to acidic mist, but this visible symptom has not been observed in the field. .Ultrastructural
8 changes in the epicuticular wax were observed both experimentally and on spruce growing at
9 high elevations. Laboratory studies indicate that visible injury usually does not occur unless the
10 pH is 3 or less (Schier and Jensen, 1992). Cape (1993) reported that, when compared with other
11 species, red spruce seedlings appeared to be more sensitive to acid mist. Huttunen (1994)
12 concluded that his studies of conifers and review of the literature suggest that acidic precipitation
13 causes direct injury to tree foliage and, also, indirect effects through the soil. The indirect effects
14 of acidic precipitation are discussed in Section 4.3.
15 Based on his review of the many studies involving field and controlled laboratory
16 experiments on crops in the literature, Cape (1993) drew a number of conclusions concerning the
17 direct effects of acidic precipitation on crops:
18 • foliar injury and growth reduction occurs below pH 3;
19 • allocation of photosynthate is altered, with increased shoot to root ratios;
20 • expanded and recently expanded leaves are most susceptible, and injury occurs first to
21 epidermal cells;
22 • leaf surface characteristics such as wettability, buffering capacity, and transport of
23 material across the leaf surface contribute to susceptibility and differ among species;
24 • data obtained from experiments in greenhouses or controlled environmental chambers
25 cannot be used to predict effects on plants grown in the field;
26 • quantitative data from experimental exposures cannot be extrapolated to field exposures
27 because of differences and fluctuations in concentrations, durations, and frequency of
28 exposure;
29 • there are large differences in response within species;
30 • timing of exposure in relation to phenology is of utmost importance;
31 • plants may be able to recover from or adapt to injurious exposures; and
March 2001
4-14
DRAFT-DO NOT QUOTE OR CITE
-------
1 • sequential exposure to acidic precipitation and gaseous pollutants is unlikely to be more
2 injurious than exposure to individual pollutants.
3 Studies by Chevone et al. (1986), Krupa and Legge (1986), and Blaschke (1990) differ with
4 the last conclusion of Cape listed above. Their studies indicate that interactions between acidic
5 deposition and gaseous pollutants do occur. Acidity affects plant responses to both O3 and SO2.
6 Chevone et al. (1986) observed increased visible injury on soybean and pinto bean when acid
7 aerosol exposure preceded O3 exposure, whereas linear decreases in dry root weight of yellow
8 poplar occurred as acidity increased with simultaneous exposures to O3 and simulated acid rain.
9 Kfupa and Legge (1986) also noted increased visible injury to pinto bean when aerosol exposure
10 preceded O3 exposure. In none of the studies cited above did acid rain per se'produce significant
11 growth changes. Blaschke (1990) observed a decrease in ectomycorrhizal frequency and short
12 root distribution caused by acid rain exposure in combination with either SO2 or O3.
13 Trace Elements. All but 10 of the 90 elements that comprise the inorganic fraction of the
14 soil occur at concentrations of less than 0.1% (1000 |ig/g) and are termed "trace" elements.
15 Trace elements with a density greater than 6 g.cm"3, referred to as "heavy metals", are of
16 particular interest because of their potential toxicity for plant and animals. Although some trace
17 metals are essential for vegetative growth or animal health, they are all toxic in large quantities.
18 Combustion processes produce metal chlorides that tend to be volatile and metal oxides that tend
19 to be nonvolatile in the vapor phase (McGowan et al., 1993). Most trace elements exist in the
20 atmosphere in particulate form as metal oxides (Ormrod, 1984). Aerosols containing trace
21 elements derive predominantly from industrial activities (Ormrod, 1984). Generally, only
22 cadmium, chromium, nickel, and mercury are released from stacks in the vapor phase (McGowan
23 et al., 1993). Concentrations of heavy metals in incinerator fly ash increase with decreasing
24 particle size.
25 Vegetational surfaces, especially the foliage, present a major reaction and filtration surface
26 to the atmosphere and act to accumulate particles deposited via wet and dry processes described
27 in Chapter 2 (Tong, 1991; Youngs et al., 1993). Particles deposited on foliar surfaces may be
28 taken up through the leaf surface. The greatest particle loading is usually on the adaxial (upper)
29 leaf surface where particles accumulate in the mid-vein, center portion of the leaves. The
March 2001
4-15
DRAFT-DO NOT QUOTE OR CITE
-------
1 mycelium of fungi becomes particularly abundant on leaf surfaces as the growing season
2 progresses and is in intimate association with deposited particles (Smith, 1990b).
3 Investigations of trace elements present along roadsides and in industrial and urban
4 environments indicate that impressive burdens of particulate heavy metals can accumulate on
5 vegetative surfaces. Foliar uptake of available metals could result in metabolic impact in above-
6 ground tissues. Only a few metals, however, have been documented to cause direct phytotoxicity
7 in field conditions. Copper, zinc, and nickel toxicities have been observed most frequently. Low
8 solubility, however, limits foliar uptake and direct heavy metal toxicity. A trace metal must be
9 brought into solution before it can enter into leaves or bark of vascular plants, m those instances
10 when trace metals are absorbed, they are frequently bound in leaf tissue and are lost when the leaf
11 drops off (Hughes, 1981). Trace metals in mixtures may interact to cause a different plant
12 response when compared with a single element; however, there has been little research on this
13 aspect (Ormrod, 1984). In experiments using chambers, Marchwinska and Kucharski (1987)
14 studied the effects of SO2 alone and hi combination with PM components (Pb, Cd, Zn, Fe, Cu,
15 and Mn) obtained from a zinc smelter bag filter. The combined effects of SO2 and PM further
16 increased the reduction in yield of beans caused by SO2, whereas the combination, though
17 severely injuring the foliage, produced little effect on carrots and parsley roots, except after
18 long-term exposures (when there was a decrease in root weight).
19 Trace metal toxicity of lichens has been demonstrated in relatively few cases. Nash (1975)
20 documented zinc toxicity in the vicinity of a zinc smelter near Palmerton, PA. Lichen species
21 richness and abundance were reduced by approximately 90% in lichen communities at Lehigh
22 Water Gap near the zinc smelter when compared with those at Delaware Water Gap. Zinc,
23 cadmium, and sulfur dioxide were present in concentrations toxic to some species near the
24 smelter; however, toxic zinc concentrations extended beyond the detectable limits of sulfur
25 dioxide (Nash, 1975). Experimental data suggest that lichen tolerance to Zn and Cd falls
26 between 200 and 600 ppm (Nash, 1975).
27 Though there has been no direct evidence of a physiological association between tree injury
28 and exposure to metals, heavy metals have been implicated because their deposition pattern is
29 correlated with forest decline. The role of heavy metals has been indicated by phytochelatin
30 measurements. Phytochelatins are intracellular metal-binding peptides that act as specific
31 indicators of metal stress. Because they are produced by plants as a response to sublethal
March 2001
4-16
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
concentrations of heavy metals, they can be used to indicate that heavy metals play a role in
forest decline (Gawel et al., 1996). Concentrations of heavy metals increased with altitude, as
did forest decline, and increased concentrations across the region showing increased levels of
forest injury, as well.
Phytochelatin concentrations were measured in red spruce and balsam fir (Abies balsamea)
needles throughout the 1993 growing season at 1000 m on Whiteface Mountain in New York.
Mean foliar concentrations in red spruce were consistently higher than in balsam fir from June
until August, with the greatest and most significant difference occurring at the peak of the
growing season in mid-July. In July, the phytochelatin concentrations were significantly higher
than at any other time measured. Balsam fir did exhibit this peak, but maintained a consistently
low level throughout the season. Both the number of dead red spruce trees and phytochelatin
concentrations increased sharply with elevation (Gawel et al., 1996). The relationship between
heavy metals and the decline of forests in northeastern United States was further tested by
sampling red spruce stands showing varying degrees of decline at 1000 m on nine mountains
spanning New Hampshire, Vermont, and New York. The collected samples indicated a
systematic and significant increase in phytochelatin concentrations associated with the extent of
tree injury. The highest phytochelatin concentrations were measured during 1994 from sites
most severely affected by forest decline in the Green Mountains, VT, and the Adirondack
Mountains, NY. These data strongly imply that metal stress is a cause of tree injury and,
therefore, contributes to forest decline in the northeastern United States (Gawel et al., 1996).
One potential direct impact of heavy metals is on the activity of microorganisms and
arthropods resident on and in the leaf surface ecosystem. The fungi and bacteria living on and in
the surfaces of leaves play an important role in the microbial succession that prepares leaves for
decay and litter decomposition after their fall (U.S. Environmental Protection Agency, 1996b).
Numerous fungi were consistently isolated from foliar surfaces, at various crown positions,
from London plane trees growing in roadside environments in New Haven, CT. Those existing
primarily as parasites included Aureobasidium pullulans, Chaetomium sp., Cladosporium sp.,
Epicoccum sp., and Philaphora verrucosa. Those existing primarily as parasites included
Gnomoniaplatani, Pestalotiposis sp., andPleurophomella sp. The following cations were tested
in vitro for their ability to influence the growth of these fungi: cadmium, copper, manganese,
aluminum, chromium, nickel, iron, lead, sodium, and zinc. Results indicated variable fungal
March 2001
4-17
DRAFT-DO NOT QUOTE OR CITE
-------
1 response with no correlation between saprophytic or parasitic activity and sensitivity to heavy
2 metals. Both linear extension and dry weight data indicated that the saprophytic Chaetomum sp.
3 was very sensitive to numerous metals. Aureobasidium pullulans, Epicoccum sp., and especially
4 P.verrucosa, on the other hand, appeared to be much more tolerant. Of the parasites, G. platani
5 appeared to be more tolerant than Pestalotiopsis sp. and Pleurophomella sp. Metals exhibiting
6 the broadest spectrum growth suppression were iron, aluminum, nickel, zinc, manganese, and
7 lead (Smith and Staskawicz, 1977; Smith, 1990c). These in vitro studies employed soluble
8 compounds containing heavy metals. Trace metals probably occur naturally on leaf surfaces as
9 low-solubility oxides, halides, sulfates, sulfides, or phosphates (Clevenger et al., 1991; Koslow
10 et al., 1977). In the event of sufficient solubility and dose, however, changes in microbial
11 community structure on leaf surfaces because of heavy metal accumulation are possible.
12 Organic Compounds. Fine particles in the atmosphere reacting with volatilized chemical
13 compounds are partitioned between the gas and particle phases, depending on the liquid phase
14 vapor pressure at the ambient atmospheric temperature, the surface area of the particles per unit
15 volume of air, the nature of the particles and of the chemical being adsorbed and can be removed
16 by wet and dry deposition (McLachlan, 1996a). Materials as diverse as DDT, polychlorinated
17 biphenyls (PCBs), and polynuclear aromatic hydrocarbons (PAHs) are being deposited from the
18 atmosphere on rural as well as urban landscapes (Kylin et al., 1994). Motor vehicles emit
19 particles to the atmosphere from several sources in addition to the tailpipe. Rogge et al. (1993)
20 inventoried the organic contaminants associated with fine particles (diameter <2.0 ^m) in road
21 dust, brake lining wear particles, and tire tread debris. In excess of 100 organic compounds were
22 identified in these samples, including n-alkanols, benzoic acids, benaldehydes, polyalkylene
23 glycol ethers, PAHs, oxy-PAH, steranes, hopanes, natural resins, and other compound classes.
24 A large number of PAHs, ranging from naphthalene (C10H8) to 5- and 6-ring and higher PAHs,
25 their alkyl-substituted analogues, and their oxygen- and nitrogen-containing derivatives are
26 emitted from motor vehicle sources (Seinfeld, 1989).
27 Plants may be used as environmental monitors to compare the deposition of PAH, POPs, or
28 SOCs between sites (e.g., urban versus rural) (Wagrowski and Kites, 1997; Ockenden et al.,
29 1998; McLachlan, 1999). Vegetation can be used qualitatively to indicate organic pollutant
30 levels as long as the mechanism of accumulation is considered. The substance may enter the
March 2001
4-18
DRAFT-DO NOT QUOTE OR CITE
-------
1 plant via the roots or, as mentioned above, deposited as a particle onto the waxy cuticle of leaves
2 or uptake thorough the stomata. The pathways are a function of the chemical and physical
3 properties of the pollutant, such as its lipophilicity, water solubility, vapor pressure (which
4 controls the vapro-particle partitioning) and Henry's law constant; environmental conditions,
5 such as ambient temperature and the organic content of the soil; and the plant species, which
6 controls the surface area and lipids available for accumulation (Simonich and Hites, 1995).
7 Ockenden et al. (1998) have observed that, for lipophilic POPs, atmospheric transfer to plant has
8 been the main avenue of accumulation. Plants can differentially accumulate POPs. Results have
9 shown differences between species with higher concentrations in the lichen (Hypogymnia
10 physiodes) than in pine needles (Pinus sylvestris). Even plants of the same species, because they
11 have different growth rates and different lipid contents (depending on the habitat in which they
12 are growing), have different rates of sequestering pollutants. These facts confound data
13 interpretations and must be taken into account when considering their use as passive samplers.
14 Vegetation itself is an important source of hydrocarbon aerosols. Terpenes, particularly
15 a-pinene, p-pinene, and limonene released from tree foliage, may react in the atmosphere to form
16 submicron particles. These naturally generated organic particles contribute significantly to the
17 blue haze aerosols formed naturally over forested areas (Smith, 1990d).
18 The low water solubility with high lipoaffinity of many of these organic xenobiotics
19 strongly control their interaction with the vegetative components of natural ecosystems. The
20 cuticles of foliar surfaces are covered with a wax layer that helps protect plants from moisture
21 and short-wave radiation stress. This epicuticular wax, consisting mainly of long-chain esters,
22 polyesters, and paraffins, has been demonstrated to accumulate lipophilic compounds. Organic
23 air contaminants, in the particulate or vapor phase, are absorbed to and accumulate in the
24 epicuticular wax of vegetative surfaces (Gaggi et al., 1985; Kylin et al., 1994). Direct uptake of
25 organic contaminants through the cuticle or the vapor-phase uptake through the stomates are
26 characterized poorly for most trace organics.
27 The phytotoxiciry and microbial toxicity of organic contaminants to soil microorganisms is
28 not well studied (Foster, 1991).
March 2001
4-19
DRAFT-DO NOT QUOTE OR CITE
-------
1 4.2.2 Particulate Matter Effects on Natural Ecosystems
2 4.2.2.1 Introduction
3 Human existence on this planet depends on nature and the life-support services ecosystems
4 provide. Ecosystem services (Table 4-2) are the conditions and processes through which natural
5 ecosystems, and the species of which they are comprised, sustain and fulfill human life (Daily,
6 1997). Both ecosystem structure and function play an essential role in providing societal
7 benefits. Society derives two types of benefits from the structural aspects of an ecosystem:
8 (1) products with market value such as fish, minerals, forage, forest products, biomass fuels,
9 natural fiber, and many Pharmaceuticals and the genetic resources of valuable species (e.g.,
10 plants for crops and timber, animals for domestication); and (2) ecosystem services (Table 4-2)
11 include the use and appreciation of ecosystems for recreation, aesthetic enjoyment, and study
12 (Westman, 1977; Daily, 1997). Economic benefits and values associated with ecosystem
13 functions and services and the need to preserve them because of their value to human life are
14 discussed by Costanza et al. (1997) and (Pimentel et al., 1997). Services usually are not
15 considered to be items with market value.
TABLE 4-2. ECOSYSTEM SERVICES
• Purification of air and water
• Mitigation of floods and droughts
• Detoxification and decomposition of wastes
• Generation and renewal of soil and soil fertility
• Pollination of crops and natural vegetation
• Control of the vast majority of potential agricultural pests
• Dispersal of seeds and translocation of nutrients
• Maintenance of biodiversity, from which humanity has derived key elements of its
agricultural, medicinal, and industrial enterprises
• Protection from the sun's harmful rays
• Partial stabilization of climate
• Moderation of temperature extremes and the force of winds and waves
• Support of diverse human cultures
• Providing of aesthetic beauty and intellectual stimulation that lift the human spirit
Source: Daily (1997).
March 2001
4-20
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Ecosystems are structurally complex biotic communities consisting of populations of
plants, animals, insects, and microorganisms interacting with one another and with their abiotic
environment (Odum, 1993). They are dynamic, self-adjusting, self-maintaining, complex
adaptive systems in which patterns at higher levels of organization emerge from localized
interactions and selection processes. Macroscopic ecosystem properties such as structure,
diversity-productivity relationships and patterns of nutrient flux emerge from the interactions
among components and may feed back to influence subsequent development of those
interactions. The relationship between structure and function is a fundamental one in ecosystem
science (Levin, 1998). Structure refers to the species, their biodiversity, abundance, mass, and
arrangement within an ecosystem. Ecosystem functions, energy flow, nutrient flux, and water
and material flow, are characterized by the way in which ecosystem components interact.
Elucidating these interactions across scales is fundamental to understanding the relationships
between biodiversity and ecosystem functioning (Levin, 1998). To function properly and
maintain themselves, ecosystem components must have an adequate supply of energy, chemical
nutrients, and water. It is the flows of nutrients and energy, that provide the interconnectedness
between ecosystem parts and transforms the community from a random collection of species into
an integrated whole, an ecosystem in which the biotic and abiotic parts are interrelated (Levin,
1998).
Growth of new trees and other vegetation requires energy in the form of carbon
compounds. Plants accumulate, store, and use carbon compounds to build their structures and
maintain physiological processes. Plants, using energy from sunlight, hi their leaves combine
carbon dioxide from the atmosphere and water from the soil to produce the carbon compounds
(sugars) that provide the energy required by vegetation for growth and maintenance (Waring and
Schlesinger, 1985). Energy is transferred through an ecosystem from organism to organism in
food webs and, finally, is dissipated into the atmosphere as heat (Odum, 1993). Chemical
nutrients, such as nitrogen, phosphorus, or sulfur, on the other hand, are taken up from the soil by
plants and are transferred to other species through the food webs. The process is cyclic with the
chemical nutrients eventually returning to the soil. This process is referred to as biogeochemical
cycling (Odum, 1993). The biogeochemistry of an ecosystem is influenced by vegetation growth
characteristics (Herbert et al., 1999).
March 2001
4-21
DRAFT-DO NOT QUOTE OR CITE
-------
1 Ecosystem functions are characterized by the way components interact. These are the
2 functions that maintain clean water, pure air, a green earth, and a balance of organisms, the
3 functions that enable humans to survive. They are the dynamics of ecosystems. The benefits
4 they impart include absorption and breakdown of pollutants, cycling of nutrients, binding of soil,
5 degradation of organic waste, maintenance of a balance of gases in the air, regulation of radiation
6 balance, climate, and the fixation of solar energy (Table 4-2; Westman, 1977; Daily, 1997).
7 Concern has risen in recent years regarding the consequences of changing biological
8 diversity of ecosystems (Tilman, 2000; Ayensu et al, 1999; Wall, 1999; Hooper and Vitousek,
9 1997; Chapin et al., 1998). The concerns arise because human activities are creating
10 disturbances that are causing the loss of biodiversity and altering the complexity and stability of
11 ecosystems and producing changes in nutrient cycling (structure and function) (Pimm, 1984;
12 Levin, 1998; Chapin et al., 1998; Peterson et al., 1998; Tilman, 1996; Tilman and Downing,
13 1994; Wall, 1999; Daily and Ehrlich, 1999). There are few ecosystems on earth today that are
14 not influenced by humans (Freudenburg and Alario, 1999; Vitousek et al., 1997; Matson et al.
15 1997; Noble and Dirzo, 1997). The scientific literature is filled with references discussing the
16 importance of ecosystem structure and function. Ecorisk, complexity, stability, biodiversity,
17 resilience, sustainability, managing earth's ecosystems, and ecosystem health are frequently
18 . discussed topics. There is a need, therefore, to understand how ecosystems respond to both
19 natural and anthropogenic stresses and, especially, the ways that anthropogenic stresses are
20 impacting ecosystem services and products.
21
22 4.2.2.2 Ecosystem Responses to Stress
23 Ecosystem responses to stresses begin at the population level. Population changes
24 however, begin with the response of individual plants or animals. Plant responses, both
25 structural and functional, must be scaled hi both tune and space and propagated from the
26 individual to the more complex levels of community interaction to produce observable changes
27 in an ecosystem (Figure 4-1). At least three levels of biological interaction are involved: (l)the
28 individual plant and its environment, (2) the population and its environment, and
29 (3) the biological community composed of many species and its environment (Billings, 1978).
30 The response of individual organisms within a population based on their genetic constitution
31 (genotype), stage of growth at time of exposure, and the microhabitats in which they are growing
March 2001
4-22
DRAFT-DO NOT QUOTE OR CITE
-------
"X. Reactic
Level of^
Organizatior
Leaf
(cm2)
Branch
(cm2)
Tree
(m2)
Stand
(ha)
nTime
Minute
Day
V
•^
Year
»•
k» •
9r\
m •
^S 1
1-^3
Decade
2
3
»4
»-5
*•§
>9
|-— ^
Vte
M*
1
Century
10
11
12
13
L.14
16
Injury Symptom
Needle necrosis
and abscission
Branch length,
bifurcation ratio,
and ring-width
growth altered
Reduction in
diameter and death
of tree
Decreases in
stand productivity,
increases in mortality
and alterations in
regeneration patterns
Key Changes in Processes
Reduced carbon assimilation
because of reduced radiation
Reduced carbon available for foliage
replacement and branch growth/
export Synergistic interaction
between mistletoe and tephra
deposition
Reduced carbon available for
height, crown, and stem growth
Influence of crown class on initial '
impact and subsequent recovery
Interaction between stand
composition and recovery
For a given level, the dot associated with a line begins with a process (e.g., photosynthesis for #1 under leaf) an
ends with the associated structure (e.g., the needle).
Evaluating Impacts Within a Level of Organization
Leaf Level
Branch Level
Carbon exchange-1
Carbon pools-2
Needle number and size-3
Needle retention/abscission-4
Carbon allocation-5
Branch growth-6
Branch morphology-7
Branch vigor-8
Branch retention-9
Tree Level Height and diameter growth-10
Crown shape and size-1 1
Tree vigor-12
Mortality-13
Stand Level Productivity-14
Mortality-15
Species composition-16
Evaluating Interactions Between Different Levels of Organization
The diagonal arrow indicates the interaction between any two levels of organization.
The types of interaction are due to the properties of variability and compensation.
A - Refers to the interaction between the leaf and branch levels, where, for example,
variability at the branch level determines leaf quantity, and compensation at the leaf
level in photosynthesis may compensate for the reduction in foliage amount.
B - Refers to the interaction between the branch and the tree, where variability in branches
determines initial interception, branch vigor, and branch location in the crown;
compensation may be related to increased radiation reaching lower branches.
C - Refers to the interaction between the tree and the stand. Both genetic and
environmental variability, inter- and intraspecific compensations, and tree historical
and competitive synergisms are involved.
Figure 4-1. Effects of environmental stress on forest trees are presented on a hierarchial
scale for the leaf, branch, tree, and stand levels of organization. The
evaluation of impacts within a level of organization are indicated by horizontal
arrows. The evaluation of interactions between different levels of organization
are indicated by diagonal arrows.
Source: Hinckleyetal. (1992).
March 2001
4-23
DRAFT-DO NOT QUOTE OR CITE
-------
1 vary in their ability to withstand the stress of environmental changes (Levin, 1998). Individual
2 organisms within a population vary in their ability to withstand the stress of environmental
3 changes. The range within which these organisms can exist and function determines the ability
4 of the population to survive. Those able to cope with the stresses survive and reproduce.
5 Competition among the different species results in succession (community change over time) and
6 ultimately produces ecosystems composed of populations of plant species that have the capability
7 to tolerate the stresses (Rapport and Whitford, 1999; Guderian, 1985).
8 The number of species in a community usually increases during succession in unpolluted
9 atmospheres. Productivity, biomass, community height, and structural complexity increase.
10 Severe stresses, on the other hand, divert energy from growth and reproduction to maintenance,
11 and return succession to an earlier stage (Waring and Schlesinger, 1985). Ecosystems are subject
12 to natural periodic stresses, such as drought, flooding, fire, and attacks by biotic pathogens (e.g.,
13 fungi, insects). Ecosystem perturbation by natural stresses can be only a temporary setback.
14 Extremely severe natural perturbations return succession to an earlier stage, reduce ecosystem
15 structure (scarcity of life forms and no symbiotic interactions) and functions, disrupt the plant
16 processes of photosynthesis and nutrient uptake, carbon allocation and transformation that are
17 directly related to energy flow and nutrient cycling, shorten food chains, and reduce the total
18 • nutrient inventory (Odum, 1993). This transformation, however, sets the stage for recovery,
19 which permits the perturbed ecosystem to adapt to changing environments (Rolling, 1986).
20 Therefore, these perturbations are seldom more than a temporary setback, and recovery can be
21 rapid (Odum, 1969).
22 In contrast, anthropogenic stresses usually are severe, debilitating stresses. Severely
23 stressed ecosystems do not recover readily, but may be further degraded (Odum, 1969; Rapport
24 and Whitford, 1999). Anthropogenic stresses can be classified into four main groups:
25 (1) physical restructuring (e.g., changes resulting from land use); (2) introduction of exotic
26 species; (3) over harvesting; and (4) discharge of toxic substances into the atmosphere, onto land,
27 and into water. Ecosystems lack the capacity to adapt to the above stresses and maintain their
28 normal structure and functions unless the stress is removed (Rapport and Whitford, 1999). These
29 stresses result in a process of degradation marked by a decrease in biodiversity, reduced primary
30 and secondary production, and a lower capacity to recover and return to its original state.
31 In addition, there is an increased prevalence of disease, reduced nutrient cycling, increased
March 2001
4-24
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
dominance of exotic species, and increased dominance by smaller, short-lived opportunistic
species (Odum, 1985; Rapport and Whitford, 1999). Discharge of toxic substances into the
atmosphere, onto land, and into water can cause acute and chronic stresses and, once the stress is
removed, a process of succession begins which can ultimately return the ecosystem to a
semblance of its former structure. Air pollution stresses, if acute, are usually short term and the
effects soon visible. Chronic stresses, on the other hand, are long-term stresses whose effects
occur at different levels of ecosystem organization and appear only after long-term exposures, as
in the case of acidic deposition in the northeast or ozone in California (Shortle and Bondietti,
1992; U.S. Environmental Protection Agency, 1996b).
The possible effects of air pollutants on ecosystems have been categorized by Guderian
(1977) as follows:
(1) accumulation of pollutants in the plant and other ecosystem components (such as soil
and surface- and groundwater),
(2) damage to consumers as a result of pollutant accumulation,
(3) changes in species diversity because of shifts in competition,
(4) disruption of biogeochemical cycles,
(5) disruption of stability and reduction in the ability of self-regulation,
(6) breakdown of stands and associations, and
(7) expanses of denuded zones.
How changes in these functions can result from PM deposition and influence ecosystems is
discussed in the following text. It should be remembered that, although the effects of PM are
being emphasized, the vegetational components of ecosystems also are responding to multiple
stresses from other sources.
4.2.2.3 Ecosystem Response to Direct Plant Effects
The presence of PM in the atmosphere may affect vegetation directly, following physical
contact with the foliar surface (Section 4.2), but in most cases, the more significant impacts are
indirect. These impacts may be mediated by suspended PM (i.e., through effects on radiation and
climate) and by particles that pass through the vegetative canopies to the soil. Particulate matter,
as considered in this chapter is a heterogeneous mixture of particles differing in size, origin, and
chemical constituents and their impacts vary depending on the chemical nature of PM being
March 2001
4-25
DRAFT-DO NOT QUOTE OR CITE
-------
1 deposited on vegetation or soil. Particulate inputs and ecosystem cycling of key elements are
2 considered below.
3 The majority of studies dealing with direct effects of particulate dust and trace metals on
4 vegetation have focused on responses .of individual plant species and were conducted in the
5 laboratory or in controlled environments (Saunders and Godzik, 1986). A few have considered
6 the effects of particles on populations, communities, and ecosystems. Most of these focused on
7 ecosystems in industrialized areas heavily polluted by deposits of both chemically inert and
8 active dusts. Effects can result from direct deposition or indirectly by deposition onto the soil.
9 Reductions in growth, yield, flowering, and reproduction of plants from particulate deposition
10 have been reported (Saunders and Godzik, 1986). Sensitivities of individual species have been
11 associated with changes in composition and structure of natural ecosystems.
12 Evidence from studies of effects of PM deposition, specifically chemically inert and active
13 dusts indicates that, within a population, plants exhibit a wide range of sensitivity, which is the
14 basis for the natural selection of tolerant individuals (Saunders and Godzik, 1986). Rapid
15 evolution of certain populations of tolerant species at sites with heavy trace element and nitrate
16 deposition has been observed. Tolerant individuals present in low frequencies in populations
17 when growing in unpolluted areas have been selected for tolerance at both the seedling and adult
18 . stages when exposed to trace metal or nitrate deposition (Ormrod, 1984; U.S. Environmental
19 Protection Agency, 1993). Chronic pollutant injury to a forest community may result in the loss
20 of sensitive species, loss of tree canopy, and maintenance of a residual cover of pollutant-tolerant
21 herbs or shrubs that are recognized as successional species (Table 4-3; Smith, 1974). Frequently,
22 trace metals that penetrate the above-ground plant parts are less injurious than when taken up
23 through the roots (Guderian, 1986).
24 Responses of ecosystems to stresses (unless severe or catastrophic) are difficult to
25 determine because the changes are subtle (Garner, 1991). This is particularly true of responses to
26 particles. Changes in the soil may not be observed until accumulation of the pollutant has
27 occurred for 10 or more years except in the severely polluted areas around heavily industrialized
28 point sources (Saunders and Godzik, 1986). hi addition, the presence of other co-occurring
29 pollutants makes it difficult to attribute the effects to PM alone. In other words, the potential for
30 alteration of ecosystem function and structure exists, but it is difficult to quantify, especially
March 2001
4-26
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-3. ECOSYSTEM FUNCTIONS IMPACTED BY AIR POLLUTION
EFFECTS ON TEMPERATE FOREST ECOSYSTEMS
Forest Soil and Vegetation: Activity and Response
Ecosystem Consequence and Impact
1. Forest tree reproduction, alteration, or inhibition
2. Forest nutrient cycling, alteration
a. Reduced litter decomposition
b. Increased plant and soil leaching and soil
weathering
c. Disturbance of microbial symbioses
3. Forest metabolism
a. Decreased photosynthesis
b. Increased respiration
c. Altered carbon allocation
4. Forest stress, alteration
a. Phytophagous insects, increased or decreased
activity
b. Microbial pathogens, increased or decreased
activity
c. Foliar damage increased by direct air pollution
influence
1. Altered species composition
2. Reduced growth, less biomass
3. Reduced growth, less biomass
4. Altered ecosystem stress:
increased or decreased insect infestations;
increased or decreased disease epidemics;
and reduced growth, less biomass, and
altered species composition
Source: Smith (1974).
1 . when there are other pollutants present in the ambient air, which may produce additive or
2 synergistic responses, even though PM concentrations may not be elevated.
3
4 Physical Effects
5 The direct effects of limestone dust on plants and ecosystems has been known for many
6 years. Long-term changes in the structure and composition of the seedling-shrub and sapling
7 strata of an experimental site near limestone quarries and processing plants in Giles County in
8 southwestern Virginia were reported by Brandt and Rhoades (1972, 1973). Dominant trees in the
9 control area, a part of the oak-chestnut association of the eastern deciduous forests of eastern
10 North America, were chestnut oak (Quercus prinus), red oak (Q. rubra), and red maple (Acer
11 rubrum). An abundance of uniformly distributed saplings and seedlings were visible under the
12 tree canopy, and herbs appeared in localized areas in canopy openings. Q. prinus dominated the
13 area, and the larger trees were 60 to 80 years old. The dusty site was dominated by white oak
March 2001
4-27
DRAFT-DO NOT QUOTE OR CITE
-------
1 (Q. alba), whereas Q. rubra and Tulip poplar (Liriodendron tulipifera) were subcodominants.
2 The largest trees were 100 years old and had necrotic leaves, peeling bark, and appeared to be in
3 generally poor condition except for L. tulipifera (which thrived in localized areas). The site
" 4 contained a tangled growth of seedlings and shrubs, a few saplings, and a prevalence of green
5 briar (Smilax spp.) and grape (Vitis spp). The sapling strata in the area was represented by Acer
6 rubrum, hickory (Carya spp.), dogwood (Cornusflorida), and hop-hornbeam (Ostrya
7 virginiana). Saplings of none of the leading dominant trees were of importance in this stratum.
8 The most obvious form of vegetation hi the seedling-shrub stratum, because of their tangled
9 appearance, were C.florida, Ostrya virginiana, redbud (Cercis canadensis), and sugar maple
10 (Acersaccarum).
11 Crust formation reduced photosynthesis, induced premature leaf fall, destruction of leaf
12 tissues, inhibited growth of new tissue and reduced the formation of carbohydrate needed for
13 normal growth and storage (Brandt and Rhoades, 1973). The authors (Brandt and Rhoades,
14 1972), citing Odum (1969), also stated that a result of the accumulation of toxic pollutants in the
15 biosphere as the result of human activities, is the simplification of both plant and animal
16 communities. In plant communities, structure is determined by sampling various strata within
17 the community. Each stratum comprises a particular life form (e.g., herbs, seedlings, saplings,
18 ' trees). Dust accumulation favored growth of some species and limited others. For example,
19 Acer saccharum was more abundant in all strata of the dusty site when compared with the control
20 site where it was present only as a seedling. The growth of L. tulipifera, C.florida,
21 O. virginiana, black haw (Viburnum prunifolium), and C. canadensis appeared to be favored by
22 the dust. Growth of conifers and acidophiles such as rhododendron (Rhododendron maximum),
23 however, was limited. Although dust accumulation began in 1945, the heaviest accumulation
24 occurred between 1967 and 1972 during the time of the study.
25 Changes in community composition were associated closely with changes in the growth of
26 the dominant trees. Decrease hi density of seedlings and saplings and in mean basal area, as well
27 as lateral growth of A. rubrum, Q. prinus, and Q. rubra, occurred in all strata. On the other hand,
28 all of these characteristics increased hi L. tulipifera, which was a subordinate species before dust
29 accumulation began but had assumed dominance at the tune of the study. Reduction in growth of
30 the dominant trees had apparently given L. tulipifera competitive advantage because of its ability
31 to tolerate dust. Changes in soil alkalinity occurred because of the heavy deposition of limestone
March 2001
4-28
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
dust; however, the facilities necessary for critical analysis of the soils were not available. From
the foregoing, it is obvious that PM physical effects in the vicinity of limestone quarries and
processing plants can impact ecosystems.
Acidic Deposition
The effects of acidic deposition have been discussed in several previous reports. The 1982
EPA document, Air Quality Criteria for Paniculate Matter and Sulfur Oxides, devoted a chapter
to the effects of acidic deposition (U.S. Environmental Protection Agency, 1982). In 1984, EPA
published The Acidic Deposition Phenomenon and Its Effects (Altshuller and Linthurst, 1984),
and, in 1991, NAPAP published the result of its extensive study, Acidic Deposition: State of
Science and Technology (Irving, 1991). The major effects of acidic deposition occur through the
soil and are discussed under indirect effects. However, included among the direct responses of
forest trees to acidic deposition are increased leaching of nutrients from foliage; accelerated
weathering of leaf cuticular surfaces; increased permeability of leaf surfaces to toxic materials,
water, and disease agents; and altered reproductive processes (Altshuller and Linthurst, 1984).
Trace Elements , ,
Possible direct responses of trace elements on vegetation result from their deposition and .
residence on the phyllosphere (i.e., leaf surfaces). Fungi and other microorganisms living on the
leaves of trees and other vegetation play an important role in leaf decomposition after litterfall
(Miller and McBride, 1999; Jensen, 1974; Millar, 1974). Early needle senescence and abscission
in the San Bernardino Forest changed fungal microflora successional and decomposition patterns
by altering the taxonomic diversity and population density of microflora that normally develop
on needles while they are on the tree. Changing the fungal community on the needles weakened
the decomposer community, decreasing the rate of decomposition, and altered nutrient cycling
(Bruhn, 1980). Nutrient availability was influenced by accumulation of carbohydrates and
mineral nutrients in the heavy litter under those stands with the most severe needle injury and
defoliation (U.S. Environmental Protection Agency, 1996b). Possible impacts of heavy metals
on nutrient cycling and their effects on leaf microflora appear not to have been studied.,
A trace metal must be brought into solution before it can enter into the leaves or bark of
vascular plants. Low solubility limits entry. In those instances when trace metals are absorbed,
March 2001
4-29
DRAFT-DO NOT QUOTE OR CITE
-------
1 they frequently are bound in the leaf tissue and then are lost when the leaf drops off (Hughes,
2 1981) and can affect litter decomposition, an important source of soil nutrients. Changes in litter
3 decomposition processes influence nutrient cycling in the soil and limit the supply of essential
4 nutrients. Both Cotrufo et al. (1995) and Niklinska et al. (1998) point out that heavy metals have
5 impacts on forest litter decomposition. Cotrufo et al. (1995) observed that decomposition of oak
6 leaves containing Fe, Zn, Cu, Cr, Ni, and Pb was influenced strongly during the early stages by
7 metal contamination. Fungal mycelium was significantly less abundant in litter and soil in
8 contaminated sites, when compared with control sites. Niklinska et al. (1998) stated that toxic
9 effects of heavy metals on soil respiration rate have been reported by many scientists, and that, in
10 polluted environments, this results in accumulation of undecomposed organic matter. However,
11 they state that results of experiments should identify the most important "natural" factors
12 affecting soil/litter sensitivity because the effects of heavy metals on respiration rates depend on
13 the dose of heavy metals, the type of litter, types of metals deposited, and the storage time before
14 respiration tests are made.
15 Trace metals, particularly heavy metals (e.g., cadmium, copper, lead, chromium, mercury,
16 nickel, zinc) have the greatest potential for influencing forest growth (Smith, 1991).
17 Experimental data indicate that the broadest spectrum of growth suppression of foliar microflora
18 • resulted from iron, aluminum, and zinc. These three metals also inhibited spore formation, as did
19 cadmium, chromium, manganese, and nickel (see Smith, 1990e). In the field, the greatest injury
20 occurs from pollution near rnining, smelting, and other industrial sources (Ormrod, 1984). Direct
21 metal phytotoxicity can occur only if the metal can move from the surface into the leaf or directly
22 from the soil into the root.
23
24 Organic Compounds
25 Secondary organic compounds formed in the atmosphere, the effects of some of which are
26 discussed below, have been referred to under the following terms: toxic substances, pesticides,
27 hazardous air pollutants (HAPS), air toxics, semivolatile organic compounds (SOCs), and
28 persistent organic pollutants (POPS). Again, it should be noted that the chemical substances
29 denoted by such headings are not criteria air pollutants controlled by the NAAQS under
30 Section 109 of the Clean Air Act (CAA) (U.S. Code, 1991), but rather are controlled under
31 Sect. 112, Hazardous Air Pollutants. Their possible effects on humans and ecosystems are
March 2001
4-30
DRAFT-DO NOT QUOTE OR CITE
-------
1 discussed in a number of government documents and in many other publications. They are
2 mentioned here because, in the atmosphere, many of the chemical compounds are partitioned
3 between gas and particle phases. As particles, they can become airborne, be distributed over a
4 wide area, and impact remote ecosystems. Some of the chemical compounds are of concern
5 because they may reach toxic levels in food chains of both animals and humans, whereas others
6 tend to decrease or maintain the same toxicity as they move through the food chain. Some
7 examples of movement through food chains are provided below.
8 Many chemical compounds from a variety of anthropogenic sources are released into the
9 ambient air (See Section 4.2.1). In the atmosphere, the emitted compounds initially go through a
10 mixing process, and the airborne particles then are distributed over a wide area and ultimately
11 deposited on ecosystem components. Atmospheric deposition of polychlorinated dibenzo-p-
12 dioxins and dibenzofurans (PCDD/Fs), as an example, can be divided into three different forms:
13 (1) dry gaseous, (2) dry particle-bound, and (3) wet deposition. Dry particle-bound deposition
14 occurs when the PM containing the pollutant is deposited on the plant surface, whereas wet
15 deposition ranges from hail through rain to fog and dew fall (McLachlan, 1996b).
16 Human exposure to PCDD/Fs has been demonstrated to be caused almost exclusively by
17 the ingestion of animal fat from fish, meat, and dairy products. Almost half of human exposure
18 to PCDD/Fs is caused by consumption of beef and dairy products (McLachlan, 1996b). Cattle
19 obtain most of their PCCD/Fs though grass. Therefore, the grass-cattle-milk/beef pathway is
20 critical for human exposure. It has been shown that root uptake/translocation is an insignificant
21 pathway of PCDD/Fs to aerial plant parts. Wet and dry particle deposition are the most
22 important for the accumulation of the higher chlorinated cogeners in vegetation. The persistence
23 of PCDD/Fs in plants has not been investigated extensively; however, biodegradation probably
24 does not occur in that these compounds are found primarily in the lipophilic cuticle and are very
25 resistant to microbial degradation (McLachlan, 1996b). Feed contaminated with soil containing
26 the pollutant also can be another source of exposure of beef and dairy cattle as well as chickens.
27 The PCDD/Fs are near a steady state in milk cows and laying hens; however, animals raised for
28 meat production (such as beef cattle and pigs) may accumulate them. The beef cattle and pigs
29 cannot excrete the contaminants in a lipid-rich matrix such as milk or eggs. All of the PCDD/Fs,
30 ingested are stored in the body. In agricultural food chains, there is a biodilution of PCDD/Fs,
31 with the fugacity decreasing by up to three orders of magnitude between the air and cows milk
March 2001
4-31
DRAFT-DO NOT QUOTE OR CITE
-------
1 (McLachlan, 1996b). Fiirst et al. (1993), based on surveys to determine the factors that influence
2 the presence of PCDD/PCDF in cows milk, earlier concluded that regardless of which pathway,
3 soil - grass - cow or air - grass - cow, it was the congener of the chemical that was most
4 important.
5 Persistent polychlorinated pollutants (POPS), such as PCBs, PCDFs, and PCDDs, can be
6 transported as particles through the atmosphere from industrial and agricultural sources; be
7 brought down via wet and dry deposition in remote regions, such as the Arctic; and have been
8 detected in all levels of the Arctic food chain (Oehme et al., 1995). High concentrations of PCB
9 (1 to 10 ppm) were found in seals; but the concentrations increased to 10 to 100 ppm in polar
10 bears. The polar bear is the top predator in the Arctic and feeds preferentially on ringed seals and
11 also, to a lesser extent, on other seal species. Bioconcentration factors of organochlorines in the
12 Arctic food web, reaching 107 for fish and seals, are biomagnified in polar bears (Oehme et al.,
13 1995). Polychlorinated dibenzo-/?-dioxins (PCDDs) and polychlorinated dibenzofurans
14 (PDCF/s) also have also been found in seals (Oehme et al., 1995). Milk taken from
15 anaesthetized polar bears was also found to contain PCDD/PCDF. Very little is known regarding
16 the intake of milk by polar bear cubs. However, estimates of the intake of milk containing
17 detectable levels of PCDD/PCDF and PCB and the additional consumption of seal blubber
18 ' confirm that these pollutants are passed on to the next generation (Oehme et al., 1995).
19 Section 112 of the CAA, provides the legislative basis for U.S. hazardous air pollutant
20 (HAP) programs. In response to mounting evidence that air pollution contributes to water
21 pollution, Congress included Section 112m (Atmospheric Deposition to Great Lakes and Coastal
22 Waters) in the 1990 CAA Amendments, which directs the Environmental Protection Agency
23 (EPA) to establish a research program on atmospheric deposition of HAPS to the "Great
24 Waters".
25 Actions taken by EPA and others to evaluate and control sources of Great Waters pollutants
26 of concern appear to have positively affected trends in pollutant concentrations measured in air,
27 sediment, and biota. Details concerning these effects may be found in "Deposition of Air
28 Pollutants to the Great Waters", Third Report to Congress (U. S. Environmental Protection
29 Agency, 2000a). The Third Report (EPA-453/R-00-005, June 2000), like the First and Second
30 Reports to Congress, focuses on 15 pollutants of concern, including pesticides, metal
31 compounds, chlorinated organic compounds, and nitrogen compounds. The new scientific
March 2001
4-32
DRAFT-DO NOT QUOTE OR CITE
-------
1 information in the Third Report supports and builds on three broad conclusions presented in the
2 previous two EPA Reports to Congress and discussed below.
3 (1) Atmospheric deposition from human activities can be a significant contributor of toxic
4 chemicals and nitrogen compounds to the Great Waters. The relative importance of
5 atmospheric loading for a particular chemical in a water body depends on many factors (e.g.,
6 characteristics of the water body, properties of the chemical, and the kind and amount of
7 atmospheric deposition versus or water discharges).
8 (2) A plausible link exists between emissions into the air of Great Waters toxic pollutants of
9 concern; the atmospheric deposition of these pollutants (and their transformation products);
10 and the concentrations of these pollutants found in the water, sediments, and biota, especially
11 fish and shellfish. For mercury, fate and transport modeling and exposure assessments
12 predict that the anthropogenic contribution to the total amount of methylmercury in fish is, in
13 part, the result of anthropogenic mercury releases from industrial and combustion sources
14 increasing mercury body burdens (i.e., concentrations) in fish. Also, the consumption of fish
15 is the dominant pathway of exposure to methylmercury for fish-consuming humans and
16 wildlife. However, what is known about each stage of this process varies with each pollutant
17 (for instance, the chemical species of the emissions and its transformation in the
18 atmosphere).
19 (3) Airborne emissions from local as well as distant sources, from both within and outside the
20 United States, contribute pollutant loadings to waters through atmospheric deposition.
21 Determining the relative roles of particular sources—local, regional, national, and possibly
22 global, as well as anthropogenic, natural, and reemission of pollutants—contributing to
23 specific water bodies is complex, requiring careful monitoring, atmospheric modeling, and
24 other analytical techniques.
25
26 4.2.2.4 Indirect Effects of Particulate Matter In Ecosystems
27 The presence of PM in the atmosphere directly affects vegetation following physical
28 contact with foliar surfaces (as discussed above in Section 4.2.2.2), but in many cases the more
29 significant impacts are indirect. These impacts may be mediated by suspended PM (i.e., through
30 effects on radiation and climate) and by particles that pass through vegetative canopies to reach
31 the soil. Effects mediated in the atmosphere are considered briefly below and in greater detail
March 2001
4-33
DRAFT-DO NOT QUOTE OR CITE
-------
1 later, under Section 4.5. Indirect plant responses are chiefly soil mediated and depend primarily
2 on the chemical composition of the individual elements deposited in PM. The individual
3 elements must be bioavailable to have an effect. The soil environment, composed of mineral and
4 organic matter, water, air, and a vast array of bacteria, fungi, algae, actinomycetes, protozoa,
5 nematodes, and arthropods, is one of the most dynamic sites of biological interactions in nature
6 (Wall and Moore, 1999; Alexander, 1977). The quantity of organisms in soils varies by locality.
7 Bacteria and fungi are usually most abundant in the rhizosphere, the soil around plant roots that
8 all mineral "nutrients must pass through. Bacteria and fungi benefit from the nutrients in the root
9 exudates (chiefly sugars) in the soil and, in turn, they play an essential role by making mineral
10 nutrients available for plant uptake (Wall and Moore, 1999; Rovira and Davey, 1974). Their
11 activities create chemical and biological changes in the rhizosphere by decomposing organic
12 matter and making inorganic minerals available for plant uptake. Bacteria are essential in the
13 nitrogen and sulfur cycles and make these elements available for plant uptake and growth (see
14 Section 4.3.3). Fungi are directly essential to plant growth. Attracted to the roots by the
15 exudates, they develop mycorrhizae, a mutualistic, symbiotic relationship, that is integral in the
16 uptake of the mineral nutrients (Allen, 1991). The impact in ecosystems of PM, particularly
17 nitrates, sulfates, and metals, is determined by their affect on the growth of the bacteria involved
18' in nutrient cycling and the fungi involved in plant nutrient uptake.
19
20 Particulate Matter-Related Atmospheric Turbidity: Effects on Vegetative Processes
21 Photosynthetic processes underlie the contribution of vegetative surfaces to nutrient and
22 energy cycling. Photosynthesis and the heat-driven processes of water cycling depend on net
23 receipts and characteristics of the radiation environment. These characteristics may be altered
24 substantially when the atmosphere becomes turbid because of particulate loading.
25 Specific wavelengths of interest depend on the vegetation process under consideration.
26 Canopy temperature and water relations are particularly sensitive to long-wave, infrared
27 radiation, whereas primary photosynthetic charge separations depend on short-wave radiation in
28 the visible and photosynthetically active range (0.4 to 0.7/mi).
29 Effects of anthropogenic aerosols on the radiation environment at the Earth's surface are
30 difficult to assess. The residence time of suspended particles varies with size and environmental
31 conditions (seconds to months or years), and concentrations are spatially and temporally variable.
March 2001
4-34
DRAFT-DO NOT QUOTE OR CITE
-------
1 In particularly polluted urban and near-urban areas, unambiguous particulate impacts on radiation
2 and local climate may be observed. Visibility was degraded by 50% in a large plume originating
3 in the St. Louis urban area during the midweek, midday period (Pueschel, 1993). In contrast,
4 visibility was reduced by only 20% on weekends, when traffic and industrial emissions were
5 reduced. The area affected by this plume includes highly productive agricultural land.
6 Empirical relationships between mass of specific components of the aerosol and radiation
7 scattering have been developed (e.g., Pueschel, 1993), from which regional visibility (or
8 radiation attenuation) isopleths can be constructed if appropriate mass data are available. These
9 estimates support trends observed by direct measurement of turbidity (e.g., Flowers et al., 1969;
10 U.S. Environmental Protection Agency, 1982).
11 Sulfates, nitrates, and elemental carbon dominate effects on visibility, in part, because they
12 frequently dominate the mass profiles and, in part, because they exhibit particularly large
13 absorption coefficients (see Section 4.3). Absorption by particles containing carbon may range
14 from 5 to 10% in rural areas to up to 50% in urban areas (U.S. Environmental Protection Agency,
15 1982). In west-coast cities with contrasting particulate sources and loadings, the common
16 component that related PM to visibility degradation was sulfate between 0.65 and 3.6 yum
17 (Barone et al., 1978). For example, in Los Angeles, sulfate and nitrate had similar effects on
18 visibility (White, 1976), despite the dominance of nitrate from transportation sources in the
19 aerosol, although this is changing with controls on point sources of sulfate (Farber et al., 1994).
20 No long-term global trend of increasing atmospheric optical depth has been documented
21 (Bolle et al., 1986; Pueschel, 1993), although seasonal and regional impacts are substantial. The
22 classic study by Flowers et al. (1969) demonstrated large regional distinctions in turbidity across
23 the United States. Typically, the western deserts, plains, and Rocky Mountains exhibited low
24 mean annual turbidity, whereas the more humid and densely vegetated eastern half of the country
25 exhibited much greater turbidities. In the mid-1970s, visible range hi the mountainous southwest
26 exceeded 110 km and radiation attenuation was ca. 2.6%; whereas, in the east, visible range was
27 below 24 km and radiation attenuation was ca. 10%. Visibility in the eastern United States has
28 decreased generally since the 1940s (Flowers et al., 1969; Trijonis and Shapland, 1979; U.S.
29 Environmental Protection Agency, 1982). Correlative trends in visibility degradation and
30 emissions of sulfur oxides suggest that particulate sulfate may account for much of the turbidity.
March 2001
4-35
DRAFT-DO NOT QUOTE OR CITE
-------
1 These trends are typical of urban industrial areas around the world. Turbidity has increased
2 above Mexico City (Binenko and Harshvardhan, 1993) since the 1911 to 1928 period. During
3 this early period, a single annual peak of turbidity coincided with the end of the dry period, and
4 natural sources dominated. By 1957 to 1962, the number of annual peaks had increased, as
5 anthropogenic sources came to dominate. During this period, atmospheric transmission of direct-
6 beam solar radiation decreased by about 10% (Binenko and Harshvardhan, 1993). Visibility in
7 the Los Angeles basin has improved very slightly in the past decades (Farber et al., 1994), as
8 sulfate emissions have been controlled by regulation. The composition of the aerosol has
9 changed, particularly in inland areas, as the former dominance of sulfate shifts to a
10 preponderance of secondary organics.
11 Particles interact with solar radiation through scattering and absorption. Absorption of
12 short-wavelength solar radiation reduces the amount of radiation reaching the Earth's surface and
13 leads to atmospheric heating. If the absorbing particles reradiate in the infrared range, then some
14 of this energy is lost as long-wave reradiation to space. This loss mechanism is minimized
15 because most of the anthropogenic aerosol in the troposphere resides in the planetary boundary
16 layer (Bolle et al., 1986), even within the lower 500 m (Binenko and Harshvardhan, 1993), where
17 the temperature is similar to that of the surface. Some of this energy is captured at the surface as
18 down-welling infrared radiation.
19 These wavelengths directly impact canopy temperatures and influence transpirational water
20 use by vegetation. The presence of absorbing aerosols reduces the ratio of photosynthetically
21 active radiation to total radiation received at the surface, potentially reducing photosynthetic
22 water use efficiency. The net effect of aerosol absorption on the surface depends on the relative
23 magnitudes of the particulate absorption coefficients in the visible and infrared area and on the
24 albedo of the Earth's surface. In general, absorption is not a dominant particulate effect.
25 Scattering of radiation dominates the effects of particulate loading on visibility and
26 turbidity. Nonabsorbing, scattering aerosols raise the overall albedo of the atmosphere and
27 reduce the amount of radiation reaching the surface by the amount reflected or backscattered to
28 space. As atmospheric turbidity increases, so does the scattering of light, including forward
29 scattering of photosynthetically active radiation that intercepts the Earth's surface (Hoyt, 1978).
30 The largest effect is described by Mie-scattering theory. Forward scattering reduces the
31 intensity of direct radiation by disrupting the solar beam, thereby increasing the path length and
March 2001
4-36
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
probability of absorption and also increases the intensity of diffuse (sky) radiation. In a clear
atmosphere, diffuse radiation may be on the order of 10% of total solar radiation (Choudhury,
1987). However, in highly turbid, humid conditions, this fraction may increase, even up to 100%
of solar radiation in extreme cases. The direct-to-diffuse-radiation ratio is highest at solar noon
and lowest near dawn or dusk, when the path length through the atmosphere is longest.
Particle scattering is wavelength dependent, causing objects to appear blue- or red- tinged,
depending on viewing and illumination angles and on the light quality, the alteration of which is
a minor contributor to photosynthetic light-use efficiency. The wavelength dependence of
scattering decreases rapidly from extreme sensitivity for very fine particles to little dependence at
10 /j,m. Equations relating scattering at a reference wavelength to scattering at wavelengths of
interest are rigorously applicable only to spherical particles but may be extended to nonspherical
particles of equal volume (Janzen, 1980).
World Meteorological Organization (WHO) data summarized in U.S. Environmental
Protection Agency (1982) indicated that turbidity in the eastern United States commonly resulted
in radiation losses of ca. 3.5% because of backscattered radiation and ca. 3.5% because of
absorption, with a resulting total reduction of incident radiation to ca. 93% of total solar
radiation. However, 28% of the radiation reaching the surface was converted from direct
radiation to diffuse, or sky, radiation. Under more polluted condition's, losses were ca.' 9%
backscattered and 9% absorbed, reducing total radiation to 82% of total solar radiation and
converting 72% from direct beam to diffuse radiation. Photosynthetically active radiation (0.4 to
0.7 //m) typically is enriched in diffuse radiation relative to total or direct beam radiation.
Altered Radiative Flux: Effects on Vegetative Processes
Canopy photosynthesis is typically a nearly linear function of incident radiation,
overcoming saturation exhibited by individual leaves by distributing the light throughout the
multilayer canopy. Light penetration into canopies limits photosynthetic productivity (Rosenberg
et al., 1983). The uppermost leaves of many canopies are at or above light saturation for
photosynthetic processes. The simplest radiative transfer functions describing plant canopies
relate total down-welling radiation (direct plus diffuse radiation measured above the canopy) to
radiation interception at each leaf level through a Beer's Law analogy. The expected exponential
decline in radiation through the canopy depends only on total radiation and a bulk canopy
March 2001
4-37
DRAFT-DO NOT QUOTE OR CITE
-------
1 extinction coefficient that depends on leaf size, orientation, and distribution, as well as on
2 reflectance and absorption in wavelengths of interest. These simplified models predict radiation
3 distribution adequately for homogeneous canopies. Turbidity affects canopy processes only by
4 attenuating the total radiation impinging on the canopy surface.
5 In more complex, and more realistic, canopy-response models (e.g., Choudhury, 1987),
6 radiation is considered in its direct and diffuse components. Foliar interception by canopy
7 elements is considered for both up- and down-welling radiation (a two-stream approximation).
8 In this case, the effect of atmospheric PM on turbidity affects canopy processes both by radiation
9 attenuation and by influencing the efficiency of radiation interception throughout the canopy
10 through conversion of direct to diffuse radiation (Hoyt, 1978). Diffuse radiation is more
11 uniformly distributed throughout the canopy and increases canopy photosynthetic productivity by
12 distributing radiation to lower leaves. The treatment of down-welling direct-beam radiation in
13 the two-stream approach remains an elaboration of the simplified Beer's Law analogy, with solar
14 angle, leaf area distribution, and orientation individually parameterized (Choudhury, 1987).
15 Diffuse down-welling radiation is a function of diffuse and direct radiation at the top of the
16 canopy and penetration within the canopy, according to cumulative leaf area density and foliage
17 orientation. Up-welling (diffuse) radiation results from scattering and reflectance within the
18 canopy, and by the soil, of both direct and diffuse down-welling radiation.
19 The effect of the altered distribution between diffuse and direct radiation impacts
20 photosynthesis in upper, exposed leaves as a function of leaf angle and in total canopy
21 photosynthesis as a function of penetration of radiation within the canopy. This depends on
22 canopy structure, leaf optical properties, and leaf area density, as well as on solar angle and
23 atmospheric turbidity. Absorption of radiation by particles heats the upper atmosphere and
24 results in reduced vertical temperature gradients. This could reduce the intensity of atmospheric
25 turbulent mixing. The magnitude of such potential effects on turbulent transport within canopies
26 remains unknown, although damping of eddy transport could inhibit canopy gas exchange.
27 Suppressed tropospheric mixing also could intensify local temperature inversions and increase
28 the severity of pollution episodes (Pueschel, 1993), with direct inhibitory effects on
29 photosynthetic processes.
30 The most significant effect of aerosols on vegetation is probably through their role as cloud
31 condensation nuclei because clouds have substantial impact on radiation receipts at the surface.
March 2001
4-38
DRAFT-DO NOT QUOTE OR CITE
-------
1 An important characteristic of fine particles is their ability to affect the flux of solar radiation
2 passing through the atmosphere directly, by scattering and absorbing solar radiation, and
3 indirectly, by acting as cloud condensation nuclei which in turn influence the optical properties
4 of clouds (Chameides et al., 1999). Regional haze has been estimated to diminish surface solar
5 visible radiation by approximately 8%. Crop yields have been reported as being sensitive to the
6 amount of sunlight received. The potentially significant impact of regional haze on the yield of
7 crops because of reduction in solar radiation has been examined by Chameides et al. (1999).
8 Using a case study approach, Chameides et al. (1999), studied the affects of regional haze on
9 crop production in China, where regional haze is especially severe. A rudimentary assessment of
10 the direct effect of atmospheric aerosols on agriculture suggests that optimal crop yields of
11 approximately 70% of the crops are being depressed by at least 5 to 3% by regional scale air
12 pollution and its associated haze (Chameides et al., 1999).
13
14 Effects of Solar Ultraviolet Radiation on Terrestrial Ecosystems
15 The transmission of solar UV-B radiation through the earth's atmosphere is controlled by
16 ozone, clouds and particles. The depletion of stratospheric ozone caused by the release of
17 chlorofluorcarbons (CFCs) and other substances, such as halides, has resulted in heightened
18 concern about potentially deleterious increases in the amount of solar UV-B (SUVB) radiation
19 reaching the Earth's surface (see Section 4.5). One salient consideration is that, although CFC
20 production is at a peak level now, the problem likely will continue well into the future because of
21 the length of time it takes for molecules to reach the stratosphere (Greenberg et al, 1997).
22 The vulnerability of terrestrial plants to UV-B results from their requirement for sunlight
23 for photosynthesis. Each 1% decline in ozone has been predicted to decrease crops yield by 1%
24 (Greenberg et al., 1997). In addition to inhibiting photosynthesis, UV-B radiation triggers
25 numerous responses in plants (e.g., membrane, protein, and DNA damage; delayed maturation;
26 diminished growth; activation of chemical stress; flavonoid synthesis; leaf thickening)
27 (Table 4-4). It is not known which of the injury and damage effects are most detrimental to plant
28 growth (Table 4-4). Effects of increased UV-B on plant growth are likely to be incremental.
29 Because plants evolved under the selective pressure of ambient UV-B radiation in sunlight, they
30 have developed adaptive mechanisms (Greenberg et al., 1997). Although inhibition of
31 photosynthesis is a detrimental growth effect, flavonoid synthesis represents acclimation.
March 2001
4-39
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-4. TYPES OF PLANT RESPONSES TO ULTRAVIOLET-B RADIATION8
Acclimation and Morphological Responses Damage and Injury Responses -
Altered biomass distribution
Altered leaf cell division
Cotyledon curling
Increased DNA repair
Increased flavonoid biosynthesis
Increased leaf thickness
Increased leaf number
Increased number of tillars
Leaf wrinkling
Reduced leaf area
Reduced hypocotyl growth
Reduced shoot height
Reduced stomatal density
Altered gene expression
Degradation of auxin
Degradation of chlorophyll and carotenoids
Degradation of proteins
Diminished biomass
Epidermal collapse
Inhibition of growth
Inhibition of photosynthesis
Increased stomatal conductance
Lower seed yield
Oxidation of DNA
Peroxidation of lipids
Prymidine dimer formation
"Entries in alphabetical order.
1 Plants growing under full light have been shown to be protected against UV-B effects but not
2 when growing under weak visible light (Bjorn, 1996). A common adaptation is alteration in leaf
3 transmission properties, which results in attenuation of UV-B in the epidermis before it can reach
4 the leaf interior.
5 Plant species vary enormously in their response to UV-B exposures, and large differences
6 in response occur among different genotypes within a species. In general, dicotyledonous plants
7 are more sensitive than monocotyledons from similar environments. In addition, plant responses
8 may differ depending on stage of development. Therefore, extrapolation of experimental
9 responses from seedlings to mature plants must be taken with caution (Bjorn, 1996). The above
10 facts are especially important when considering the effects of UV-B on agricultural plants.
11 For example, among soybeans and rice, there are varieties for which growth and crop yield are
March 2001
4-40
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
severely decreased by increased UV-B radiation and other varieties that are not affected or may
even be stimulated. On the other hand, the growth of the same sensitive soybeans when grown
under water stress was not inhibited. Many crop plants grown in temperate regions originated in
more tropical areas, hence, a gene pool for more resistant varieties is likely to exist (Bjorn, 1996).
Crop plants, unlike forest trees and vegetation in natural ecosystems, are only exposed for one
generation, and thus, it may be possible to readily change the genotype if a variety proves to be
sensitive.
Trees, forests, and perennial evergreen plants are long-lived when compared to agricultural
systems, making it possible for UV-B exposure impacts to accumulate with time. Saplings and
young and small trees react differently when compared to mature trees; also, on evergreen trees,
needles of different ages respond differently (Bjorn, 1996). Breeding and testing trees is a slow
process, and, for this reason, much care needs to be taken when planting large areas with trees of
a single species and one provenance (e.g., Stika Spruce [Picea sitchensis] in Britain). The
response of only a few broad-leaved trees have been studied. The most investigated genus has
been loblolly pine (Pinus taeda) (Bjorn, 1996).
A few studies indicate that the photomorphogenesis (changes in leaf thickness under UV-B
that results in a transition from shade to sun leaves, Table 4-4) and the variable responses of
native plants in ecosystems to UV-B exposures results in changes in interactions between various
plants species, changes between plants and other organisms, and between plants and their abiotic
environment. These preliminary studies suggest that in natural ecosystems, composed of many
different plant species, with complex interactions between plants and between plants and other
organisms, there may develop effects of UV-B that cannot be determined from experiments on
single plant species. The effects of UV-B on natural plant systems, therefore, should be of
greater concern than on agricultural crops (Bjorn, 1996).
Nitrogen Deposition Effects
Nitrogen has long been recognized as the nutrient most important for plant growth. Plants
usually absorb nitrogen through their roots by absorbing NH4 + or NO3" or informed by symbiotic
organisms in the roots. Plants, however, vary in their ability to absorb ammonium and nitrate
(Chapin et al., 1987). Nitrogen is of overriding importance in plant metabolism and, to a large
extent, governs the utilization of phosphorus, potassium, and other nutrients. Most of the
March 2001
4-41
DRAFT-DO NOT QUOTE OR CITE
-------
1 nitrogen in soils is associated with organic matter. Typically, the availability of nitrogen via the
2 nitrogen cycle controls net primary productivity and possibly the decomposition rate of plant
3 litter. Photosynthesis is influenced by nitrogen uptake in that ca. 75% of the nitrogen in a plant
4 leaf is used during the process of photosynthesis. The nitrogen-photosynthesis relationship is,
5 therefore, critical to the growth of trees and other plants (Chapin et al, 1987).
6 Because nitrogen is not readily available and is usually in shortest supply, it is the chief
7 element in agricultural fertilizers. Atmospherically deposited nitrogen also can act as a fertilizer
8 in soil low in nitrogen. Not all plants, however, are capable of utilizing extra nitrogen. Inputs of
9 nitrogen to natural ecosystems that alleviate deficiencies and increase growth of some plants can
10 impact competitive relationships and alter species composition and diversity (Ellenberg, 1987;
11 Kenk and Fischer, 1988; U.S. Environmental Protection Agency, 1993).
12 The impact of increasing nitrogen inputs (e.g., NOX, nitrates, nitric acid) on the nitrogen
13 cycle and forests, wetlands, and aquatic ecosystems is discussed in detail elsewhere (U.S.
14 Environmental Protection Agency, 1993,1997a; Garner, 1994; World Health Organization,
15 1997). The most important effects of nitrogen deposition are accumulation of nitrogen
16 compounds resulting in the enhanced availability of nitrate or ammonium, soil-mediated effects
17 of acidification, and increased susceptibility to stress factors (Bobbink et al., 1998). A major
18 concern is "nitrogen saturation", the result of the deposition of large amounts of particulate
19 nitrates. Nitrogen saturation results when additions to soil background nitrogen (nitrogen
20 loading) exceed the capacity of plants and soil microorganisms to utilize and retain nitrogen
21 (Aber et al., 1989,1998; Garner, 1994; U.S. Environmental Protection Agency, 1993). Under
22 these circumstances, ecosystems become unable to utilize excessive nitrogen inputs and
23 disruptions of ecosystem functioning may result (Hornung and Langan, 1999).
24 Growth of most forests in North America is limited by the nitrogen supply. Severe
25 symptoms of nitrogen saturation, however, have been observed in high-elevation, nonaggrading
26 spruce-fir ecosystems in the Appalachian Mountains, as well as in the eastern hardwood
27 watersheds at Fernow Experimental Forest near Parsons, WV. Mixed conifer forests and
28 chaparral watersheds with high smog exposure in the Los Angeles Air Basin also are nitrogen
29 saturated and exhibit the highest stream water NO3" concentrations for wildlands in North
30 America (Bytnerowicz and Fenn, 1996; Fenn et al., 1998). Not all forest ecosystems react in the
31 same manner to nitrogen deposition. High-elevation alpine watersheds in the Colorado Front
March 2001
4-42
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Range and a deciduous forest in Ontario, Canada, also are naturally saturated even though
nitrogen deposition has been moderate (=8 kg.ha ha'1 .year1). The Harvard Forest hardwood
stand in Massachusetts, however, has absorbed >900 kg N/ha without significant NO3' leaching
during a nitrogen amendment study of 8 years (Table 4-5; Fenn et al.,1998). Johnson et al.
(199 la) reported that measurements showing the leaching of nitrates and aluminum (A13+) from
high elevation forests in the Great Smoky Mountains indicates that these forests have reached
saturation.
Possible ecosystem responses to nitrate saturation, as postulated by Aber and coworkers
(Aber et al., 1989), include a permanent increase in foliar nitrogen and reduced foliar phosphorus
and lignin caused by the lower availability of carbon, phosphorus, and water; reduced
productivity in conifer stands because of disruptions of physiological function; decreased root
biomass and increased nitrification and nitrate leaching; and (4) reduced soil fertility, resulting
from increased cation leaching, increased nitrate and aluminum concentrations in streams, and
decreased water quality. Saturation implies that some resource other than nitrogen is limiting
biotic function.
Water and phosphorus for plants and carbon for microorganisms are the resources most
likely to be the secondary limiting factors. The appearance of nitrogen in soil solution is an early
symptom of excess nitrogen. In the final stage, disruption of forest structure becomes visible
(Garner, 1994).
Changes in nitrogen supply can have a considerable impact on an ecosystem's nutrient
balance (Waring, 1987). Large chronic additions of nitrogen influence normal nutrient cycling
and alter many plant and soil processes involved in nitrogen cycling (Aber et al., 1989).
Among the processes affected are (1) plant uptake and allocation, (2) litter production,
(3) immobilization (includes ammonification [the release of ammonia] and nitrificatrion
[conversion of ammonia to nitrate during decay of litter and soil organic matter]), and (4) nitrate
leaching and trace gas emissions (Figure 4-2; Aber et al., 1989).
Subsequent studies have shown that, although initially, there was an increase in nitrogen
mineralization (i.e., the conversion of soil organic matter to nitrogen in available form [see item
3 above]), nitrogen mineralization rates were reduced under nitrogen-enriched conditions. Also,
studies suggest that, during saturation, soil microbial communities change from predominantly
fungal (mycorrhizal) communities to those dominated by bacteria (Aber et al., 1998).
March 2001
4-43
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-5. NITROGEN-SATURATED FORESTS IN NORTH AMERICA, INCLUDING
ESTIMATED N INPUTS AND OUTPUTS
Location
Forest Type
Elevation N Input N Output
(m) (kg-ha"' -year') (kg-ha"' -year'1)
Reference
Adirondack Mts.
northeastern New York
Catskill Mts.,
southeastern New York
Turkey Lakes Watershed,
Ontario, Canada
Whitetop Mt.,
southwestern Virginia
Northern hardwoods or '
hardwood/conifer mix
Mainly hardwood;
some eastern hemlock
Sugar maple and yellow
birch
Red spruce
396-661
,335-675
350-400
1650
9.3",
10,2s
7.0-7.7 (as
throughfall)
32C
Stage 1 N
loss"
Stage 1 and 2
N loss"
17.9-23.6
47C
Driscoll and
Van Dreason (1993)
Stoddard (1994)
Foster etal. (1989)
Johnson and
Lindberg(1992a)
Joslin and Wolfe
(1992)
Joslin etal. (1992)
Fernow, West Virginia
Great Smoky Mts.
National Park, Tennessee
Great Smoky Mts.
National Park, Becking
Site, North Carolina
Great Smokey Mts.
National Park, Tower
Site, North Carolina
Front Range, Colorado
San'Dimas, San Gabriel
Mts. southern California
Camp Paivika,
San Bemadino Mts.,
southern California
Klamath Mts,
northern California
Thompson Forest, Cascade
Mts Washington
Mixed hardwood
American beech
Red spruce
Red spruce
Alpine tundra,
subalpine conifer
Chapparral and
grasslands
Mixed conifer
Western coniferous
Red alder
735-870 15-20
1600
1800
1740
3.1",
10.3d
26.6
3000-4000 7.5-8.0
580-1680 23.3"
1600 30
NA
6.1
2.9
19.2
20.3
7.5
0.04-19.4
7-26f
NA8
Mainly
geologic8
220 4.7 plus >, 100 38.9
as N, fixation
Gilliametal. (1996)
Peterjohn etal. (1996)
Johnson and Lindberg
(1992b)
Johnson etal. (199 la)
Johnson etal. (1991a)
Williams etal. (1996)
Riggan etal. (1985)
Fenn etal. (1996)
' ! • •
Dahlgren (1994)
Johnson and Lindberg
(1992b) '
"Estimated total N deposition from wet deposition data is from Driscoll et al. (1991) for the Adirondacks, and from Stoddard and
Murdoch (1991) for the Catskills. Total deposition was estimated based on the wet deposition/total N deposition ratio (0.56) at
Huntington Forest in the Adirondacks (Johnson and Lindberg, 1992b). Nitrogen deposition can be higher in some areas, especially
at high-elevation sites such as Whiteface Mountain (15.9 kg-ha-'-year'1; Johnson and Lindberg, 1992b).
'Stage 1 and 2 of N loss according to the watershed conceptual model of Stoddard (1994). Nitrogen discharge (kg-ha"'-year') data
are not available; only stream water NO3" concentration trend data were collected. ,
^Values appear high compared to other sites, especially N leaching losses. Joslin and Wolfe (1992) concede that "there is
considerable uncertainty associated with the estimates of atmospheric deposition and leaching fluxes." However, elevated NO3"
concentrations in soil solution, and lack of a growth response to N fertilization (Joslin and Wolfe, 1994) support the hypothesis that
the forest at Whitetop Mountain is N saturated.
•"Estimated total N deposition from throughfall data. Total deposition was estimated based on the throughfall/total N deposition
ration (0.56) from the nearby Smokies Tower site (Johnson and Lindberg, 1992b).
'Annual throughfall deposition to the chaparral ecosystem.
'Nitrogen output is from unpublished streamwater data (M.E. Fenn and M.A. Poth). The low value represents a year of average
precipitation, and the high value is for 1995, when precipitation was nearly double the long-term average. Nitrogen output
includes N export in streamwater and to groundwater.
'Annual input and output data are not known, although N deposition in this forest'is probably typical for much of the rural western
United States (2-3 kg N-ha^-year'1 (Young et al., 1988). Excess N is from weathering of ammonium in mica schist bedrock. The
ammonium was rapidly nitrified, leading to high NO3- concentrations in soil solution (Dahlgren, 1994).
March 2001
4-44
DRAFT-DO NOT QUOTE OR CITE
-------
Nitrogen
Oxides in
Atmosphere
Gaseous
Nitrogen in
Atmosphere
Photosynthesis
Plant
Utilization
Animal
Proteins
Trace
Gas
Emissions
Litter
Production
(Death)
Nitrogen
Fixation
Microbial
Decomposition
Nitrates
in Soil
Microbial
Utilization
Urea
(Ammonia)
Nitrates in
Streams,
Rivers, Lakes
and Oceans
Nitrites
by
Bacteria
—B Process altered by
nitrogen saturation
Figure 4-2. Nitrogen cycle (dotted lines indicate processes altered by nitrogen satuation).
Source: Garner (1994).
1
2
3
The availability of nutrients is an important factor in determining species composition, and
nitrogen is usually the growth-limiting nutrient (Bobbink, 1998). Most of the plants growing iti
nutrient-poor habitats have become adapted to them over time and can only compete successfully
March 2001
4-45
DRAFT-DO NOT QUOTE OR CITE
-------
1 on soils low in nitrogen (Bobbink, 1998; Chapin, 1991). All plants growing in low resource
2 environments (e.g., infertile soil, shaded understory, deserts, tundra) have been observed to have
3 certain similar characteristics: a slow growth rate, low photosynthetic rate, and low capacity for
4 nutrient uptake. An important feature to plants adapted to low-resource environments is that they
5 grow slowly and tend to respond less, even when provided with an optimal supply and balance of
6 resources (Pearcy et al., 1987; Chapin, 1991). Plants adapted to cold, moist environments grow
7 more leaves than roots as the relative availability to nitrogen increases; however, other nutrients
8 may soon be limiting. The capacity of gymnosperms in general, and in subalpine and boreal
9 species in particular, to reduce nitrates in either roots or leaves appears to be limited. In addition,
10 the ability of trees to use nitrogen varies with the age of the tree and the density of the stand
11 (Waring, 1987).
12 Because the competitive equilibrium of plants in any community is finely balanced, the
13 alteration of one of a number of environmental parameters, (e.g., continued nitrogen additions),
14 can change the vegetation structure of an ecosystem (Bobbink, 1998; Skeffington and Wilson,
15 1988). Increases in soil nitrogen play a selective role. When nitrogen becomes more readily
16 available, plants adapted to living in an environment of low nitrogen availability will be replaced
17 by plants capable of using increased nitrogen because they have a competitive advantage.
1 g The long-term impacts of increased nitrogen deposition have been studied in several
19 western and central European plant communities: lowland heaths, species-rich grasslands,
20 mesotrophic fens, ombrotrophic bogs, upland moors, forest-floor vegetation, and freshwater
21 lakes (Bobbink, 1998). Large changes in species composition have been observed in regions
22 with high nitrogen loadings or infield experiments after years of nitrogen addition (Bobbink
23 et al., 1998). The increased input of nitrogen gradually increased availability of nitrogen in the
24 soil, and its retention because of low rates of leaching and denitrification resulted in faster litter
25 decomposition and rate of mineralization. Faster growth and greater height of nitrophilic species
26 enables these plants to shade out the slower growing species, particularly those in oligotrophic or
27 mesotrophic conditions (Bobbink, 1998; Bobbink et al., 1998). Excess nitrogen inputs to
28 unmanaged heathlands in the Netherlands has resulted in nitrophilous grass species replacing
29 slower growing heath species (Roelofs et al., 1987; Garner, 1994). Van Breemen and Van Dijk
30 (1988) noted that over the past several decades the composition of plants in the forest herb layers
March 2001
4-46
DRAFT-DO NOT QUOTE OR CITE
-------
1 has been shifting toward species commonly found on nitrogen-rich areas. It also was observed
2 that the fruiting bodies of mycorrhizal fungi had decreased in number.
3 Other studies in Europe point out the effects of excessive nitrogen deposition on mixed-oak
4 forest vegetation along a deposition gradient largely controlled by soil acidity, nitrogen supply,
5 canopy composition, and location of sample plots (Brunet et al., 1998; Falkengren-Grerup,
6 1998). Results of the study, using multivariate methods, suggest that nitrogen deposition has
7 affected the field-layer vegetation directly by increased nitrogen availability and, indirectly, by
8 accelerating soil acidity. Time series studies indicate that 20 of the 30 field-layer species
9 (nonwoody plants) that were associated most closely with high nitrogen deposition increased in
10 frequency in areas with high nitrogen deposition during the past decades. Included in the field-
11 layer species were many generally considered nitrophilous; however, there were several acid
12 tolerant species (Brunet et al, 1998). Falkengren-Grerup (1998), in an experimental study
13 involving 15 herbs and 13 grasses, observed that species with a high nitrogen demand and a
14 lesser demand for other nutrients were particularly competitive in areas with acidic soils and high
15 nitrogen deposition. The grasses grew better than herbs with the addition of nitrogen. It was
16 concluded that, at the highest nitrogen deposition, growth was limited for most species by the
17 supply of other nutrients; and, at the intermediate nitrogen concentration, the grasses were more
18 efficient than the herbs in utilizing nitrogen. Nihlgard (1985) suggested that excessive nitrogen
19 deposition may contribute to forest decline in other specific regions of Europe. Also, Schulze
20 (1989), Heinsdorf (1993), and Lamersdorf and Meyer (1993) attribute magnesium deficiencies in
21 German forests, in part, to excessive nitrogen deposition.
22 Plant succession patterns and biodiversity are affected significantly by chronic nitrogen
23 additions in some North American ecosystems (Figure 4-3). Fenn et al. (1998) report that
24 long-term nitrogen fertilization studies in both New England and Europe, as well, suggest that
25 some forests receiving chronic inputs of nitrogen may decline in productivity and experience
26 greater mortality. Long-term fertilization experiments at Mount Ascutney, Vermont, suggest that
27 declining coniferous forest stands with slow nitrogen cycling may be replaced by deciduous
28 fast-growing forests that cycle nitrogen rapidly (Fenn et al., 1998).
29 In experimental studies of nitrogen deposition conducted by Wedin and Tilman (1996) over
30 a 12-year period on Minnesota grasslands, plots dominated by native warm-season grasses
31 shifted to low-diversity mixtures dominated by cool-season grasses at all but the lowest rates of
March 2001
4-47
DRAFT-DO NOT QUOTE OR CITE
-------
N-Saturated Ecosystems
in North America
Review of Ecosystem Effects
and Responses to Excess N
1. Nitrogen Inputs: •
> Atmospheric deposition, N2 fixation, fertilization
Nitrogen Retention:
>• In plant biomass and soil organic matter
> The role of soil microbes and woody residues
> Abiotic retention
Nitrogen Outputs:
>-Hydrologic transport, gaseous emissions from soil
>Removal in harvest, fire emissions, and soil erosion
' 2. Characteristics Predisposing Forests
to N Saturation:
>• Stand vigor and succession, forest type
>• Previous land use-stand history
••Soil N accumulation
> Topography and climate
> Nitrogen deposition
3. Ecosystem Responses to Excess Nitrogen:
>• Nitrate leaching and export
> Eutrophicationof estuaries
>• Toxicity of surface waters
> Foliar nutrient responses
>• Nitrogen mineralization and nitrification
>• Effects on soil organic matter
>• Soil acidification, cation depletion, Al toxicity
>• Foliar nutrient responses
> Greenhouse gas fluxes
4. Regional N Saturation Conceptual Models:
>• New England forests
>• California forests
> Colorado alpine ecosystems
Figure 4-3. Diagrammatic overview of excess nitrogen (N) in North America.
Source: Fennelal.(1998).
March 2001
4-48
DRAFT-DO NOT QUOTE OR CITE
-------
1 nitrogen addition. Grasslands with high nitrogen retention and carbon storage rates were the
2 most vulnerable to loss of species and major shifts in nitrogen cycling. The shift to low-diversity
3 mixtures was associated with the decrease in biomass carbon to nitrogen (C:N) ratios, increased
4 nitrogen mineralization, increased soil nitrate, high nitrogen losses, and low carbon storage
5 (Wedin and Tilman, 1996). Naeem et al. (1994) experimentally demonstrated (under controlled
6 environmental conditions) that loss of biodiversity, in addition to loss of genetic resources, loss
7 of productivity, loss of ecosystem buffering against ecological perturbation, and loss of aesthetic
8 and commercially valuable resources, also may alter or impair ecosystems services.
9 The C:N ratio of the forest floor can be changed by nitrogen deposition over time. This
10 change appears to occur when the ecosystem becomes nitration saturated (Gundersen et al.,
11 1998a). Long-term changes in C:N status have been documented in Central Europe and indicate
12 that nitrogen deposition has changed the forest floor. In Europe, low C:N ratios coincide with
13 high deposition regions (Gundersen et al,, 1998a). A strong decrease in forest floor root biomass
14 has been observed with increased nitrogen availability. Roots and the associated rnycorrhizae
15 appear to be an important factor in the accumulation of organic matter in the forest floor at
16 nitrogen limited sites. If root growth and mycorrhizal formation are impaired by nitrogen
17 deposition, the stability of the forest floor may be affected by stimulating turnover and decreasing
18 the root litter input to the forest floor and thus decrease the nitrogen that can be stored in the
19 forest floor pool (Gundersen et al., 1998b). Nitrogen-limited forests have a high capacity for
20 deposited nitrogen to be retained by the plants and microorganisms competing for available
21 nitrogen (Gundersen et al., 1998b). Nitrate leaching has been correlated significantly with nitrate
22 status but not with nitrate depositions. Forest floor C:N ratio has been used as a rough indicator
23 of ecosystem nitrogen status in mature coniferous forests and the risk of nitrate leaching;
24 analyses of European databases indicated an empirical relationship between forest floor C:N ratio
25 and nitrate leaching (Gundersen et al., 1998a). Nitrate leaching was observed when the
26 deposition received was more than 10kg N/ha. All of the data sets supported a threshold at
27 which nitrate leaching seems to increase at a C:N ratio of 25. Therefore, to predict the rate of
28 changes in nitrate leaching it is necessary to be able to predict the rate of changes in the forest
29 floor C:N ratio. Understanding the variability in forest ecosystem response to nitrogen input is
30 essential in assessing pollution risks (Gundersen et al., 1998a).
March 2001
4-49
DRAFT-DO NOT QUOTE OR CITE
-------
1 The plant root is an important region of nutrient dynamics. The rhizosphere includes the
2 soil that surrounds and is influenced by plant roots (Wall and Moore, 1999). The mutualistic
3 relationship between plant roots, fungi, and microbes is critical for the growth of the organisms
4 involved. The plant provides shelter and carbon, whereas the symbiont provides access to a
5 limiting nutrient such as nitrogen and phosphorus. As indicated above, changes in soil nitrogen
6 influence the mycorrhizal-plant relationship. Mycorrhizal fungal diversity is associated with
7 above-ground plant biodiversity, ecosystem variability, and productivity (Wall and Moore, 1999).
8 Aber et al. (1998) showed a close relationship between mycorrhizal fungi and the conversion of
9 dissolved inorganic nitrogen to soil nitrogen. During nitrogen saturation, soil microbial
10 communities change from being fungal, and probably being dominated by mycorrhizae, to being
11 dominated by bacteria. The loss of mycorrhizal function has been hypothesized as the key
12 process leading to increased nitrification and nitrate mobility. Increased nitrate mobility leads to
13 increased cation leaching and soil acidification (Aber et al., 1998).
14 The interrelationship of above- and below-ground flora is illustrated by the natural invasion
15 of heath lands by oaks (Quercus robur). Soils are dynamic entities, the features of which can
16 change like the rest of the ecosystem with age and management. The soil-forming factors under
17 the heath have been vegetation type during the last 2000 years, whereas the invasion by oaks has
18 been taking place for only a few decades. Clear changes in the ground floor and soil morphology
19 takes place when trees colonize heath (Nielsen et al., 1999). The distribution of roots also
20 changed under the three different vegetation types. Under both heather and the Sitka spruce
21 plantation, the majority of roots are confined to the uppermost horizons, whereas under oak, the
22 roots are distributed more homogeneously. There was also a change in the C:N ratio when
23 heather was replaced by oaks. Also, the spontaneous succession of the heath by oaks changed
24 the biological nutrient cycle into a deeper vertical cycle, when compared to the heath where the
25 cycle is confined to the upper soil horizons. Soils similar to those described in this study
26 (Jutland, Denmark), with mainly an organic buffer system, seem to respond quickly to changes in
27 vegetation (Nielsen et al., 1999).
28 In addition to excess nitrogen deposition effects on terrestrial ecosystems of the types noted
29 above (e.g., dominant species shifts and other biodiversity impacts), direct atmospheric nitrogen
30 deposition and increased nitrogen inputs via runoff into streams, rivers, lakes, and oceans can
31 have notable impacts on aquatic ecosystems as well. One illustrative example is recently
March 2001
4-50
DRAFT-DO NOT QUOTE OR CITE
-------
1 reported research (summarized by Paerl et al., in press) characterizing impacts of nitrogen
2 deposition on the Pamlico Sound, NC, estuarine complex, which serves as a key fisheries nursery
3 supporting an estimated 80% of commercial and recreational finfish and shellfish catches in the
4 southeastern U.S. Atlantic coastal region. Such direct atmospheric nitrogen deposition onto
5 waterways feeding into the Pamlico Sound or onto the sound itself and indirect nitrogen inputs
6 via runoff from upstream watersheds contribute to conditions of severe water oxygen depletion,
7 formation of algae blooms in portions of the Pamlico Sound estuarine complex, and altered fish
8 distributions, catches, and physiological states and incidence of disease. Under extreme
9 conditions of especially high rainfall rate events (e.g., hurricanes) affecting watershed areas
10 feeding into the sound, the effects of nitrogen runoff (in combination with excess loadings of
11 metals or other nutrients) can be massive (e.g., creation of the widespread "dead-zone" affecting
12 large areas of the Pamlico Sound for many months after hurricane Fran in 1996 and hurricanes
13 Dennis, Floyd, and Irene in 1999 impacted eastern North Carolina).
14
15 Sulfur Deposition Effects
16 Sulfur is an essential plant nutrient and, as such, is a major component of plant proteins.
17 The most important source of sulfur is sulfate taken up from the soil by plant roots even though
18 plants can utilize atmospheric SO2 (Marschner, 1995). The availability of organically bound
19 sulfur in soils depends largely on microbial decomposition, a relatively slow process. The major
20 factor controlling the movement of sulfur from the soil into vegetation is the rate of release from
21 the organic to the inorganic compartment (May et al., 1972; U. S. Environmental Protection
22 Agency, 1982; Marschner, 1995). Sulfur plays a critical role in agriculture as an essential
23 component of the balanced fertilizers needed to grow and increase worldwide food production
24 (Ceccotti and Messick, 1997). Atmospheric deposition is an important component of the sulfur
25 cycle. This is true not only in polluted areas where atmospheric deposition is very high, but also
26 in areas of low sulfur input. Additions of sulfur into the soil in the form of SO4 2" could alter the
27 important organic-sulfur/organic-nitrogen relationship involved in protein formation in plants.
28 The biochemical relationship between sulfur and nitrogen in plant proteins indicates that neither
29 element can be assessed adequately without reference to the other. There is a regulatory coupling
30 of sulfur and nitrogen metabolism. Sulfur deficiency reduces nitrate reductase and, to a similar
31 extent, also glutamine synthetase activity. Nitrogen uptake in forests, therefore, may be loosely
March 2001
4-51
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
regulated by sulfur availability, but sulfate additions in excess of needs do not necessarily lead to
injury (Turner and Lambert, 1980; Hogan et al., 1998).
Only two decades ago, there was little information comparing sulfur cycling in forests with
nutrients, especially nitrogen. With the discovery of deficiencies in some unpolluted regions
(Kelly and Lambert, 1972; Humphreys et al., 1975; Turner et al., 1977; Schnug, 1997) and
excesses associated with acidic deposition in other regions of the world (Meiwes and Khanna,
1981; Shriner and Henderson, 1978; Johnson et al., 1982a,b), interest in sulfur nutrition and
cycling in forests has heightened. General reviews of sulfur cycling in forests have been written
by Turner and Lambert (1980), Johnson (1984), Mitchell et al. (1992a,b), and Hogan et al.
(1998). The salient elements of the sulfur cycle as it may be affected by changing atmospheric
deposition are summarized by Johnson and Mitchell (1988). Sulfur has become the most
important limiting factor in European agriculture because of the desulfurization of industrial
emissions (Schnug, 1997).
Most studies dealing with the impacts of sulfur deposition on plant communities have been
conducted in the vicinity of point sources and have investigated above-ground effects of SO2 or
acidifying effects of sulfate on soils (Krupa and Legge, 1998; Dreisinger and McGovern, 1970;
Legge, 1980; Winner and Bewley, 1998a,b; Laurenroth and Michunas, 1985; U.S. Environmental
Protection Agency, 1982). Krupa and Legge (1986), however, observed a pronounced increase
in foliar sulfur concentrations in all age classes of needles of the hybrid pine lodgepole x jack
pine (Finns contorta x P. banksiana). This vegetation had been exposed to chronic low
concentrations of sulfur gas pollution (SO2), hydrogen sulfide (H2S), and fugitive sulfur aerosol
for more than 20 years. Observations under the microscope showed no sulfur deposits on the
needle surfaces and led to the conclusion that the sulfur was derived from the soil. The oxidation
of elemental sulfur and the generation of protons is well known for the soils of Alberta, CN.
This process is mediated by bacteria of the Thiobacillus sp. As elemental sulfur gradually is
converted to protonated SO4, it can be leached downward and readily taken up by plant roots.
The activity of Thiobacillus sp. is stimulated by elemental sulfur additions (Krupa and Legge,
1986).
March 2001
4-52
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Effects of Acidic Deposition on Forest Soils
Substantial and previously unsuspected changes in soils are occurring both in polluted areas
of eastern North America, the United Kingdom, Sweden, and Central Europe and in less polluted
regions of Australia and western North America (reviewed by Johnson et al., 1991b; see review
by Huntington, 2000). In some cases, trends are toward more acidic soils (e.g., Markewitz et al.,
1998), and, in others, there are no consistent trends, with some soils showing increases and some
showing decreases at different sampling times, and some showing no change (e.g., Johnson and
Todd, 1998; Trettin et al., 1999; Yanai et al., 1999).
Significant changes have occurred at many sites in the eastern United States during recent
decades. Temporal trends in tree ring chemistry were examined as indicators of historical
changes in the chemical environmental of red spruce. Chemical changes in tree-ring chemistry
reflect changing inputs of regional pollutants to forests. If significant base cation mobilization
and depletion of base cations from eastern forest soils has occurred, a temporal sequence of
changes in uptake patterns and possibly in tree growth would be expected. Patterns of tree-ring
chemistry principally at high-elevation sites in the eastern United States, leads to the conclusion
that significant changes in soil chemistry have occurred in many of these sites during recent
decades leading to changes in growth (Bondietti and McLaughlin, 1992). These changes are
spatially and temporally consistent with emissions of SO2 and NO2 across the region, suggesting
that increased acidification of forest soils has occurred.
Increases in levels of Al and Fe typically occur as base cations are removed from soils by
tree uptake. A region-wide Ca increase above expected levels followed by a decrease suggests
that increased mobilization began perhaps 30 to 40 years ago (Bondietti and McLaughlin, 1992).
The period of Ca mobilization coincides with a region-wide increase in growth rate of red spruce,
whereas the period of decreasing levels of Ca in wood corresponds temporally with patterns of
decreasing radial growth at high elevation sites throughout the region during the past 20 to 30
years. The decline in wood Ca suggests that Ca loss may have been increased to the point at
which base saturation of soils has, been reduced (Bondietti and McLaughlin, 1992).
Studies by Shortle and Bondietti, (1992) support the view that changes in soil chemistry in
eastern North America forest sites occurred many decades ago "before anybody was looking".
Sulfur and nitrogen emissions began increasing in eastern North America in the 1920s and
continued to increase into the 1980s, when sulfur began to decrease but nitrogen emissions have
March 2001
4-53
DRAFT-DO NOT QUOTE OR CITE
-------
1 not (Garner et al., 1989). Shortle and Bondietti (1992) present evidence that, from the late 1940s
2 into the 1960s, the mor humus layer of acid-sensitive forest sites in eastern North America
3 underwent a significant change that resulted in the loss of exchangeable essential base cations
4 and interrupted the critical base nutrient cycles between mature trees and the root-humus
5 complex. The timing of the impact appears to have coincided with the period when the SOX and
6 NOX emissions in eastern North America subject to long-range transport were increasing the most
7 rapidly (See above; Shortle and Bondietti, 1992). Although forest ecosystems other than the
8 high-elevation spruce-fir forests are not currently manifesting symptoms of injury directly
9 attributable to acid deposition, less sensitive forests throughout the United States are
10 experiencing gradual losses of base cation nutrient, which in many cases will reduce the quality
11 of forest nutrition over the long term (National Science and Technology Council, 1998). In some
12 cases, it may not even take decades, because these forests already have been receiving sulfur and
13 nitrogen deposition for many years. The current status of forest ecosystems in different U.S.
14 geographic regions varies, as does their sensitivity to nitrogen and sulfur deposition. Variation in
15 potential future forest responses or sensitivity are caused, in part, by differences in deposition of
16 sulfur and nitrogen, ecosystem sensitivities to sulfur and nitrogen additions, and responses of
17 soils to sulfur and nitrogen inputs (National Science and Technology Council, 1998).
lg Acidic deposition has played a major role in recent soil acidification in some areas of
19 Europe and, to a more limited extent, eastern North America. Examples include the study by
20 Hauhs (1989) at Lange Bramke, Germany, which indicated that leaching was of major
21 importance in causing substantial reduction in soil-exchangeable base cations over a 10-year
22 period (1974-1984). Soil acidification and its effects result from the deposition of nitrate (NO3')
23 and sulfate (SO4 *') and the associated hydrogen (H +) ion. The effects of excessive nitrogen
24 deposition on soil acidification and nutrient imbalances have been well established in Dutch
25 forests (Van Breemen et al., 1982; Roelofs et al., 1985; Van Dijk and Roelofs, 1988).
26 For example, Roelofs et al. (1987) proposed that NH3 /NH4+ deposition leads to heathland
27 changes via two modes: (1) acidification of the soil and the loss of cations K+, Ca2+, and Mg2+;
28 and (2) nitrogen enrichment that results in "abnormal" plant growth rates and altered competitive
29 relationships. Nihlgard (1985) suggested that excessive nitrogen deposition may contribute to
30 forest decline in other specific regions of Europe. Falkengren-Grerup (1987) noted that, during
March 2001
4-54
DRAFT-DO NOT QUOTE OR CITE
-------
1 about 50 years, unexpectedly large increases in growth of beech (Fagus sylvatica Z.) were
2 associated with decreases in pH and exchangeable cations in some sites in southernmost Sweden.
3 Likens et al. (1996) suggested that soils are changing at the Hubbard Brook Watershed,
4 NH, because of a combination of acidic deposition and reduced base cation deposition. They
5 surmised, based on long-term trends in stream-water data, that large amounts of Ca and Mg have
6 been lost from the soil-exchange complex over a 30-year period from approximately 1960 to
7 1990. The authors speculate that the declines in base cations in soils may be the cause of recent
8 slowdowns in forest growth at Hubbard Brook. In a follow-up study, however, Yanai et al.
9 (1999) found no significant decline in Ca and Mg concentrations in forest floors at Hubbard
10 Brook over the period 1976 to 1997. They also found both gains and losses in forest floor Ca
11 and Mg between 1980 and 1990 in a regional survey. Thus, they concluded that "Forest floors in
12 the region are not currently experiencing rapid losses of base cations, although losses may have
13 preceded the onset of these three studies."
14 Hydrogen ions entering a forest ecosystem first encounter the forest canopy, where they are
15 often exchanged for base cations that then appear in throughfall (Figure 4-4 depicts a model of
16 H+ sources and sinks). Base cations leached from the foliage must be replaced through uptake
17 from the soil, or foliage cations will be reduced by the amounts leached. In the former case, the
18 acidification effect is transferred to the soil, where H + is exchanged for a base cation at the
19 root-soil interface. Uptake of base cations or NH4 + by vegetation or soil microorganisms causes
20 the release of H + in order to maintain charge balance. Uptake of nutrients in anionic form (NO3',
21 SO4 2", PO4 3") causes the release of OH" in order to maintain charge balance. Thus, the net
22 acidifying effect of uptake is the difference between cation and anion uptake. The form of ions
23 taken up is known for all nutrients but nitrogen, where either NH4+ or NO3" can be taken up.
24 In that, nitrogen is a nutrient taken up in great quantities, the uncertainty in the ionic form of
25 nitrogen taken up creates great uncertainty in the overall H+ budget for soils (Johnson 1992).
26 The cycles of base cations differ from those of N, P, and S in several respects. The fact that
27 Ca, K, and Mg exist primarily as cations in solution whereas N, P, and S exist primarily as anions
28 has major implications for the cycling of the nutrients and the effects of acid deposition on these
29 cycles. The most commonly accepted model of base cation cycling in soils is one in which base
30 cations are released by weathering of primary minerals to cation exchange sites, where they are
31 then available for either plant uptake or leaching (Figure 4-4). The introduction of H + by
March 2001
4-55
DRAFT-DO NOT QUOTE OR CITE
-------
Deposition
so?
Soil
Organism
Uptake
2H++NO3
Nitrification
CO2 + H2O
Carbonic Acid Formation
R-COOH
Organic Acid Formation
2OH
Soil
Organism
Uptake
20H
Leaching
Figure 4-4. Schematic of sources and sinks of hydrogen ions in a forest (from Taylor et al.,
1994).
1 atmospheric deposition or by internal processes will impact directly the fluxes of Ca, K, and
2 Mg via cation exchange or weathering processes. Therefore, soil leaching is often of major
3 importance in cation cycles, and many forest ecosystems show a net loss of base cations
4 (Johnson, 1992a).
5 Two basic types of soil change are involved: (1) a short-term intensity type change
6 resulting from the concentrations of chemicals in soil water, and (2) a long-term capacity change
March 2001
4-56
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
based on the total content of bases, aluminum and iron stored in the soil (Reuss and Johnson,
1986; Van Breemen, 1983). Changes in intensity factors can have a rapid impact on the
chemistry of soil solutions. Increases in the amounts of sulfur and nitrogen in acidic deposition
can cause immediate increases in acidity and mobilization of aluminum in soil solutions.
Increased aluminum concentrations and an increase in the Ca/Al ratio in soil solution have been
linked to a significant reduction in the availability of essential base cations to plants, an increase
in plant respiration, and increased biochemical stress (National Science and Technology Council,
1998).
Rapid changes in intensity, resulting from the addition of increased amounts of nitrogen or
sulfur in acidic deposition, can have a rapid impact on the chemistry of soil solutions by
increasing the acidity and mobilizing aluminum. Increased concentrations of aluminum and an
increase in the ratio of calcium-to-aluminum in soil solution have been linked to significantly
reduced availability of essential cations to plants.
Capacity changes are the result of many factors acting over long time periods. The content
of base cations (calcium, magnesium, sodium, and potassium) in soils result from additions from
the atmospheric deposition, decomposition of vegetation and geologic weathering. Loss of base
cations may occur through plant uptake and leaching. Increased leaching of base cations may
result in nutrient deficiencies in soils as has been happening in some sensitive forest ecosystems
(National Science and Technology Council, 1998).
A major concern has been that soil acidity would lead to nutrient deficiency. Calcium is
essential in the formation of wood and the maintenance of cells, the primary plant tissues
necessary for tree growth. Trees obtain Ca from the soil, but to be taken up by roots, the Ca
(a positively charged ion) must be dissolved in soil water (Lawrence and Huntington, 1999).
Tree species may be adversely affected if high Al to nutrient ratios limit uptake of Ca and Mg
and create a nutrient deficiency (Shortle and Smith, 1988; Garner, 1994). Acid deposition by
lowering the pH of aluminum-rich soil can increase aluminum concentrations in soil water
through dissolution and ion-exchange processes. When in solution, aluminum can be taken up
by roots, transported through the tree and, eventually, deposited on the forest floor in leaves and
branches. Aluminum is more readily taken up than is Ca because it has a higher affinity for
negatively charged surfaces than does Ca. When present in the forest floor, Al tends to displace
adsorbed Ca and causes it to be more readily leached. The continued buildup of Al in the forest
March 2001
4-57
DRAFT-DO NOT QUOTE OR CITE
-------
1 floor layer, where nutrient uptake is greatest, can decrease the availability of Ca to the roots
2 (Lawrence et al., 1995) and lower the efficiency of Ca uptake because Al is more readily taken up
3 than is Ca 2* when the ratio of Ca to Al in soil water is less than one (Lawrence and Huntington,
4 1999). A 1968 Swedish report to the United Nations postulated a decrease in forest growth of ca.
5 1.5%/year as result of Ca 2+ loss by leaching (Johnson and Taylor, 1989). The concern that soil
6 acidification and nutrient deficiency may result in forest decline remains extant today.
7 Aluminum toxicity is a possibility in acidified soils. Atmospheric deposition (or any other
8 source of mineral anions) can increase the concentration of Al, especially A13+, in soil solution
9 without causing significant soil acidification (Johnson and Taylor, 1989). Aluminum can be
10 brought into soil solution in two ways: (1) by acidification of the soil and (2) by an increase in
11 the total anion and cation concentration of the soil solution. The introduction of mobile, mineral
12 acid anions to an acid soil will cause increases in the concentration of aluminum in the soil
13 solution, but extremely acid soils in the absence of mineral acid anions will not produce a
14 solution high in aluminum. An excellent review of the relationships among the most widely used
15 cation-exchange equations and their implications for the mobilization of aluminum into soil
16 solution is provided by Reuss (1983).
17 . Aluminum toxicity may influence forest tree growth, where acid deposition and natural
18 acidifying processes increase soil acidity. Aluminum concentrations have been observed to
19 exhibit a strongly descending gradient from bulk soil through the rhizosphere to the root (Smith,
20 1990a). Once it enters the forest tree roots, Al accumulates in root tissue (Thornton et al., 1987;
21 Vogt et al., 1987a,b). There is abundant evidence that Al is toxic to plants. Reductions of Ca
22 uptake by roots has been associated with increases in Al uptake (Clarkson and Sanderson, 1971).
23 Calcium plays a major role in cell membrane integrity and cell wall structure. A number of
24 studies have suggested that the toxic effect of aluminum on forest trees could be caused by Ca2*
25 deficiency (Shortle and Smith, 1988; Smith, 1990a). Mature trees have a high Ca2+ requirement
26 relative to agricultural crops (Rennie, 1955). Shortle and Smith (1988) attributed the decline of
27 red spruce in eight stands across northern New England from Vermont to Maine to an imbalance
28 of Al3* and Ca2+ in the fine root environment. Aluminum in the soil solution reduces Ca uptake
29 by competing for binding sites in the cortex of fine roots. Reduction in Ca uptake suppresses
30 cambial growth and reduces the rate of wood formation (annual ring formation), decreases the
31 amount of functional sapwood and live crown, and predisposes trees to disease and injury from
March 2001
4-58
DRAFT-DO NOT QUOTE OR CITE
-------
1 stress agents when the functional sapwood becomes less than 25% of cross-sectional stem area
2 (Smith, 1990a).
3 Air pollution is not the sole cause of soil change. High rates of acidification are occurring
4 in less polluted regions of the western United States and Australia because of internal soil
5 processes, such as tree uptake of nitrate and nitrification associated with excessive nitrogen
6 fixation (Johnson et al., 1991b). Many studies have shown that acidic deposition is not a
7 necessary condition for the presence of extremely acid soils, as evidenced by their presence in
8 unpolluted, even pristine forests of the northwestern United States and Alaska (Johnson et al.,
9 1991b). The soil becomes acidic when H+ ions attached to NH4+ or HNO3 remain in the soil after
10 nitrogen is taken up by plants. For example, Johnson et al. (1982b) found significant reductions
11 in exchangeable K + over a period of only 14 years in a relatively unpolluted Douglas fir
12 Integrated Forest Study (IPS) site in the Washington Cascades. The effects of acid deposition at
13 this site were negligible relative to the effects of natural leaching (primarily carbonic acid) and
14 nitrogen tree uptake (Cole and Johnson, 1977). Even in polluted regions, numerous studies have
15 shown the importance of tree uptake of NH4+ and NCy in soil acidification. Binkley et al. (1989)
16 attributed the marked acidification (pH decline of 0.3 to 0.8 units and base saturation declines of
17 30 to 80%) of abandoned agricultural soil in South Carolina over a 20-year period to NH4+ and
18 NO3" uptake by a loblolly pine plantation.
19 An interesting example of uptake effects on soil acidification is that of Al uptake and
20 cycling (Johnson et al., 1991b). Aluminum accumulation in the leaves of coachwood
21 (Ceratopetalum apetalum) in Australia has been found to have a major impact on the distribution
22 and cycling of base cations (Turner and Kelly, 1981). The presence of C. apetalum as a
23 secondary tree layer beneath brush cox (Lophostemon confertus) was found to lead to increased
24 soil exchangeable Al3+ and decreased soil exchangeable Ca 2+ (Turner and Kelly, 1981). The
25 constant addition of aluminum-rich litter fall obviously has had a substantial effect on soil
26 acidification, even if base cation uptake is not involved directly.
27 Given the potential importance of particulate deposition for base cation status of forest
28 ecosystems, the findings of Driscollet al. (1989) and Hedin et al. (1994) are especially relevant.
29 Driscoll et al. (1989) noted a decline in both SO42' and base cations in both atmospheric
30 deposition and stream water over the past two decades at Hubbard Brook Watershed, NH. The
March 2001
4-59
DRAFT-DO NOT QUOTE OR CITE
-------
1 decline in SO42" deposition was attributed to a decline in emissions, and the decline in stream
2 water SO42" was attributed to the decline in sulfur deposition.
3 Hedin et al. (1994) reported a steep decline in atmospheric base cation concentrations in
4 both Europe and North America over the past 10 to 20 years. The reductions in SO 2 emissions
5 in Europe and North America in recent years have not been accompanied by equivalent declines
6 in net acidity related to sulfate in precipitation. These current declines in sulfur deposition have,
7 in varying degrees, been offset by declines in base cations and may be contributing "to the
8 increased sensitivity of poorly buffered systems." Analysis of the data from the Integrated Forest
9 Studies (IFS) supports the authors' contention that atmospheric base cation inputs may seriously
10 affect ecosystem processes. Johnson et al. (1994a) analyzed base cation cycles at the Whiteface
11 Mountain IFS site in detail and concluded that Ca losses from the forest floor were much greater
12 than historical losses, based on historical changes in forest floor Ca observed in an earlier study
13 (Johnson et al., 1994b). Further, the authors suggest that the difference between historical and
14 current net loss rates of forest floor Ca may be caused by sharply reduced atmospheric inputs of
15 calcium after about 1970 and exacerbated by sulfate leaching (U.S. Environmental Protection
16 Agency, 1999).
17 . The calcium/aluminum molar ratio has been suggested as a valuable ecological indicator of
18 an approximate threshold beyond which the risk of forest injury from Al stress and nutrient
19 imbalances increases (Cronan and Grigal, 1995). The Ca/Al ratio also can be used as an
20 indicator to assess forest ecosystem changes over time in response to acidic deposition, forest
21 harvesting, or other process that contribute to acid soil infertility. This ratio, however, may not
22 be a reliable indicator of stress in areas with both high atmospheric deposition of ammonium and
23 magnesium deficiency via antagonism involving ammonium rather than aluminum, and in areas
24 with soil solutions with calcium concentrations greater than 500 micromoles per liter (National
25 Science and Technology Council, 1998). Cronan and Grigal (1995) based on a review of the
26 literature have made the following estimates for determining the adverse impact of acidic
27 deposition on tree growth or nutrition:
28 • forests have a 50% risk of adverse impacts if the Ca/Al ration is 1.0,
29 • the risk is 75% if the ratio is 0.5, and
30 • the risk approaches 100% if the ratio is 0.2.
March 2001
4-60
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
The Ca/Al ratio of soil solution provides only an index of the potential for Al stress. Cronan and
Grigal (1995), state that the overall uncertainty of the Ca/Al ratio associated with a given
probability ratio is considered to be approximately ±50%. Determination of thresholds for
potential forest impacts requires the use of the four successive measurement endpoints in the soil,
soil solution, and plant tissue listed below.
(1) Soil base saturation less than 15% of effective cation exchange capacity
(2) Soil solution Ca/Al molar ratio less than 1.0 for 50% risk
(3) Fine roots tissue Ca/Al molar ratio less than 0.2 for 50% risk
(4) Foliar tissue Ca/Al molar ratio less than 12.5 for 50% risk
The application of the Ca/Al ratio indicator for assessment and monitoring of forest health risks
has been recommended for sites or in geographic regions where the soil base saturation <15%.
Critical Loads
In Europe, the critical load concept generally has been accepted as the basis for abatement
strategies to reduce or prevent injury to the functioning and vitality of forest ecosystems caused
by long-range transboundary .acidic deposition (Lokke, et al., 1996). The critical load has been
defined as a "quantitative estimate of an exposure to one or more pollutants below which
significant harmful effects on specified sensitive elements of the environment do not occur
according to present knowledge" (Lokke et al., 1996). A biological indicator, a chemical
criterion, and a critical value are the elements used in the critical load concept. The biological
indicator is the organism used to indicate the status of the receptor ecosystem, the chemical
criterion is the parameter that results in harm to the biological indicator, and the critical value is
the value of the chemical criterion below which no significant harmful response occurs to the
biological indicator (Lokke et al., 1996). Trees, and sometimes other plants, are used as the
biological indicators in the case of critical loads for forests. The critical load calculation using
the current methodology, is essentially an acidity/alkalinity mass balance calculation. The
chemical criterion must be expressible in terms of alkalinity. Initially, the Ca/Al ratio was used,
but, recently, the (Ca+Mg+K)/Al ratio has been used (Lokke et al., 1996).
Ideally, changes in acidic deposition should result in changes in the status of the biological
indicator used in the critical load calculation. However, the biological indicator is the integrated
response to a number of different stresses. Furthermore, there are other organisms more sensitive
March 2001
4-61
DRAFT-DO NOT QUOTE OR CITE
-------
1 to acid deposition than trees. At high concentrations, Al3+ is known to be toxic to plants,
2 inhibiting root growth and, ultimately, plant growth and performance (Lokke et al., 1996;
3 National Science and Technology Council, 1998). Sensitivity to Al varies considerably between
4 species and within species because of changes in nutritional demands and physiological status,
5 which are related to age and climate. Experiments have shown that there are large variations in
6 Al sensitivity, even among ecotypes.
7 Mycorrhizal fungi as possible biological indicators have been suggested by Lokke et al.
8 (1996) because they are intimately associated with tree roots, depend on plant assimilates, and
9 play an essential role in plant nutrient uptake, influencing the ability of their host plants to
10 tolerate different anthropogenically generated stresses. Mycorrhizas and fine roots are an
11 extremely dynamic component of below-ground ecosystems and can respond rapidly to stress.
12 They have a relatively short life span, and their turnover appears to be strongly controlled by
13 environmental factors. Changes in mycorrhizal species composition or the loss of dominant
14 mycorrhizal species in areas where diversity is already low may lead to increased susceptibility of
15 plant to stress (Lokke et al., 1996). Stress affects the total amount of carbon fixed by plants and
16 modifies carbon allocation to biomass, symbionts and secondary metabolites. Because
17 . mycorrhizal fungi are dependent for their growth on the supply of assimilates from the host
18 plants, stresses that shift the allocation of carbon reserves to the production of new leaves at the
19 expense of supporting tissues will be reflected rapidly in decreased fine root and mycorrhizzal
20 biomass (Winner and Atkinson, 1986). The physiology of carbon allocation has also been
21 suggested as an indicator of anthropogenic stress (Andersen and Rygiewicz, 1991). Soil
22 dwelling animals are important for decomposition, soil aeration, and nutrient redistribution in the
23 soil. They contribute to decomposition and nutrient availability mainly by increasing the
24 accessibility of dead plant material to microorganisms. Earthworms decrease in abundance and
25 in species number in acidified soils Lokke et al., 1996).
26
27 Biogeochemical Cycling—The Integrated Forest Study
28 The Integrated Forest Study (IPS) (Johnson and Lindberg, 1992a) has provided the most
29 extensive data set available on wet and dry deposition and the effects of deposition on the cycling
30 of elements in forest ecosystems. The overall patterns of deposition and cycling have been
31 summarized by Johnson and Lindberg (1992a), and the reader is referred to that reference for
March 2001
4-62
DRAFT-DO NOT QUOTE OR CITE
-------
1 details. The following is a summary of particulate deposition, total deposition, and leaching in
2 the IPS sites.
3 Particulate deposition in the IPS was separated at the 2-fj.m level; a decision was made to
4 include total particulate deposition in this analysis and may include the deposition of particles
5 larger than 10 //m.
6 Particulate deposition contributes considerably to the total impact of base cations to most of
7 the IPS sites. On average, particulate deposition contributes 47% to total calcium deposition
8 (range: 4 to 88%), 49% of total potassium deposition (range: 7 to 77%), 41% to total magnesium
9 deposition (range: 20 to 88%), 36% to total sodium deposition (range: 11 to 63%), and 43% to
10 total base cation deposition (range: 16 to 62%). Of the total particulate deposition, the vast
11 majority (>90%) is >2 //m.
12 Figures 4-5 through 4-8 summarize the deposition and leaching of calcium, magnesium,
13 potassium, and total base cations for the IPS sites. As noted in the original synthesis (Johnson
14 and Lindberg, 1992a), some sites show net annual gains of base cations (i.e., total deposition
15 > leaching), some show losses (total deposition < leaching), and some are approximately in
16 balance. Not all cations follow the same pattern at each site. For example, calcium shows net
17 accumulation at the Coweeta, TN; Durham (Duke), NC; and Florida sites (Figure 4-5), potassium
18 shows accumulation at the Duke; Florida; Douglas-fir; red alder; Thompson, WA; Huntingdon
19 Forest, NY; and Whiteface Mountain, NY, sites (Figure 4-7), and magnesium accumulated only
20 at the Florida sites (Figure 4-6). Only at the Florida site is there a clear net accumulation of total
21 base cations (Figure 4-8).
22 The factors affecting net calcium accumulation or loss include the soil-exchangeable cation
23 composition, as noted previously; base cation deposition rate; the total leaching pressure because
24 of atmospheric sulfur and nitrogen inputs, as well as natural (carbonic and organic) acids; and
25 biological demand (especially for potassium). At the Florida site, which has a very cation-poor,
26 sandy soil (an Ultic Haploquod derived from marine sand), the combination of all these factors
27 leads to net base cation accumulation from atmospheric deposition (Johnson and Lindberg,
28 1992a). The site showing the greatest net base cation losses, the red alder stand in Washington
29 state, is one that is under extreme leaching pressure by nitrate produced because of excessive
30 fixation by that species (Van Miegroet and Cole, 1984). hi the red spruce site in the Smokies,
31 the combined effects of SO42" and NO3" leaching are even greater than hi the red alder site
March 2001
4-63
DRAFT-DO NOT QUOTE OR CITE
-------
1,000
500
CO
s,
-500
JD
1
g. -1,000
01
-1,500 -
-2,000
63%
41%
E
H
K < 2|jm
HI Wet
^ Leaching
Pe
55%
rcent c
65%
I
I
)f total
40%
1
[zJ
depos
26%
R
M
tion a:
25%
fl
I
I
I
> partic
49%
m
;les:
58%
I
88%
0
4%
IP
i
i
[69
|
1
>
CP DL GS LP FS
- Warmer Sites
DF RA NS
-> •* — —
FF MS WF
- Colder Sites—
ST
Figure 4-5. Calcium deposition in >2-/^m particles, <2-/zm particles, and wet forms (upper
bars) and leaching (lower bars) in the Integrated Forest Study sites.
CP = Pinus strobus, Coweeta, TN; DL = Pinus taeda, Durham (Duke), NC;
GS = Pinus taeda, B. F. Grant Forest, GA; LP = Pinus taeda, Oak Ridge, TN;
FS = Pinus ettottii, Bradford Forest, FL; DF = Psuedotsuga menziesii,
Thompson, WA; RA = Alnus rubra; Thompson WA; NS = Picea abies,
Nordmoen, Norway; HF = northern hardwood, Huntington Forest, NY;
MS = Picea rubens, Rowland, ME; WF = Picea rubens, Whiteface Mountain,
NY; and ST = Picea rubens, Clingman's Dome, NC.
1 (Figure 4-9), but a considerable proportion of the cations leached from this extremely acid soil
2 consist of H1" and A13+ rather than of base cations (Johnson and Lindberg, 1992a). Thus, the red
3 spruce site in the Smokies is approximately in balance with respect to calcium and total base
4 cations, despite the very high leaching pressure at this site (Figures 4-5 and 4-8).
5 ! The relative importance of particulate base cation deposition varies widely with site and
6 cation and is not always related to the total deposition rate. The proportion of calcium deposition
7 in particulate form ranges from a low of 4% at the Whiteface Mountain site to a high of 88% at
8 the Maine site (Figure 4-5). The proportion of potassium deposition as particles ranges from
9 7% at the Smokies site to 77% at the Coweeta site (Figure 4-7), and the proportion of total base
March 2001
4-64
DRAFT-DO NOT QUOTE OR CITE
-------
1,000
400
v- 200
CO
CO
-200
CT
UJ -400
-600 -
-800
40%
^
1
k^d
49%
PS
I
[ | > 2 |jm
| < 2 |jm
^j V\fet
^ Leaching
P
44%
srcent
48%
Essl
I
I
of tota
36%
••M
y
depo
27%
PI
1
sition a
27%
PP
1
i
I
s parti
39%
1
cles:
46%
0
88%
i
20%
I
m
26%
§
i
CP DL GS LP FS DF RA NS FF MS WF ST
— Warmer Sites *• * Colder Sites
Figure 4-6. Magnesium deposition in >2-/an particles, <2-//m particles, and wet forms
(upper bars) and leaching (lower bars) in the Integrated Forest Study sites.
See Figure 4-5 for legend.
1
2
3
4
5
6
7
8
9
10
11
12
cation deposition ranges from 16% at the Whiteface site to 62% at the Maine site (Figure 4-8).
Overall, participate deposition at the site in Maine accounted for the greatest proportion of
calcium, magnesium, potassium, and base cation deposition (88, 88,57, and 62%, respectively),
even though total deposition was relatively low. At some sites, the relative importance of
particulate deposition varies considerably by cation. At the Whiteface Mountain site, particulate
deposition accounts for 4, 20, and 40% of calcium, magnesium, and potassium deposition,
respectively. At the red spruce site in the Smokies, particulate deposition accounts for 46, 26%,
7% of calcium, magnesium, and potassium deposition, respectively.
As observed in the IPS synthesis, SO42' and NOj leaching often are dominated by
atmospheric sulfur and nitrogen (Johnson and Lindberg, 1992a). The exceptions to this are in
cases where natural nitrogen inputs are high (i.e.,the nitrogen-fixing red alder stand), as are NOj
leaching rates, even though nitrogen deposition is low, and where soils adsorb much of the
March 2001
4-65
DRAFT-DO NOT QUOTE OR CITE
-------
co
Q>
1
S
400-
300-
200-
100-
0-
-100-
-200-
-300-
-400-
-500.
77%
76%
m
|
1
H< 2(jm
H! Wet
^ Leaching
Pe
61%
F1
^
jrcent
48%
i
3f total
54%
H
depos
40%
H
E
it on a:
40%
•
^
%
s partic
68%
N
;ulates
23%
n
57%
n
^
40%
H
7%
1
1
^
X^
^
^
I
CP
DL
GS LP FS DF RA NS FF MS WF
Warmer Sites ^ +— Colder Sites-
ST
Figure 4-7. Potassium deposition in >2-/zm particles, <2-Atm particles, and wet forms
(upper bars) and leaching (lower bars) in the Integrated Forest Study sites.
See Figure 4-5 for legend.
1 atmospherically deposited SO42', thus reducing SO42" leaching compared to atmospheric sulfur
2 input.
3 Sulfate and NO3" leaching have a major effect on cation leaching in many of the IPS sites
4 (Johnson and Lindberg, 1992a). Figure 4-9 shows the total cation leaching rates of the IPS sites
5 and the degree to which cation leaching is balanced by SO42' + NO3" deposition. The SO42' and
6 NO3' fluxes are subdivided further into that proportion potentially derived from particulate sulfur
7 and nitrogen deposition (assuming no ecosystem retention, a maximum effect) and other sulfur
8 and nitrogen sources (wet and gaseous deposition, internal production).
9 As noted in the IPS synthesis, total SO42' and NO3" inputs account for a large proportion
10 (28 to 88%) total cation leaching in most sites. The exception is the Georgia loblolly pine site,
11 where there were high rates of HCO{ and Cl" leaching (Johnson and Lindberg, 1992a). The role
12 of particulate sulfur and nitrogen deposition in this leaching is generally very small (<10%),
13 however, even if it is assumed that there is no ecosystem sulfur or nitrogen retention.
March 2001
4-66
DRAFT-DO NOT QUOTE OR CITE
-------
3,000
2,000
L 1,000
OJ
-1,000
cr
-2,000
-3,000
-4,000
53%
1
47%
S
1
| | > 2 |jm
Jj < 2|jm
H Wet
^ Leaching
P
48%
1
ercent
62%
5S
I
n
of tote
49%
^^ ^^
1
1
1 depo
28%
H
1
sition e
28%
i
I
I
I
is part
47%
H
1
culate
44%
i
s:
62%
1
16%
1
31%
1
I
i
CP DL GS LP FS DF
"* Warmer Sites —
NS FF MS WF ST
Colder Sites >•
Figure 4-8. Base cation deposition in >2-yum particles, <2-//m particles, and wet forms
(upper bars) and leaching (lower bars) in the Integrated Forest Study sites.
See Figure 4-5 for legend.
1
2
3
4
5
6
7
8
9
10
11
12
It was noted previously in this chapter that the contribution of particles to total deposition
of nitrogen and sulfur at the IPS sites is lower than is the case for base cations. On average,
particulate deposition contributes 18% to total nitrogen deposition (range: 1 to 33%) and 17%
to total sulfur deposition (range: 1 to 30%). Particulate deposition contributes only a small
amount to total ET deposition (average = 1%; range: 0 to 2%). (It should be noted, however,
that particulate H+ deposition in the >2-jum fraction was neglected.)
Based on the IPS data, it appears that the particulate deposition has a greater effect on base
cation inputs to soils than on base cation losses associated with inputs of sulfur, nitrogen, and H+.
It cannot be determined what fraction of the mass of these particles are <10 yum, but only a very
small fraction is <2 /zm. These inputs of base cations have considerable significance, not only to
the base cation status of these ecosystems, but also to the potential of incoming precipitation to
acidify or alkalize the soils in these ecosystems. As noted above, the potential of precipitation to
March 2001
4-67
DRAFT-DO NOT QUOTE OR CITE
-------
(0
7,000
6,000
5,000
4,000 - •
% of total cation leaching balanced by SOf and NOs from Particles (P) and other (O) sources
P: 4 %
O:28%
7%
55%
1%
6%
8%
78%
9%
30%
10%
42%
Other Anions
Particulate Sulphur and Nitrogen
Other Sulphur and Nitrogen Sources
10%
88%
8%
55%
6%
73%
18%
69%
1%
77%
9%
69%
MS WF ST
Figure 4-9. Total cation leaching (total height of bar) balanced by sulfate and nitrate
estimated from particulate deposition (assuming no ecosystem retention,
particulate sulfur and nitrogen) and by other sources (both deposition and
internal) of sulfate and nitrate (other sulfur and nitrogen sources) and by
other anions in the Integrated Forest Study sites. See Figure 4-5 for legend.
1 acidify or alkalize soils depends on the ratio of base cations to H+ in deposition, rather than
2 simply on the inputs of H+ alone. In the case of calcium, the term "lime potential" has been
3 applied to describe this ratio; the principle is the same with respect to magnesium and potassium.
4 Sodium is a rather special case, in that it is a poorly absorbing cation, and leaching tends to
5 balance input over a relatively short term.
6 Net balances of base cations tell only part of the story as to potential effects on soils; these
7 net losses or gains must be placed in the perspective of the soil pool size. One way to express
8 this perspective is to simply compare soil pool sizes with the net balances. This comparison is
9 made for exchangeable pools and net balances for a 25-year period in Figures 4-10 to 4-12.
10 It readily is seen that net leaching losses of cations pose no threat in terms of depleting
11 soil-exchangeable Ca2"1", K+, or magnesium ion within 25 years at the Coweeta, Duke, Georgia,
March 2001
4-68
DRAFT-DO NOT QUOTE OR CITE
-------
350,000
Soil Exchangeable
(Dep-Leaching)*25
DL GS LP FS
Warmer Sites
HF MS WF ST
Colder Sites - >•
Figure 4-10. Soil exchangeable Ca2+ pools and net annual export of Ca2+ (deposition minus
leaching times 25 years) in the Integrated Forest Study sites. See Figure 4-5
for legend.
1
2
3
4
5
6
7
8
9
10
11
12
Oak Ridge, or Douglas-fir sites. There, however, is a potential for significant depletion at the red
alder, Whiteface Mountain (magnesium), and Smokies red spruce sites.
The range of values for soil-exchangeable turnover is very large, reflecting variations in
both the size of the exchangeable pool and the net balance of the system. Soils with the highest
turnover rates are those most likely to experience changes in the shortest time interval, other
things being equal. Thus, the Whiteface Mountain, Smokies, and Maine red spruce sites; the
Thompson red alder site; and the Huntington Forest northern hardwood site appear to be most
sensitive to change. The actual rates, directions, and magnitudes of changes that may occur in
these soils (if any) will depend on weathering inputs and vegetation outputs, in addition to
deposition and leaching. It is noteworthy that each of the sites listed above as sensitive has a
large store of weatherable minerals, whereas many of the other soils, with larger exchangeable
cation reserves, have a small store of weatherable minerals (e.g., Coweeta white pine, Duke
March 2001
4-69
DRAFT-DO NOT QUOTE OR CITE
-------
I I Soil Exchangeable
• (Dep - Leaching)*25
GS LP FS
Warmer Sites
HF MS WF
Colder Sites
Figure 4-11. Soil exchangeable Mg2+ pools and net annual export of Mg2"1" (deposition
minus leaching times 25 years) in the Integrated Forest Study sites.
See Figure 4-5 for legend.
1 loblolly pine, Georgia loblolly pine, and Oak Ridge loblolly pine) (Johnson and Lindberg, 1992a;
2 April and Newton, 1992).
3 Base cation inputs are especially important to the Smokies red spruce site because of
4 potential aluminum toxicity and calcium and magnesium deficiencies. Johnson et al. (1991a)
5 found that soil solution aluminum concentrations occasionally reached levels found to inhibit
6 calcium uptake and cause changes in root morphology in solution culture studies of red spruce
7 (Raynal et al., 1990). In a follow-up study, Van Miegroet et al. (1993) found a slight but
8 significant growth response to calcium and magnesium fertilizer in red spruce saplings near the
9 Smokies red spruce site. Joslin et al. (1992) reviewed soil and solution characteristics of red
10 spruce in the southern Appalachians, and it would appear that the IPS site is rather typical.
11 Wesselink et al. (1995) reported on the complicated interactions among changing
12 deposition and soils at this site (including repeated sampling of soil exchangeable base cation
13 pools) from 1969 to 1991 and compared these results with those of a simulation model. They
March 2001
4-70
DRAFT-DO NOT QUOTE OR CITE
-------
160000 •} •
140 000 •
<0
9J, 120 000 •
x: 100000 •
8"
o 80000 •
^ 60 000 •
cr
-------
1 variety of forest ecosystems and to determine if these effects are in any way related to current or
2 potential forest decline. Acidic deposition is having a significant effect on nutrient cycling in
3 most of the forest ecosystems studied in the EPS project. The exceptions were the relatively
4 unpolluted Douglas fir, red alder, and Findley Lakes in Washington state. The nature of the
5 effects, however, varies from one location to another (Johnson, 1992). hi all but the relatively
6 unpolluted Washington sites, atmospheric deposition was having a significant, often
7 overwhelming effect on cation leaching from the soils. In general, nutrient budget data from IPS
8 and literature suggest that the susceptibility of southeastern sites to base cation depletion from
9 soils and the development of cation deficiencies by that mechanism appears to be greater than in
10 northern sites (Johnson, 1992).
11 Atmospheric deposition may have affected significantly the nutrient status of some IPS
12 sites through the mobilization of Al. Soil solution Al levels in the Smokies sites approach and
13 sometimes exceed levels noted to impede cation uptake in solution culture studies. It is therefore
14 possible that the rates of base cation uptake and cycling in these sites have been reduced because
15 of soil solution Al. To the extent that atmospheric deposition has contributed to these elevated
16 soil solution Al levels, it has likely caused a reduction in base cation uptake and cycling rates at
17 ' these sites. Nitrate and sulfate are the dominant anions in the Smokies sites, and nitrate pulses
18 are the major cause of Al pulses in soil solution (Johnson, 1992). The connection between Al
19 mobilization and forest decline is not clear. The decline in red spruce certainly has been more
20 severe in the Northeast than in the Southeast, yet all evidence indicates that Al mobilization is
21 most pronounced in the southern Appalachians. However, at the Whiteface Mountain site
22 selected for study because it was in a state of decline, soil solution levels there are lower than in
23 the Smokies, which are in a visibly obvious state of decline (e.g., no dieback other than the fir
24 killed by the balsam wooly adelgid, no needle yellowing, etc). Thus, Al mobilization constitutes
25 a situation worthy of further study (Johnson, 1992).
26 The simple calculations shown above give some idea of the importance of particulate
27 deposition in these forest ecosystems, but they cannot account for the numerous potential
28 feedbacks between vegetation and soils nor for the dynamics through time that can influence the
29 ultimate response. One way to examine some of these interactions and dynamics is to use
30 simulation modeling. The nutrient cycling model (NuCM) has been developed specifically for
31 this purpose and has been used to explore the effects of atmospheric deposition, fertilization, and
March 2001
4-72
DRAFT-DO NOT QUOTE OR CITE
-------
1 harvesting on some of the IPS sites (Johnson et al., 1993). The NuCM model is a stand-level
2 model that incorporates all major nutrient cycling processes (uptake, translocation, leaching,
3 weathering, organic matter decay, and accumulation).
4 Johnson et al. (1999) used the NuCM model to simulate the effects of reduced S, N, and
5 base cation (CB) deposition on nutrient pools, fluxes, soil, and soil solution chemistry in two
6 contrasting southern Appalachian forest ecosystems: (1) the red spruce and (2) Coweeta
7 Hardwood sites from the EPS project. The scenarios chosen for these simulations included
8 "no change", 50% N and S deposition, 50% CB deposition, and 50% N, S, and CB deposition
9 (50% N, S, CB). The NuCM simulations suggested that, for the extremely acid red spruce site,
10 S and N deposition is the major factor affecting soil solution Al concentrations and CB deposition
11 is the major factor affecting soil solution CB concentrations. The effects of S and N deposition
12 were largely through changes in soil solution SO42' and NO3" and, consequently, mineral acid
13 anion (MAA) concentrations rather than through changes in soils. This is illustrated in
14 Figures 4-13 and 4-14, which shows simulated soil solution mineral acid anions, base cations,
15 Al, and soil base saturation in B horizon from in the red spruce site. The 50% S and N scenario
16 caused reductions in soil solution SO42', NO3" and, therefore, MAA concentrations, as expected,
17 and this, in turn, caused short-term reductions in base cation concentrations. However, by the
18 end of the 24-year simulation, base cations in the 50% S, N scenario were nearly as high as in the
19 no change scenario because base saturation had increased and the proportion of cations as Al
20 decreased. The 50% CB scenario had virtually no effect on soil solution SO42", NO3" and,
21 therefore, MAA concentrations, as expected, but did cause a long-term reduction in base cation
22 concentrations. This was caused by a long-term reduction in base saturation (Figure 14). Thus,
23 the effects of CB deposition were solely through changes in soils rather than through changes in
24 soil solution MAA, as postulated by Driscoll et al. (1989). In the less acid Coweeta soil, base
25 saturation was high and little affected by scenario (not shown), Al was unimportant, and S and
26 N deposition had a much greater effect than CB deposition in all respects (Figure 15).
27 In summary, Johnson et al. (1999) found that the results of the red spruce simulations
28 support the hypothesis of Driscoll et al. (1989) in part: CB deposition can have a major effect on
29 CB leaching through time in an extremely acid system. This effect occurred through changes in
30 the soil exchanger and not through changes in soil solution MAA concentration. On the other
31 hand, S and N deposition had a major effect on Al leaching at the Noland Divide site. This
March 2001
4-73
DRAFT-DO NOT QUOTE OR CITE
-------
Red Spruce
500 —
400
300
200 —
100
Mineral Acid Anions
No Change
50%N,S
50% BC
200
1 23 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Year
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Figure 4-13. Simulated soil solution mineral acid anions and base cations in the red spruce
site with no change, 50% N and S deposition, and 50% base cation
deposition. Redrawn from Johnson et al. (1999).
March 2001
4-74
DRAFT-DO NOT QUOTE OR CITE
-------
Red Spruce
500
400
300-
No Change
50%N,S
50% BC
200 —
100 —
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Year
5 —
Base Saturation
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Figure 4-14. Simulated soil solution Al and soil base saturation in the red spruce site with
no change, 50% N and S deposition, and 50% base cation deposition.
Redrawn from Johnson et al. (1999).
March 2001
4-75
DRAFT-DO NOT QUOTE OR CITE
-------
Coweeta
100
Mineral Acid Anions
No Change
50% S,N
50% BC
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Base Cations
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Year
Figure 4-15. Simulated soil solution mineral acid anions and base cations in the Coweeta
site with no change, 50% N and S deposition, and 50% base cation
deposition. Redrawn from Johnson et al. (1999).
March 2001
4-76
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
occurred primarily because of changes in soil solution MAA concentration. At the less acidic
Coweeta site, CB deposition had a minor effect on soils and soil solutions, whereas S and N
deposition had delayed but major effects on CB leaching because of changes in SO42~ and MAA
concentrations.
Trace Element Effects
Trace metals are natural elements that are ubiquitous in small (trace) amounts in soils,
ground water and vegetation. Many are essential elements required for growth by plants and
animals as micronutrients. Naturally occurring surface mineralizations can produce metal
concentrations in soils and vegetation that are as high, or higher, than those in the air and
deposited near man-made sources (Freedman and Hutchinson, 1981). The occurrence and
concentration of trace metals in any ecosystem component depend on the sources of the metal via
the soil or as particulate. Even when air pollution is the primary source, continued deposition
can result in the accumulation of trace metals in the soil (Martin and Coughtrey, 1981). Many
metals are deposited into soils by chemical processes and are not available to plants (Saunders
andGodzik, 1986).
When aerial deposition is the primary source of metal particles, both the chemical form and
particle size deposited determine the heavy metal concentration in the various ecosystem
components (Martin and Coughtrey, 1981). Human activities introduce heavy metals into the
atmosphere and have resulted in the deposition of antimony, cadmium, chromium, copper, lead,
molybdenum, nickel, silver, tin, vanadium, and zinc (Smith, 1990c). Extensive evidence
indicates that heavy metals deposited from the atmosphere to forests accumulate either in the
richly organic forest floor or in the soil layers immediately below, areas where the activity in
roots and soil is greatest. The greater the depth of soil, the lower the metal concentration. The
accumulation of metal in the soil layers where the biological activity is greatest, therefore, has the
potential for being toxic to roots and soil organisms and interfering with nutrient cycling (Smith,
1990e). Though all metals can be directly toxic at high levels, only toxicity from copper, nickel,
and zinc have been documented frequently. Toxicity of cadmium, cobalt, and lead has been seen
only under unusual conditions (Smith, 1990e). Exposures at lower concentrations have the
potential, over the long-term, for interfering with the nutrient-cycling processes when they affect
mycorrhizal function.
March 2001
4-77
DRAFT-DO NOT QUOTE OR CITE
-------
1 Accumulation of heavy metals in litter presents the greatest potential for interference with
2 nutrient cycling. Accumulation of metals in the litter occurs chiefly around brass works and lead
3 and zinc smelters. There is some evidence that invertebrates inhabiting soil litter do accumulate
4 metals. Earthworms from roadsides were shown to contain elevated concentrations of cadmium,
5 nickel, lead, and zinc; however, interference with earthworm activity was not cited (Martin and
6 Coughtrey, 1981). It has been shown, however, that when soils are acidic, earthworm abundance
7 decreases and bioaccumulation of metals from soil may increase exponentially with decreasing
8 pH (Lokke et al, 1996). Organisms that feed on earthworms living in soils with elevated levels
9 of Cd, Ni, Pb, and Z for extended periods could accumulate lead and zinc to toxic levels (Martin
10 and Coughtrey, 1981). Increased concentrations of heavy metals have been found in a variety of
11 small mammals living in areas with elevated heavy metal concentrations in the soils.
12 Studies by Babich and Stotsky (1978) support the concept that increased accumulation of
13 litter in metal-contaminated areas is the result of effects on the microorganismal populations.
14 Cadmium toxicity to microbial populations was observed to decrease and prolong logarithmic
15 rates of microbial increase, to reduce microbial respiration and fungal spore formation and
16 germination, to inhibit bacterial transformation, and to induce abnormal morphologies. Also, the
17 effects of cadmium, copper, nickel, and zinc on the symbiotic activity of fungi, bacteria, and
18 actinomycetes were reported by Smith (1991). The formation of mycorrhizae by Glomus
19 mosseae with onions was reduced when zinc, copper, nickel, or cadmium was added to the soil.
20 The relationship of the fungus with white clover, however, was not changed. It was suggested
21 that the effect of heavy metals on vesicular-arbuscular mycorrhizal fungi will vary from host to
22 host (Gildon and Tinker, 1983). Studies with ericoid plants indicated that, in addition to Calluna
23 vulgaris, mycorrhizae also protect Vaccinium macrocarpa and Rhodendron ponticum from heavy
24 metals (Bradley et al., 1981). Heavy metals tend to accumulate in the roots, and shoot toxicity is
25 prevented.
26 The effects of sulfur deposition on litter decomposition in the vicinity of smelters also must
27 be considered. Metal smelters emit SO2 as well as heavy metals. Altered litter decomposition
28 rates have been well documented near SO2 sources (Prescott and Parkinson, 1985). The presence
29 of sulfur in litter has been associated with reduced microbial activity (Bewley and Parkinson,
30 1984). Additionally, the effects on symbiotic activity of fungi, bacteria and actinomycetes were
31 reported by Smith (1990b).
March 2001
4-78
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
The potential pathways of accumulation of trace metals in terrestrial ecosystems, as well as
the possible consequences of trace metal deposition on ecosystem functions is summarized in
Figure 4-16. The generalized trophic levels found in an ecosystem and the various physiological
and biological processes that could be affected by trace metals are shown in the figure.
Reduction in physiological processes can affect productivity, fecundity, and mortality (Martin
and Coughtrey, 1981). Therefore, any effects on structure and function of an ecosystem are
likely to occur through the soil and litter (Tyler, 1972).
Trace metals deposited from the atmosphere to forests accumulate either in the richly
organic forest floor or in the soil layers immediately below, layers where greatest biological
activity occurs. The shallow-rooted species plant species are those most likely to take up metals
from the soil (Martin and Coughtrey, 1981).
Certain species of plants are tolerant of metal contaminated soils (e.g., soils from mining
activities) (Antonovics et al., 1971). Certain species of plants also have been used as
bioindicators of metals (e.g., Astragalius is an accumulator of selenium). The sources of both
macroelements and trace metals in the soil of the Botanical Garden of the town of Wroclow,
Poland, were determined by measuring the concentrations of the metals in Rhododendron
catawbiense, Ilex aquifolium, and Mahonia aquifolium growing in the garden and comparing the
results with the same plant species growing in two other botanical gardens in nonpolluted areas.
Air pollution deposition was determined as the source of metals in plants rather than the soil
(Samecka-Cymerman and Kempers, 1999).
Biological accumulation of metals through the plant-herbivore and litter-detrivore chains
can occur. A study of the accumulation of cadmium, lead, and zinc concentrations in
earthworms suggested that cadmium and zinc were concentrated, but not lead. Studies indicate
that heavy metal deposition onto the soil, via food chain accumulation, can cause excessive
levels and toxic effects in certain animals. Cadmium appears to be relatively mobile within
terrestrial food chains; however, the subsequent mobility of any metal after it is ingested by a
herbivorous animal depends on the site of accumulation within body tissues. Although food
chain accumulation may not in itself cause death, it can reduce the breeding potential in a
population (Martin and Coughtrey, 1981).
In actual case studies, it was observed that the deposition of copper and zinc particles
around a brassworks resulted in an accumulation of incompletely decomposed litter. In one
March 2001
4-79
DRAFT-DO NOT QUOTE OR CITE
-------
vo
1—I
4
I
March 2001
4-80
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
study, litter accumulation was reported up to 7.4 km from the stack of a primary smelter in
southeastern Missouri. Similar results were reported around a metal smelter at Avonmouth,
England. In the latter case, litter accumulation was associated closely with concentrations
specifically of cadmium, as well as with those of lead, copper, and zinc (Martin and Coughtrey,
1981). Experimental data (using mesh bags containing litter) supports the hypothesis that
reduced decomposition occurs close to heavy metal sources.
Accumulations of metals emitted in particulate matter also were reported in soil litter close
to a metal smelter at Palmerton, PA, in 1975 and 1978. The continued presence of cadmium,
lead, zinc, and copper in the upper soil horizons (layers) were observed 6 years after the smelter
terminated operation in 1980. Metal levels were highest near the smelter. The relationship of
decreasing amounts of metal in body tissues also held true for amphibians and mammals. Levels
of cadmium in kidneys and liver of white-tailed deer (Odocoileus virginaus) were five times
higher at Palmerton than in those collected 180 km southwest downwind. The abnormal
amounts of metal in the tissues of terrestrial vertebrates and the absence or low abundance of
wildlife at Palmerton indicated that ecological processes within 5 km of the zinc smelter
continued to be markedly influenced even 6 years after its closing (Storm et al., 1994).
The effects of lead in ecosystems are discussed in fo&'Air Quality Criteria for Lead
(U.S. Environmental Protection Agency, 1986b). Studies have shown that there is cause for
concern in three areas where ecosystems may be extremely sensitive to lead: (1) delay of
decomposition because the activity of some decomposer microorganisms and invertebrates is
inhibited by lead, (2) subtle shifts toward plant populations tolerant of lead, and (3) lead in the
soil and on the surfaces of vegetation circumvent the processes of biopurification. The problems
cited above arise because lead is deposited on the surface of vegetation, accumulates in the soil,
and is not removed by the surface and ground water of the ecosystem (U.S. Environmental
Protection Agency, 1986b).
4.2.3 Ecosystem Goods and Services and Their Economic Valuation
Human existence on this planet depends on ecosystems and the services and products they
provide. The essential services and products provided by the planet's collective biodiversity (the
earth's flora, fauna, and microorganisms) are clean air, clean water, clean soil, and clean energy
(Table 4-6). Today, governments around the world pursue a "bottom line" that driven is by an
March 2001 4-81 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-6. PRIMARY GOODS AND SERVICES PROVIDED BY ECOSYSTEMS
Ecosystem Goods Services
Agroecosvstems
Coastal ecosystems
Forest ecosystems
Freshwater
Grassland
ecosystems
• Food crops
• Fiber corps
• Crop genetic resources
• Fish and shellfish
• Fishmeal (animal feed)
• Seaweeds (for food and industrial
use)
•Salt
• Genetic resources
• Timber
• Fuelwood
• Drinking and irrigation water
• Fodder
• Nontimber products (vines,
bamboos, leaves, etc.)
• Food (honey, mushrooms fruit,
and other edible plants; game)
• Genetic resources
• Drinking and irrigation water
•Fish
• Hydroelectricity
• Genetic resources
• Livestock (food, game, hides, and
fiber)
• Drinking and irrigation water
• Genetic resources
• Maintain limited watershed functions (infiltration,
flow control, and partial soil protection)
• Pro vide-habitat for birds, pollinators, and soil
organisms important to agriculture
• Sequester atmospheric carbon
• Provide employment
• Moderate storm impacts (mangroves, barrier
islands)
• Provide wildlife (marine and terrestrial) habitat
and breeding areas/hatcheries/nurseries
• Maintain biodiversity
• Dilute and treat wastes
• Provide harbors and transportation routes
• Provide human and wildlife habitat
• Provide employment
• Contribute aesthetic beauty and provide recreation
• Remove air pollutants, emit oxygen
• Cycle nutrients
• Maintain array of watershed functions (infiltration,
purification, flow control, soil stabilization)
• Maintain biodiversity
• Sequester atmospheric carbon
• Moderate weather extremes and impacts
• Generate soil
• Provide employment
• Provide human and wildlife habitat
• Contribute aesthetic beauty and provide recreation
•Buffer water flow (control timing and volume)
• Dilute and carry away wastes
• Cycle nutrients
• Maintain biodiversity
• Provide aquatic habitat
• Provide transportation corridor
• Provide employment
• Contribute aethetic beauty and provide recreation
•Maintain array of watershed functions (infiltration,
purification, flow control, and soil stabilization)
• Cycle nutrients
• Remove air pollutants and emit oxygen
• Maintain biodiversity
• Generate soil
• Sequester atmospheric carbon
• Provide human and wildlife habitat
• Provide employment
• Contribute aesthetic beauty and provide recreation
Source: World Resources (2000-2001).
March 2001
4-82
DRAFT-DO NOT QUOTE OR CITE
-------
1 economy that is disconnected from the natural world and is fundamentally destructive of local
2 ecosystems (Suzuki, 1997). For this reason, human society needs to be reconnected to the
3 biologically diverse ecosystems and the natural world of which they are a part (Suzuki, 1997).
4 There is a need to understand the biodiversity that encompasses all levels of biological
5 organization, including populations, individuals, species and ecosystems (Wilson, 1997).
6 Populations, geographical entities within a species of organisms, usually distinguished
7 ecologically or genetically, are essential to the conservation of species diversity. Their number
8 and size influence the probability of the future existence of the entire species (Hughes et al.,
9 1997). The number, biodiversity, structure, and functions of ecosystem populations, provide
10 ecosystem products (goods) and services. For any given population, the number of individuals,
11 the genetic variation between individuals, and the area occupied affects ecosystem functioning
12 and the delivery of ecosystem services and other benefits provided by that population (Hughes,
13 et al., 1997). Loss of population diversity means loss of the benefits described in Table 4-6 and,
14 in particular, with time, the loss of the life-support systems on which humanity relies (Hughes
15 etal., 1997).
16 Attempts have been made to value biodiversity and the world's ecosystem services and
17 natural capital (Pimentel et al., 1997; Costanza et al., 1997). Pimentel et al. (1997) estimated
18 economic and environmental benefits for services contributed from all biota (biodiversity) in the
19 United States, including their genes, at $319 billion per year. Costanza et al. (1997) have
20 estimated the total value of ecosystem services by biome for the entire bioshere. Ecosystems
21 provide at least $33 trillion worth of services annually. Approximately, 63% of the estimated
22 value is contributed by marine ecosystems ($20.9 trillion per year), most of which comes from
23 coastal ecosystems ($10.6 trillion per year). About 38% of the estimated value comes from
24 terrestrial ecosystems, mainly from forests ($4.7 trillion per year) and wetlands ( $4.9 trillion per
25 year). Costanza et al. (1997) state that it may never be possible to make a precise estimate of the
26 services provided by ecosystems. Their estimates, however, indicate the relative importance of
27 ecosystem services.
28 Heal (2000), however, feels that attempts to value ecosystems and their services are
29 probably misplaced. "Economics cannot estimate the importance of natural environments to
30 society: only biology can do that" (Heal, 2000). The role of economics is to help design
31 institutions that will provide incentives to the public and policy makers for the conservation of
March 2001
4-83
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
important natural systems and for mediating human impacts on the biologically diverse
ecosystems and the biosphere so that they are sustainable. The approach of Harwell et al. (1999)
also deals with the need to understand human impacts on ecosystems so that ecosystem
management can define what ecological conditions are desired. Further, they state that the
establishment of ecological goals involves a close linkage between scientists and decision
makers, in which science informs decision makers and the public by characterizing the ecological
conditions that are achievable under particular management regimes. Decision makers then can
make choices that reflect societal values, including issues of economics, politics, and culture.
For management to achieve their goals, the general public, scientific community, resource
managers, and decision makers need to be routinely apprised of the condition or integrity of
ecosystems, so that ecological goals may be established (Harwell et al., 1999).
The above assessment of new information leads to the clear conclusion that atmospheric
PM at levels currently found in the United States has the potential to alter ecosystem structure
and function in ways that may reduce their ability to meet societal needs. The possible direct
effects of airborne PM on individual plants were discussed in Section 4.2.1 above. The major
impacts of airborne PM on ecosystems, however, are the indirect effects on plant populations that
occur through the soil and affect the cycling of nutrients necessary for plant growth and vigor, as
discussed in Section 4.2.2. By altering the cycling of nitrogen, nitrogen deposition changes the
biodiversity of ecosystems and their functioning and, by altering the vigor of forest tree stands,
alters forest succession. Also, nitrogen deposition in combination with the deposition of sulfur in
the form of acid rain alters the biogeochemical cycling of soil mineral nutrients and changes the
biodiversity and functioning of forest ecosystems. The changes in the ability of forest vegetation
and soil microorganisms to utilize nutrients results in the leaching of nitrates and other minerals
from the soils. The nitrate and mineral runoff impacts coastal and aquatic ecosystems and, thus,
influences the services important to human life provided by these ecosystems as well (Table 4-6).
March 2001
4-84
DRAFT-DO NOT QUOTE OR CITE
-------
1 4.3 EFFECTS ON VISIBILITY
2 4.3.1 Introduction
3 Visibility is defined as the degree to which the atmosphere is transparent to visible light and
4 the clarity (transparency) and color fidelity of the atmosphere (National Research Council, 1993).
5 Visibility impairment is defined as any humanly perceptible change in visibility (light extinction,
6 visual range, contrast, or coloration). Visual range is described as the farthest distance at which a
7 large black object can be distinquished against the horizontal sky (U.S. Environmental Protection
8 Agency, 1979). For regulatory purposes, visibility impairment is classified into two principal
9 forms: (1) "reasonably attributable" impairment, attributable to a single source or small group of
10 sources and (2) regional haze, described as any perceivable change in visibility (light extinction,
11 visual range, contrast, or coloration) from which would have existed under natural conditions
12 that is caused predominantly by a combination of many sources over a wide geographical area
13 (U.S. Environmental Protection Agency, 1999).
14 The objective of the visibility discussion in this section is to summarize the linkage
15 between air pollution, in particular particulate matter, and visibility. This section summarizes the
16 information discussed in the previous 1996 PM air quality criteria document (PM AQCD) and
17 includes additional relevant information available since publication of that document. For a
18 more detailed discussion on visibility, the reader is referred to the earlier PM AQCD entitled, Air
19 Quality Criteria for Particulate Matter (U.S. Environmental Protection Agency, 1996a), the
20 Recommendations of the Grand Canyon Visibility Transport Commission (Grand Canyon
21 Visibility Transport Commission, 1996), the National Research Council (National Research
22 Council, 1993), the National Acid Precipitation Assessment Program (Trijonis et al., 1991), and
23 the U.S. Environmental Protection Agency (1995a).
24
25 4.3.2 Factors Affecting Atmospheric Visibility
26 4.3.2.1 Anthropogenic Pollutants
27' Visibility impairment may be connected to air pollutant properties, including size
28 distribution, aerosol chemical composition, and relative humidity, hi the United States, visibility
29 impairment is caused by sulfate and nitrate particles in the 0.1- to 1.0-micron (ju.m) range, and
30 organic aerosols, carbon soot, and crustal dust. Generally, sulfates are responsible for most of
March 2001
4-85
DRAFT-DO NOT QUOTE OR CITE
-------
1 the visibility impairment in the United States, as measured by light extinction, accounting for
2 approximately two-thirds of the light extinction in the eastern United States. Sulfate
3 concentrations are higher in summer months than in the wintertime (Malm et al., 1994).
4 Exceptions to the sulfate-related effects on visibility include California, where the primary cause
5 of visibility effects is ambient nitrate, and Alaska, where visibility impairment is caused by fine
6 soil plus coarse mass (classified as coarse extinction) or organics, thought to be from natural
7 sources (Sisler and Cahill, 1993).
8
9 4.3.2.2 Human Vision
10 Human vision is one of the factors that affects the way an object is viewed. Vision is the
11 response to the electromagnetic radiation that enters the eye between wavelengths of 400 and
12 700 nm. The cones, a receptor cell in the retina, govern visibility interpretations.
13 The eye perceives the lightest and brightest object in a scene as white, and determines the
14 color of other objects by comparison! The ability of the eye to perceive contrasts, the degree of
15 color difference between the lightest and darkest object in a scene, changes in response to the
16 illumination and setting. The effects of illumination on visibility are discussed in the following
17 subsection. At increasing distances the brightness of a target or object will approach the
18 brightness of the horizon making the target indistinquishable from the horizon, hence, visual •
19 range.
20
21 4.3.2.3 Characteristics of the Atmosphere
22 The appearance of a distant object is determined by illumination of the sight path by the
23 direct rays of the sun, diffused skylight, light that has been reflected from the surface of the Earth
24 (path radiance or air light), and the light reflected from the object itself. Some of the light in the
25 sight path is absorbed or scattered towards the observer. The remaining light is absorbed or
26 scattered in other directions. The portion of scattered light from the object being viewed that
27 reaches the observer is the transmitted radiance. The radiance seen by the observer looking at a
28 distant object is the sum of the transmitted radiance and the path radiance. Figure 4-17
29 demonstrates light being absorbed and scattered by the atmosphere and a target object.
30 On a clear day when the sun is high in the sky, 80 to 90% of the visible solar radiation
31 reaches the surface of the Earth without being scattered or absorbed. Rayleigh scattering by
March 2001
4-86
DRAFT-DO NOT QUOTE OR CITE
-------
Figure 4-17. Light reflected from a target toward an observer. The intervening
atmosphere scatters a portion of this light out of the sight path and scatters
light from the sun into the sight path. Some particles and gases also absorb a
portion of the light from the target. The light scattered into the sight path
increases with distance from the target, whereas the light transmitted from
the target decreases with distance from the target. The visual range is the
closest distance between the target and the observer at which the transmitted
light no longer can be distinguished from the light scattered into the sight
path.
Source: Watson and Chow (1994).
1 gases is the major component of light extinction in relatively unpolluted areas. Mie scattering is
2 the scattering of all visible wavelengths equally (Shodor Education Foundation, Inc., 1996). It is
3 the attenuation of light in the atmosphere by scattering because of particles of a size comparable
4 to the wavelength of the incident light (National Acid Precipitation Assessment Program, 1991).
5 The term, multiple scattering, is used when light is scattered more than once in a turbid medium.
6 The great majority of light absorption by particles is caused by black carbonaceous particles,
7 assumed to be elemental carbon, that are products of incomplete combustion (Rosen et al., 1978;
8 Japar et al., 1986; Watson and Chow, 1994). Malm et al. (1996) suggested that organic carbon
March 2001
4-87
DRAFT-DO NOT QUOTE OR CITE
-------
1 also acts to scatter and absorb light. The estimated natural visibility for the east and west is 60 to
2 90 mi and up to 140 mi, respectively. Current visibility conditions range from 18 to 40 mi in the
3 rural east to 35 to 90 mi in the rural west (U.S. Environmental Protection Agency, 2000b).
4 At the surface, a variable fraction of the solar radiation is reflected back upwards, referred
5 to as surface reflectance or the albedo, illuminating the atmosphere from above and below. The
6 amount of solar radiation reflected depends on the color of the terrain. Dark-colored terrain
7 reflects less radiation than light-colored terrain.
8 Visibility within a sight path longer than approximately 100 km (60 mi) is affected by
9 changes in the properties of the atmosphere over the length of the sight path. The atmosphere
10 generally will not have uniform optical properties over distances greater than a few tens of
11 kilometers. Air quality within a sight path can affect the illumination of the sight path by
12 scattering or absorbing solar radiation before it reaches the Earth's surface. The light-extinction
13 coefficient, oext, is a measure of the fraction of light that is lost as it travels through the
14 atmosphere. The light-extinction coefficient is the sum of the light-scattering coefficient, oscat,
15 and the light-absorption coefficient, oabs, expressed in units of inverse lengths of the atmosphere
16 (megameters ; Mm'1). Typical extinction coefficients range from 0.01 km'1 (10 Mm'1) in
17 relatively clean air to ~ 1000 Mm'1 in highly polluted areas (Watson and Chow, 1994).
18 The light-extinction coefficient can be divided into coefficients for the following
19 components:
20 aag, light absorption by gases,
21 asg, light scattering by gases (Rayleigh scattering),
22 aap, light absorption by particles, and
23 Ojp, light scattering by particles.
24 Light scattering by particles, a^, can be divided to indicate scattering by coarse and fine particles:
25 asfp, light scattering by fine particles and oscp, light scattering by coarse particles.
26
27 4.3.3 Optical Properties of Particles
28 Visibility impairment is typically caused by fine particles. Fine particles are small enough
29 in comparison with the wavelength of visible light that their optical properties are nearly the
30 same as those of homogeneous spheres of the same volume and average index of refraction.
31 Accordingly, Mie equations (Mie, 1908; Kerker, 1969), for calculating the optical properties of
March 2001
4-88
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
homogeneous spheres also may be used to calculate the optical properties of fine particles with
the only uncertainties being in the fine particle size distribution and index of refraction (Richards,
1973). However, within the range of indices of refraction that most commonly occur in
atmospheric fine particles, the results of Mie calculations can be scaled to account for the effect
of the index of refraction. Coarse particles have less of an impact on visibility than do fine
particles. However, in most actual cases, the dominant uncertainty in using the optical properties
for coarse particles calculated with Mie equations is the uncertainty in the particle size
distribution. Uncertainties exist in the use of Mie calculations for calculating light absorption for
course particles because the refractive index of the particle is generally not known, and the
light-absorbing particles are not spherical in shape, making the calculated light absorption
efficiency factor less reliable. Also, light absorption by elemental carbon particles can be
reduced when the particle is covered by some chemical species (Dobbins et al, 1994).
Conversely, light absorption by carbon particles can be enhanced when coated with a
nonabsorbing refractive material such as ammonium sulfate (Fuller et al., 1999).
The output of the Mie calculations includes efficiency factors for extinction, Qext,
scattering, Qscat, and absorption, Qabs. The Qext, Qscat, and Qabs give the fraction of the incident
radiation falling on a circle with the same diameter as the particle that is either scattered or
absorbed. The light scattering or absorption efficiency factor (in units of m2/g) is the change in
the light scattering or absorption efficiencies per unit change in mass of the fine particle
constituent. The scattering and absorption efficiencies are determined by estimating the size
distribution of each particle. The results of the calculations for the light absorption efficiencies
contains significant uncertainties because the components of the index of refraction is generally
unknown and the light-absorbing particles are frequently chained agglomerates that do not have a
spherical shape. Multiplying the values of the light-scattering efficiency factor by the aerosol
volume concentration (in units of ^nrVcm3) gives the value of the light-scattering coefficient, asp,
(in units of Mm"1) for these particles.
Richards et al. (1991) reported a scattering efficiency for fine particles of ammonium
sulfate of 1.2 m2/g based on Mie calculations. The value was in agreement with the value
determined using the integrating nephelometer readings and the sulfate concentrations. Sulfate
scattering efficiencies have been reported to increase by a factor of two when the size distribution
went from 0.15 to 0.5 /u.m (McMurry et al., 1996). The calculated scattering efficiencies for
March 2001
4-89
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
sulfates were 4.1 m2/g for 100% mass removal and 3.4 and 5.6 m2/g for 25% mass removal.
Calculated scattering efficiencies for carbon particles ranged from 0.9 to 8.1 m2/g (Zhang et al.,
1994; Sisler and Malm, 2000; Sloane et al., 1991). A scattering efficiency of 1.0 and 0.6 m2/g
was reported for soil and coarse mass, respectively (Trijonis et al., 1987).
Scattering efficiencies of 2.4 and 3.1 m2/g for fine particles were reported by White et al.
(1994) and Waggoner et al. (1981), using an integrating nephelometer. Coarse particle scatter
less light, resulting in lower scattering efficiencies. Scattering efficiencies for coarse particles
ranged from 0.4 to 0.6 m2/g, based on integrating nephelometer readings (White et al., 1994;
Trijonis et al., 1987; White and Macias, 1990; Watson et al., 1991).
Absorption efficiencies for elemental carbon particles have been reported to range from
9 to 10 m2/g (Japar et al., 1984; Adams et al., 1989; Sloane et al., 1991). Based on a review of
the available data, Horvath (1993) reported that measured light absorption efficiencies for light
absorbing carbon ranges from 3.8 to 17 m2/g. According to Horvath (1993), calculated
absorption efficiencies are too high, ranging from 8 to 12 m2/g for monodispersed carbon
particles. Fuller et al. (1999) suggested that isolated spheres of light absorbing carbon have a
specific absorption of less than 10 m2/g. Light absorption by carbon particles only will be greater
than 10 m2/g if the particles are internally mixed and the occluding particles are sufficiently large.
Absorption values for graphitic and amorphous carbon spheres for primary sizes typical of diesel
soot are around 5 m2/g. Light absorption by aggravated carbon at visible wavelengths is
enhanced by no more than 30% and diminishes if encapsulated by a nonabsorbing aerosol.
Malm et al. (1996) suggested a combined scattering and absorption efficiency of 10 m2/g for
organic carbon.
Light-extinction budgets may be estimated using the light extinction efficiency and the
measured species concentrations. Light-extinction budgets estimate the fraction of the total light
extinction contributed by each chemical species in the sight path; however, the values obtained
will depend on the assumptions used (Malm et al., 1996; Lowenthal et al., 1995; Sisler and
Malm, 1994).
March 2001
4-90
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
4.3.4 Effect off Relative Humidity on Particle Size and Light-Scattering
Properties
Ambient particles contain water, even on relatively dry days. As the relative humidity
increases, the particle absorbs more water and increases in size and volume. It is the increase in
particle size and volume that acts to increase the light scattering properties of most particles
(Malm etal., 1996).
Ambient particles are a mixture of chemical compounds. The amount of increase in
particle size with increasing relative humidity is dependent on the particle composition (Zhang
et al., 1993). Available data indicate that particles containing ammonium salts are in a liquid
solution at relative humidities above 80%. Particles containing inorganic salts and acids are
more hygroscopic than particles composed primarily of organic species (Day et al., 1996;
McMurry and Stolzenburg, 1989; Saxena et al., 1995; Zhang et al., 1993, 1994; Sloane et al.,
1991). Particles containing the more hygroscopic salts and acid species deliquesce and undergo
changes in particle size in response to changes in relative humidity. For sulfate and nitrate
aerosols, light-scattering properties are similar for all mixture types and compositions, as long as
there is the same particle size distribution (Tang, 1997). Saxena et al. (1995) found that the
hygroscopic properties of inorganic particles can be altered positively or negatively in the
presence of organics. Based on limited data, nonurban organics were found to add to water
absorption by inorganics, whereas the urban organics diminished the absorption of water by
inorganic particles at relative humidities of 80 to 93%. Figure 4-18 demonstrates the humidity
effect on the scattering coefficients for several internally mixed (individual particles containing
one or more species) and externally mixed (species that co-exist as separate particles) aerosols.
The total scattering computed for an aerosol is relatively insensitive to whether the sample is
internally or externally mixed (Malm et al., 1997). Figure 4-19 demonstrates changes in the
scattering coefficient ratio, ospw/aspd, where aspw is the scattering coefficient under humid
conditions, and ospd is the scattering coefficient under dry conditions. The figure demonstrates
that light scattering is a function of relative humidity and chemical composition. The monitoring
data were generated as part of the Southeastern Aerosol and Visibility Study (Day et al., 2000).
There is also a relative humidity-related effect on the scattering efficiency. Ammonium
sulfate fine-particle-scattering efficiency varied from 1.5 to 4.5 m2/g, with low relative humidity
and median particle sizes ranging from 0.07 to 0.66 /^m (McMurry et al., 1996). Sloane et al.
March 2001
4-91
DRAFT-DO NOT QUOTE OR CITE
-------
0.1
Q.
b
0
O
o
D)
C
to
o
CO
0.001-
30
(NH4)2S04
Na2SO4
o Internal Mixture
External Mixture
(0.3 |jm, 1.5) j
(0.6 |jm, 2.0) ../a
T"
40
50
60 70
%RH
T"
80
90
100
Figure 4-18. Humidity effect on scattering coefficients computed for internal and external
mixtures of the mixed-salt aerosol: Na2SO4 (x2 = 0.5)-(NH4)2 SO4 (x3 = 0.5),
for two dry-salt particle size distributions, where x is the mass fraction of the
dry solutes. Particle size distributions are stated in the parenthesis.
Source: Tang (1997).
March 2001
4-92
DRAFT-DO NOT QUOTE OR CITE
-------
4-
I
S.
o
f •
S 2'
O)
0)
1 1.
0
A Day 207 (32.6% Sol. Inorg.; 25.8% Org.; 41.6% Soil)
Day 211 (42.7% Sol. Inorg.; 51.7% Org.; 5.6% Soil)
o Day 224 (63.5% Sol. Inorg.; 32.0% Org.; 4.5% Soil)
10 20 30 40 50 60 70
Relative Humidity (%)
80 90
100
Figure 4-19. Scattering ratios, ospw/ospd, for different chemical compositions as a function
of relative humidity.
Source: Day et al. (2000).
1
2
3
4
5
6
7
8
(1991) reported scattering efficiencies of 7.1 to 8.2 m2/g for sulfate at 74% relative humidity and
2.1 to 2.9 m2/g at 38% relative humidity. Average dry scattering efficiencies for sulfate ranged
from 2.03 to 2.23 m2/g for two western sites and one eastern site (Malm and Pitchford, 1997).
The dry scattering efficiency increased with increasing particle size. Dry specific scattering
efficiencies of 3 m2/g were reported for sulfates and nitrates (Sisler and Malm, 2000). Omar
et al. (1999) reported a calculated scattering efficiency range of 1.23 m2/g for sulfate when the
relative humidity was <63% to 5.78 m2/g when the relative humidity was >75%. The calculated
scattering efficiencies for organic carbon ranged from 3.81 m2/g when the relative humidity was
March 2001
4-93
DRAFT-DO NOT QUOTE OR CITE
-------
1 <63% to 6.9 m2/g at relative humidities above 75% (Omar et al., 1999). A more detailed
2 discussion of the effects of relative humidity on the size distribution of ambient particles appears
3 in Chapter 2 of this document.
4
5 4.3.5 Measures of Visibility
6 4.3.5.1 Human Observations
7 The National Weather Service has in recent decades recorded hourly visibility readings at
8 all major airports in the United States based on human observations of the most distant targeted
9 object's perceivability. Human observation of visibility, although providing a historical record of
10 visibility readings in the United States, are dependent on the individual and the availability of a
11 target and generally are related poorly to air quality.
12
13 4.3.5.2 Light-Extinction Coefficient and Parameters Related to the Light-Extinction
14 Coefficient
15 The most frequently used indicator for visibility characterization for air quality is the
16 light-extinction coefficient because it is closely linked to air quality (U.S. Environmental
17 Protection Agency, 1996a). Various meteorological conditions (moisture and cloud cover) can
18 affect the light-extinction coefficient; however, these effects can be minimized (Husar et al.,
19 1994; Blandford, 1994; Mercer, 1994). The light-extinction coefficient can be measured directly
20 using a transmissometer (Molenar et al., 1990,1992) or can be estimated by measuring the
21 components of light extinction (scattering and absorption) and calculating the sum (Malm et al.,
22 1994).
23 The light-extinction coefficient is the quantitative measure of haziness, defined as
24 0«a= K/visual range, where K is the Koschmieder constant. The value of K is determined both
25 by the threshold sensitivity of the human eye and the initial contrast of the visible object against
26 the horizon sky.
27 The visual range may be calculated from the light-extinction coefficient using the
28 Koschmieder equation by assuming the atmosphere and the illumination over a sight path in the
29 daytime is uniform, and that the threshold contrast is 2% (Katsev and Zege, 1994; Koschmieder,
30 1924). These assumptions are, however, invalid for visual ranges greater than 100 km (U.S.
31 Environmental Protection Agency, 1996a). Visual range is an understandable, and for most
March 2001
4-94
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
purposes, an appropriate measure of the optical environment. It has the disadvantage of being
related inversely to aerosol concentration.
Visual Range = 3.
The deciview index is an atmospheric haze index that expresses uniform changes in
haziness in common increments from pristine conditions to extremely visibility impaired
environments. The deciview scale is linear with perceived visual changes, starting near zero for
a pristine atmosphere (particle-free) at a 1.8-km elevation, and increases with increasing
haziness. The deciview index may be calculated from the light-extinction coefficient for green
light. For consistency, a Rayleigh scattering value of 10 Mm"1 is used.
dv = 10 Iogj0 (a^ /10 Mm -1)
Under ideal conditions, a just noticeable change in the light-extinction coefficient should
represent a one or two deciview change in the deciview scale, about a 10 to 20% change in the
extinction coefficient. Any change in the deciview scale should have a change of similar
magnitude in the visual appearance of the scene in cases where the assumptions used to develop
the deciview scale are met (Pitchford and Malm, 1994; Sisler and Malm, 2000). Figure 4-20
illustrates the relationship of light extinction in Mn"1, deciview index, and visual range in
kilometers. Although the deciview is related to extinction, it is scaled in such a way that is
perceptually correct (Fox et al., 1999).
Extinction (Mm'1) _J°
Deciviews (dv)
20
30
40 50 70 100 200 300 400 500 700 1000
Visual Range (km)
I
0
I
I
7
I
I
11
I
I I
14 16
I I
I Illl
19 23
I Illl
I
30
I
1
34
1
1
37
1
1
39
1
1 Illl
42 46
1 Illl
200
130 100 80 60 40
20
13
Figure 4-20. Comparison of extinction (Mn"1) and visual range (km).
Source: Fox et al. (1999).
March 2001 4-95 DRAFT-DO NOT QUOTE OR CITE
-------
1 Figures 4-2 la,b illustrate a change in deciview scale based on reconstructed extinction
2 coefficients for the Great Plains Region (Badlands) using data from the Interagency Monitoring
3 of Protected Visual Environments Network (IMPROVE). Details about the IMPROVE network
4 appears in Section 4.5.6. The data are sorted by year into three groups based on the cumulative
5 frequency of occurrence of PM2 5: best visibility days (1 Oth percentile), median (50th percentile),
6 and worst visibility days (90th percentile) (Sisler and Malm, 2000).
7 Richards (1999) suggests that the deciview index may not be a good tool for measuring
8 visibility impairment in areas restricted by boundaries. The deciview index is, however, suitable
9 for measuring visibility conditions over a broad geographic region, which is consistent with the
10 definition of regional haze, uniform haze caused by pollutant sources over broad areas (U.S.
11 Environmental Protection Agency, 1999).
12
13 4.3.5.3 Light-Scattering Coefficient
14 Light-scattering by particles has been reported to account for 68 to 86% of the total
15 extinction coefficient in several cities in California (Eldering et al., 1994). The light-scattering
16 coefficient is closely linked to fine particle concentrations, making it a good tool for determining
17 small particle-related effects on visibility. When the light-scattering coefficient is increased,
18 visibility is impaired because the transmitted radiance is decreased and the path radiance is
19 increased. (See discussion in the previous sections on transmitted radiance and path radiance.)
20 The light-scattering coefficient can be measured directly with an open and enclosed integrating
21 nephelometer and a forward scatter visibility monitor (Molenar et al., 1992; National Oceanic
22 and Atmospheric Administration, 1992). The light-scattering coefficient also may be calculated
23 using analytical approximations of the particle size distributions, log normal size distributions, or
24 sectional particle size distributions. In the sectional approach, the size composition distribution
25 is represented by a set of particle size sections. The chemical composition of each size section is
26 assumed to be the same (Wu et al., 1996).
27
28 4.3.5.4 Fine Particulate Matter Concentrations
29 The influence of particles on visibility degradation is dependent on the particle
30 composition, solubility, and size (Pryor and Steyn, 1994). Fine particle species have been
31 classified into five major types: (1) sulfates, (2) nitrates, (3) organics, (4) light absorbing carbon,
March 2001
4-96
DRAFT-DO NOT QUOTE OR CITE
-------
Fine Mass PM
2.5
Q.
6-
5-
4-
3-
2-
1-
88 89 90
91 92 93 94
Sample Year
95 96
Visibility Impairment
88 89 90 91 92 93 94 95 96
Sample Year
•10th Percentile
• 50th Percentile
• 90th Percentile
Figure 4-21a,b. Plots of the 10th, 50th, and 90th percentile groups for PM25 and deciview
at the Badlands National Park. The sample year began in March of each
year.
Source: Sisler and Malm (2000).
March 2001
4-97
DRAFT-DO NOT QUOTE OR CITE
-------
1 and (5) soil (Malm et al., 1994). The coefficient of light-scattering by fine particles is primarily
2 responsible for visibility impairment making fine particle concentration a suitable indicator of.
3 particle related effects on visibility. Several studies have demonstrated a relationship between
4 the coefficient for light-scattering by particles, measured using an integrating nephelometer, and
5 fine particle concentrations (Dattner, 1995; Waggoner and Weiss, 1980; Waggoner et al., 1981;
6 White et al., 1994). Figure 4-22 demonstrates visual range based on particle concentrations and
7 extinction efficiencies for road dust and sulfate.
8
400
300 -
0)
D)
i 200
—
15
CO
100 -
• Sulfate A Road Dust
change in visibility based on 1 ug/m3
difference in concentration
—i —i 1 —i 1 1 i—
10 20 30 40 50 60 70
Concentration (pg/m3)
so
go
100
Figure 4-22. Reduction in visual range as a function of increasing fine (sulfate) and coarse
(dust) particle concentrations.
Source: Watson and Chow (1994).
1 4.3.5.5 Discoloration
2 Discoloration may be used as a quantitative measurement of atmospheric color changes in
3 urban hazes. Atmospheric color changes is a component of plume visibility models. The color
March 2001
4-98
DRAFT-DO NOT QUOTE OR CITE
-------
1 of haze will primarily depend on the scene used and human vision. For plume visibility, the
2 threshold for perception of color differences depend on the apparent width of the plume and is
3 greater for color patches separated by sharp edges. Methods for specifying the colors of hazes
4 include the CIE XYZ system of color matching, the Hunt94 color-appearance model, and the
5 visual colorimeter, VISUAL colorimeter for Atmospheric Research (Trijonis et al., 1991;
6 Mahadev and Henry, 1999).
7
8 4.3.6 Visibility Monitoring Methods and Networks
9 Visibility monitoring studies measure the properties of the atmosphere either at the sampler
10 inlets (point measurements), as is the case with air quality measurements, or by determining the
11 optical properties of a sight path through the atmosphere (path measurements). Instrumental
12 methods for measuring visibility are generally of three types: (1) direct measurement of light
13 extinction of a sight path using a transmissometer, (2) measurement of light scattering at one
14 location using an integrating nephelometer, and (3) measurement of ambient aerosol mass
15 concentration and composition (Mathai, 1995).
16 The largest instrumental visibility monitoring network in the United States is designed to
17 provide real-time data for runway visibility to aid in controlling airport operations.
18 An automated observing system, Automated Surface Observing System (ASOS), is being placed
19 at airports around the country. This monitoring network is sponsored by the National Weather
20 Service, the Federal Aviation Administration, and the Department of Defense. More than
21 500 airports are currently commissioned and an additional 500 are expected to come online in the
22 next few years.
23 The visibility sensor, instead of measuring how far one can see, measures the clarity of the
24 air using a forward scatter visibility meter. The forward scatter meter was found to correlate
25 fairly well with extinction coefficient measurements from the Optec Transmissometer. The
26 clarity is then converted to what would be perceived by the human eye using a value called
27 Sensor Equivalent Visibility (SEV). Values derived from the sensor are not affected by terrain,
28 location, buildings, trees, lights, or cloud layers near the surface. The sensor transmits an
29 average 1-min value for a 10-min period. The sensor only samples 0.75 ft of the atmosphere.
30 An algorithm processes the air passing through the sensor over the 10-min measurement period
31 to provide a generally accurate visibility measurement for within 2 to 3 mi of the site. Moisture,
March 2001
4-99
DRAFT-DO NOT QUOTE OR CITE
-------
1 dust, snow, rain, or particles in the light beam affect the amount of light scattered (National
2 Weather Service, 1998).
3 Visibility data from the ASOS network is typically reported in small increments, up to
4 10 mi, for the purposes of airport operations. However, beginning in 1998, the raw visibility
5 data, including light extinction measurements equaling to visual ranges exceeding 10 mi, have
6 been archived in databases available from the National Climatic Data Center. Data for visibility
7 at larger distances from ASOS sites are available at the sensors for only a short period of time.
8 The data can be directly downloaded from the site. The ASOS data may be useful for
9 characterizing visibility in urban and suburban areas across the country. It also may be used in
10 future analyses to better understand the effects of fine PM on visibility in non-class I areas.
11 The largest monitoring network that includes both visibility and air quality measurements is
12 the Interagency Monitoring of Protected Visual Environments (IMPROVE) network. The
13 IMPROVE network was formed as a collaborative effort between the EPA and federal land
14 management agencies (National Park Service, U.S. Forest Service, Bureau of Land Management,
15 and Fish and Wildlife Service) responsible for Class I areas and the land around them (National
16 Park Service, 1998; Malm et al., 1994; Sisler et al., 1993; U.S. Environmental Protection
17 Agency, 1995a; Eldred et al., 1997; Perry et al., 1997). The primary monitoring objectives of the
18 IMPROVE program are to establish visibility levels, identify anthropogenic sources of
19 impairment, document progress towards elimination of visibility impairment in protected areas
20 from anthropogenic sources, and promote the development of visibility monitoring equipment
21 and the collection of comparable visibility data (National Park Service, 1998; Evans and
22 Pitchford, 1991). Presently over 70 sites employ the IMPROVE program monitoring methods.
23 It is anticipated that an additional 80 sites will be added in 2000.
24 Table 4-7 contains PM25 monitoring data from 30 IMPROVE sites for the years 1988 to
25 1996. The data includes averaged PM25 mass and specific species contributions. The data are
26 divided into eastern and western regions. The eastern regions, in addition to Washington, DC,
27 include Acadia National Park and Appalachia and consist of data from Shenandoah and the
28 Great Smoky Mountains National Parks. The western regions include the Northern Great Plains,
29 West Texas, Sonora, the Colorado Plateau, Central Rockies, Cascade, Sierra Humbolt, West
30 Coast, Sierra Nevada, Southern California, and Alaska (Sisler and Malm, 2000).
31
March 2001
4-100
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-7. AVERAGED REGIONAL PM2.S MASS AND EXTINCTION
SUMMARIES FOR THE YEARS 1988 TO 1996"
Region
Alaska
Appalachia
Cascades
Colorado Plateau
Central Rockies
Coastal
Northeast
Northern Great Plains
Northern Rockies
Southern California
Sonora
Sierra Nevada
Sierra Humbolt
Washington, DC
West Texas
PM2.5
1.71
(11-9)
10.81
(97.6)
4.67
(50.6)
3.15
(17.3)
2.87
(15.8)
4.40
(43.5)
6.13
'(59.3)
4.26
(30.3)
5.15
(39.5)
8.64
(51.7)
4.09
(21.3)
4.40
(25.2)
2.67
(16.7)
16.90
(132.8)
5.11
(27.0)
Sulfate
0.55
(5.1)
6.53
(71.7)
1.30
(29.1)
1.06
(6.7)
0.80
(5.5)
1.35
(18.4)
3.32
(40.6)
1.61
(14.6)
0.98
(15.0)
1.45
(9.3)
1.52
(8.3)
0.96
(7.0)
0.52
(5.2)
7.91
(73.2)
2.13
(12.9)
Nitrate
Organics
0.06
(0.06)
0.60
(6-9)
0.23
(5.0)
0.21
(1.3)
0.18
(1.2)
0.90
(10.9)
0.40
(4.8)
0.51
(4.7)
0.31
(4.7)
3.53
(22.6)
0.24
(1.3)
0.47
(3.5)
0.16
(1.5)
2.16
(19.9)
0.25
(1.5)
Organics
0.77
(3.1)
2.73
(10.9)
2.51
(10.0)
1.08
(4.3)
1.11
(4.4)
1.65
(6.6)
1.84
(7.3)
1.35
(5.4)
2.88
(11.5)
2.29
(9.2)
1.28
(5.1)
2.16
(8.6)
1.36
(5.5)
4.44
(17.8)
1.29
(5.2)
Fine Soil
0.22
(1.0)
0.52
(4.3)
0.22
(4.1)
0.64
(1.7)
0.64
(1.4)
0.25
(2.5)
0.23
(3.4)
0.63
(1.6)
0.57
(4.1)
0.94
(4.2)
0.84
(2.0)
0.55
(2.6)
0.42
(2.0)
0.82
(15.6)
1.27
(1 7)
Elemental
Carbon
0.10
(2.2)
0.43
(3.8)
0.41
(2.3)
0.17
(3.3)
0.14
(3.2)
0.25
(5-1)
0.34
(3.0)
0.16
(4.0)
0.41
(4.1)
0.42
(6.3)
0.20
(4.6)
0.26
(3.5)
0.20
(2.5)
1.56
(6.3)
0.17
(5 7)
"Mass is in fj.g/m3. Extinction summaries in parenthesis are in Mm.
Adapted: Sisler and Malm (2000).
March 2001
4-101
DRAFT-DO NOT QUOTE OR CITE
-------
1 The U.S. Environmental Protection Agency is currently in the process of establishing a
2 national PM2.S monitoring network of approximately 1,700 monitors at over 1,100 sites. The
3 PM^s monitoring effort will be coordinated with visibility monitoring efforts currently in place,
4 such as IMPROVE, to maximize benefits of both programs. The monitoring network is expected
5 to be fully implemented by the end of 2000 or shortly thereafter (U.S. Environmental Protection
6 Agency, 1997b; U.S. Environmental Protection Agency, 2000b).
7
8 4.3.7 Visibility Modeling
9 There are several types of models available for the evaluation of pollution-related effects on
10 visibility. Plume visibility models and regional haze models are source models that simulate the
11 transport, dispersion, and transformation of chemical species in the atmosphere. Plume models
12 use the resulting air quality data to calculate the values of parameters related to human
13 perception, such as contrast and color differences. Regional haze models calculate aerosol
14 species concentrations and the light-extinction coefficient. Models for the photographic
15 representation of haze use air quality data as an input and perform the optical calculations
16 required to create images that represent the visual effects of the air quality.
17
18 4.3.7.1 Regional Haze
19 Regional haze models may be used to assess the impact of pollutant sources on an
20 identified area or region, in most cases identified class I wilderness areas, or to evaluate the
21 impact of new or existing air quality regulations. Light extinction by fine particles is used to
22 determine the effect of anthropogenic pollutants on regional visibility degradation (regional
23 haze). In the United States, these anthropogenic particles are composed primarily of sulfate
24 compounds, organic compounds, and, to a much lesser extent, nitrate compounds, with the
25 exception of California, where nitrates are the largest single contributor to light extinction. The
26 contribution to light extinction by these compounds will vary based on the particle composition
27 and size distribution. Once the particles are formed, then- size can change, resulting in a change
28 in their light extinction efficiency. Model calculations take into consideration the mass of the
29 particulate constituents and the relative humidity.
30 The model requirements for regional-scale, multiple-source haze models are nearly
31 identical to the model requirements for simulations of regional-scale, multiple-source
March 2001
4-102
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
fine-particle impacts. Hence, the Eulerian-based grid models currently under development to
support fine particle impact assessments will be relied on to provide a means for assessing
large-scale, multiple-source haze impacts.
Middleton (1997) described the findings of a Eulerian-based grid model, the Denver Air
Quality Model (DAQM). The DAQM is the principal component of the Brown Cloud H study
that is part of earlier work investigating visibility in Denver over the last 20 years. The DAQM
is derived from the Regional Acid Deposition Model (RADM) and includes aerosol processes,
meteorological modeling analysis, and visibility analysis procedures. The DAQM has been used
to determine the relationship between emissions and concentrations of fine and coarse particles
and all major gaseous pollutants under various emission scenarios and meteorological
conditions. The results of the study demonstrated an association between visibility and air
quality issues in the Colorado Front Range area.
Neff (1997), in his evaluation of the DAQM model, suggested that the meteorological
model does not address adequately mesoscale structures responsible for the initiation and
maintenance of the brown cloud episodes or cloud systems and surface moisture fluxes. Given
these model uncertainties, it was suggested that there may be errors in the quantification of
emissions and in the calculated optical extinction and scattering.
The Visibility Assessment Scoping Model (VASM) uses Monte Carlo techniques to
generate multiple realizations of daily concentrations of sulfates, nitrates, elemental carbon,
organic carbon, fine and coarse dust, and the relative humidity to determine particle effects on
regional haze. Species-specific light attenuation is calculated based on particle concentration and
relative humidity, producing short-term haze intensity or visual range information (Shannon
etal., 1997).
The Elastic Light Scattering and Interactive Efficiency (ELSIE) model was used by Omar
et al. (1999) to determine the species concentrations and to relate apportionment to the extinction
coefficient in an aerosol mixture. The model assumes the aerosol is an internal inhomogeneous
mixture of chemical species and size distributions. Model input parameters included the size
distributions, prevailing relative humidity, refractive indices of the constituents, percent
solubility of the aerosol components, and the growth function of the aerosol particles. The model
assumes that the particles grow with increasing relative humidity according to a predetermined
growth function.
March 2001
4-103
DRAFT-DO NOT QUOTE OR CITE
-------
1 Several source-oriented models have been developed to evaluate the effects of pollutants on
2 regional haze. The U.S. Environmental Protection Agency, in cooperation with the U.S. Forest
3 Service, the Fish and Wildlife Service, and the National Park Service (the Interagency
4 Workgroup for Air Quality Modeling), developed the MESOPUFFII system of assessing
5 regional haze impacts. The MESOPUFF E system uses the light extinction for sulfates and
6 nitrates for an estimated 3- to 24-h average concentration (U.S. Environmental Protection
7 Agency, 1995b). The CALPUFF modeling system can process mesoscale meteorological data
8 and address dispersive processes of a regional nature. Simulated long-range pollutant trajectories
9 have been compared successfully to results from a field study involving transport to 1000 km
10 downwind (U.S. Environmental Protection Agency, 1995c). However, Lagrangian puff
11 ' dispersion modeling involving transport of 200 km or more tend to underestimate the horizontal
12 extent of the dispersion, causing the surface concentration to be overestimated (Moran and
13 Pielke, 1994). Another source-oriented Lagrangian trajectory model capable of computing light
14 extinction and scattering and estimating visual range from gas phase and primary particle phase
15 air pollutant emissions directly from sources was reported by Eldering and Cass (1996). The
16 model is comprised of several modules that take into consideration particle size distribution and
17 chemical composition, the speciation of organic vapor emissions, atmospheric chemical
18 reactions, transport of condensible material between the gas and particle phase, fog chemistry,
19 dry deposition, and light scattering and absorption. The model is, however, not suitable for
20 predicting visibility over great distances through nonuniform hazes and for visualization of
21 pollutant effects of isolated major point source plumes. Single line Lagrangian trajectory models
22 cannot represent horizontal turbulent diffusion, the effects of wind shear, and advection by
23 turbulent wind components. Error in transport calculations have been reported of up to ± 50%
24 (Eldering and Cass, 1996).
25 Gray and Cass (1998) developed a lagrangian particle-in-cell model for predicting source
26 class contributions of fine particle total carbon and elemental carbon. The model simulates the
27 motion and deposition of pollutants in an air basin with varying meteorological conditions. The
28 model also takes into consideration the vertical mixing characteristics of pollutants in areas
29 located near the source. The model is useful in determining changes in long-term average
30 pollutant concentrations from implementing specific emission control measures.
March 2001
4-104
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
The Regional Particulate Model (RPM) simulates secondary fine particulate matter (PM25)
formation and long-range transport. The RPM is used with the Regional Acid Deposition Model
(RADM), a comprehensive acid rain model. Predictions from the RADM are used to simulate
the formation of sulfate and nitrate, ammonium particles, and secondary organic aerosols. The
external RADM includes particle physics from the RPM and operates at an 80- and 20-km
resolution. Additional work currently is being done that will incorporate the RADM/RPM and
external RADM models into a more comprehensive air quality modeling system,
Models-3/Community Multi-Scale Air Quality (CMAQ). This modeling system simulates the
processes involved in primary and secondary PM!0 and PM2 5 and ozone formation, regional haze,
acid deposition, and nutrient deposition. The modeling system includes a mesoscale
meteorological model, emission model, and a version of the CMAQ.
The Regulatory Modeling System for Aerosols and Deposition (REMSAD) also simulates
PM2 5 formation. The REMSAD was derived from the Urban Airshed Model Version V
(UAM-V) for primary and secondary PM2 5 and PM10 formation, and acid nutrient and toxic
deposition. The REMSAD system consists of a meteorological data preprocessor, the core
aerosol and toxic deposition model (ATOM), and postprocessing programs. The ATOM is a
three-dimensional Eulerian grid model designed to calculate the concentrations of both inert and
chemically reactive pollutants by simulating the physical and chemical processes in the
atmosphere that affect pollutant concentrations. The basis for the model is the atmospheric
diffusion or species continuity equation. This equation represents a mass balance in which all of
the relevant emissions, transport, diffusion, chemical reactions, and removal processes are
expressed in mathematical terms (Systems Applications International, Inc., 1998).
Zannetti et al. (1990, 1993) and Fox et al. (1997) described a semi-empirical model that
could be used to estimate the visibility impact on one region resulting from sulfur dioxide
emission controls in a different region. The model combined four different input parameters:
(1) chemical transport; (2) possible nonlinearity of pollutant chemical transformation; (3) sulfate
fraction of fine particulate matter, including the amount of water absorbed by the fine particles;
and (4) the fraction of light extinction caused by fine particles. The model uses physically
realistic concepts of atmospheric transport, chemical transformation, and physical effects.
However, actual data sets, mathematical constructs, or expert opinions also maybe used. Models
also have been developed that predict the downwind concentration of smoke particulate and other
March 2001
4-105
DRAFT-DO NOT QUOTE OR CITE
-------
1 combustion products from the burning of crude oil from accidental spills (McGrattan et al., 1995,
2 1996).
3
4 4.3.7.2 Plume Models
5 Several plume visibility models are currently available. Plume visibility models estimate
6 the value of optical parameters related to human perception, such as contrast and color
7 differences, and compare these values with perception thresholds to determine whether the plume
8 is likely to be perceptible under various simulated conditions (U.S. Environmental Protection
9 Agency, 1988; Larimer, 1988). An empirical algorithm, Probability of Detection Algorithm
10 (PROBDET), allows the prediction of the lower limit of plume contrast that can be detected
11 visually. The PROBDET can be used to estimate the detection level for plumes that fall within
12 the bounds defined by the full-length, oval, and circular plume stimuli (Ross et al., 1997).
13 A simplified dispersion model using a second-order turbulence closure scheme to account
14 for averaging time effects on the dispersion rate was described by Sykes and Gabruk (1997). The
15 lateral and vertical spread is estimated using a Gaussian plume framework. A simplified
16 representation of the turbulence spectrum is used to predict the reduced spread rate for short
17 averaging tunes.
18 Earlier plume models included PLUVUEI and H, used during the preparation of a permit
19 application to determine whether or not a proposed new facility would cause visibility
20 impairment in a Class I area (Larimer et al., 1978; Johnson et al., 1980; White et al., 1985; U.S.
21 Environmental Protection Agency, 1992). Seigneur et al. (1997) developed a plume visibility
22 model, the Reactive and Optics Model Emissions (ROME), that improves on the existing plume
23 visibility models. The model simulates the momentum and buoyancy forces of the plume rise,
24 the dispersion and chemistry, and condensation and evaporation of the aqueous phase.
25 A second-order closure algorithm is used to estimate instantaneous plume concentrations, or the
26 time-averaged plume concentration may be estimated using a first-order closure algorithm.
27 A comprehensive chemical kinetic mechanism simulates chemical transformation processes in
28 the gas, aqueous, and particle phases. Particle dynamics and chemical composition is based on
29 sectional representation of the particle size distribution. The model includes a radioactive
30 transfer module that provides optical properties using sectional particle size distributions.
31 Deposition velocities based on atmospheric stability, surface type, chemical type, and particle
March 2001
4-106
DRAFT-DO NOT QUOTE OR CITE
-------
1 size are derived using a resistance-based, dry deposition algorithm. The ROME can be used with
2 other models to estimate a stack plume opacity, the percentage of light intensity attenuated by the
3 plume near the stack after any condensed water has evaporated (Meng et al., 2000). When
4 compared with the PLUVUE II, the ROME, with the second-order dispersion algorithm, was
5 found to present a more accurate estimate of plume height, width, nitrogen oxide concentration,
6 nitrogen dioxide/nitrogen oxide ratio, and visibility. Error, bias, correlation coefficients, and
7 simulations were within a factor of two of that observed (Gabruk et al., 1999).
8
9 4.3.7.3 Photographs
10 Computer-generated photographs are sometimes used to illustrate the effects of pollution
11 on visibility. To begin, a photograph is taken on a very clean, cloud-free day to serve as the
12 initial scene image. As previously indicated, the appearance of an object is determined by the
13 path radiance and the transmitted radiance. To determine the transmitted radiance, an estimate of
14 the light-extinction coefficient from the photograph is used to determine the initial radiance for
15 each element in the scene. The transmitted radiance is equal to the initial radiance of the
16 element in the scene multiplied by the transmittance of the atmosphere in the sight path. Because
17 the path radiance changes over the distance of the sight path, the source function, the rate of
18 change over the distance of the sight path, also must be determined.
19 Eldering et al. (1996) proposed the use of a model that uses simulated photographs from
20 satellite and topographic images to evaluate the effect of atmospheric aerosols and gases on
21 visibility. Use of this model requires ground-based photography and size distribution and
22 chemical composition of atmospheric aerosols, NO2 concentration, temperature, and relative
23 humidity for a clear day, for comparison purposes. Light extinction and sky color are then
24 calculated based on differences in aerosol size distribution, NO2 concentration, temperature, and
25 relative humidity. The images created represent natural landscape elements.
26 Molenar et al. (1994) provides a discussion of existing visual air quality simulation
27 methods based on techniques under development for the past 20 years. The WinHaze visual air
28 quality modeling system is one tool that has been developed using techniques to simulate
29 changes in visibility due to changes in air quality.
30 One of the limitations in using photographic models for representation of haze is that haze
31 is assumed to be uniformly distributed throughout the scene and selected conditions are
March 2001
4-107
DRAFT-DO NOT QUOTE OR CITE
-------
1 idealized, so the full range of conditions that occur in a scene are not represented. Photographs
2 are also expensive to produce. More detailed information on the use of photographic
3 representation of haze may be found in the U.S. Environmental Protection Agency (1996b),
4 Trijonisetal. (1991), Molenaretal. (1994), and Elderingetal. (1993).
5
6 4.3.8 Trends in Visibility Impairment
7 Trends in visibility impairment or haziness, visual range, often are associated with fine
8 mass concentrations (^2.5 //g/m3). Observations of visual range, obtained by the National
9 Weather Service and available through the National Climatic Data Center of the National
10 Oceanic and Atmospheric Administration, provide one of the few truly long-term, daily records
11 of any parameter related to air pollution. After some manipulation, the visual range data can be
12 used as an indicator of fine mode particle pollution. The data reduction process and analyses of
13 resulting trends have been reported by Husar et al. (1994), Husar and Wilson (1993), and Husar
14 etal. (1981).
15 Generally, visibility impairment is greatest in the eastern United States and southern
16 California. Haziness in the southeastern United States is greatest in the humid summer months
17 because of its affinity to atmospheric water vapor, followed by the spring and fall, and winter.
18 Summer haziness in the southeastern United States has increased by approximately 80% since
19 the 1950s (Husar and Wilson, 1993) because of increased sulfate from increased SO2 emissions
20 (Husar et al., 1994). The resulting sulfate, considered to be ammonium sulfate, accounts for
21 40 to 70% of the fine particle mass (Husar and Wilson, 1993). Sulfate-related effects on
22 visibility in the southeast is a factor of 20 higher than the Great Basin area and 10 higher than the
23 desert southwest, central Rocky Mountains, and Sierra Mountains (Malm et al., 1994). For most
24 rural eastern sites, sulfates accounts for >60% of the annual average light extinction on the best
25 days and >75% of the light extinction on the worst days. A statistically significant increase in
26 summer sulfate concentrations was noted in two class I areas in the eastern United States
27 (Shenandoah and the Great Smoky Mountains) from 1982 to 1992 (Eldred et al., 1993; Cahill
28 et al., 1996). The increase was largest in the summer and decreased in the winter. The majority
29 of the southwest showed decreasing sulfur (Eldred et al., 1993; Eldred and Cahill, 1994). White
30 (1997) suggested that the increase in fine-particle sulfur may be the result of the measurement
31 method and not an upward trend in fine particle concentrations in those Class I areas. However,
March 2001 4-108 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Iyer et al. (2000), using the Spearman correlation of trend, reported an increased trend in hazy
days during the summer months in Shenandoah and the Great Smoky Mountains based on
monitoring data for the period 1979 to 1996 showing high sulfur concentrations.
Sulfates also may be a significant contributor to total light extinction in the rural western
United States, accounting for 30 to 40% of the total light extinction on the best days and 35 to
45% of the total light extinction on the haziest days. In several areas of the west, sulfates account
for over 50% of the annual average aerosol extinction (U.S. Environmental Protection Agency,
2000b).
Organics are the second largest contributor to light extinction in most areas in the United
States. Extinction caused by organic carbon is greatest in the Pacific Northwest, Oregon, Idaho,
and Montana, accounting for 40 to 45% of the total extinction. Organic carbon can contribute
between 20 to 30% to the total extinction in most of the western United States and 10 to 15% in
the remaining areas of the United States. Light absorption by carbon is relatively insignificant
but is highest in the Pacific Northwest (up to 15%) and in the eastern United States (up to 6%)
(Malm et al., 1994; U.S. Environmental Protection Agency, 2000b).
Some of the visibility impairment in northern California and Nevada, including Oregon,
southern Idaho and western Wyoming, results from coarse mass and soil, primarily considered
natural extinction. In some areas of the United States, extinction from coarse mass is almost
negligible because the overall extinction is so high. High dust concentrations from southern
California have contributed to regional haze in the Grand Canyon and other class I areas in the
southwestern United States (Vasconcelos et al., 1996). White et al. (1999) reported that some of
the worst haze near the Grand Canyon is associated with pollutant transport from southern
California and the subtropics.
Visibility impairment in southern California is primarily caused by light extinction by
nitrates. Nitrates contribute about 40% to the total light extinction in Southern California and
10 to 20% of the total extinction in other areas of the United States.
The average haze patterns across the continental United States, for five-season averages for
the years 1980 to 1985 and 1990 to 1995 are shown in Figure 4-23. Haze is indicated by the
75th percentile of the extinction coefficient that is calculated from the visual range, corrected to
60% relative humidity by the Koschmeider relationship.
March 2001
4-109
DRAFT-DO NOT QUOTE OR CITE
-------
o
o
o
r-l
a
ts
g
March 2001
4-110
DRAFT-DO NOT QUOTE OR CITE
-------
1 The trends graphs in Figure 4-24 for regions in the United States represent the 75th
2 percentile of the light extinction coefficient for the stations located within the designated region
3 over a 30-year period (1940 to 1990). The trends are presented for quarters 1 (winter) and
4 3 (summer). The northeastern United States exhibited an increase in haze during quarter 3
5 between 1960 and 1970 and a steady decline between 1973 (0.22 km'1) and 1992 (0.12 km'1).
6 In quarter 1 the haziness steadily declined from 0.15 to 0.10 km"1 in the 30-year period. The
7 Mid-Atlantic region, the Virginias and Carolinas, shows a strong increase in haziness in quarter 3
8 between 1960 and 1973, followed by a decline. The winter haze was virtually unchanged over
9 the 30-year period. The haziness over the Gulf states increased between 1960 and 1970 and
10 remained virtually unchanged since then. The central Midwest, including Missouri and
11 Arkansas, exhibit virtually no change during the winter season and a slight increase in the
12 summer (1960 to 1970). The upper Midwest shows an opposing trend for summer and winter.
13 Although summer haze has increased, mostly from 1960 to 1973, the winter haze has declined.
14 Based on PM2 5 concentrations and changes in the deciview scale, calculated from
15 reconstructed extinction coefficients, Sisler and Malm (2000) reported no significant
16 deterioration in air quality and visibility conditions at 30 IMPROVE network sites for the years
17 1988 to 1996. The sites were divided into eastern and western regions. Averaged PM2 5 mass
18 and extinction summaries for the sites appear in Table 4-7. The annual best visibility
19 (10th percentile) and median visibility days (50th percentile) are improving at approximately
20 70% of the sites. However, several sites are not showing steady improvements in either visibility
21 or PM2 5, particularly in the number of worst visibility days (90th percentile). The sites included
22 the Badlands, Big Bend, Crater Lake, Great Smoky Mountains, Mesa Verde, Shenandoah and
23 Yosemite National Parks, Chiricahua National Monument, and the District of Columbia.
24
25 4.3.9 Economics of Particulate Matter Visibility Effects
26 Given the evidence of potential economically significant effects of visibility impairtment,
27 economic analysis proceeds by quantifying in monetary terms the costs associated with different
28 ambient levels of PM. Where possible, direct economic valuation can take place using prices
29 that are determined in the marketplace. There are a variety of ways to estimate costs/benefits.
30 Avoided cost methods estimate the costs of pollution by using the expenditures that are made
31 necessary by pollution damage. For example, if ambient levels of particulate matter results in
March 2001
4-111
DRAFT-DO NOT QUOTE OR CITE
-------
e japeno
I
C5
03
5
I
I
in
C
O
•a
I
N fl
S3 4)
•2 "
b &
£ a.
s js
U *3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
increased frequency of building cleaning or repainting, then the appropriately calculated increase
in these costs is a reasonable estimate of true economic damage. Benefits associated with
reductions in the pollution levels then are represented by the avoided costs of these damages.
Estimating benefits for visibility is a more difficult and less precise exercise because the
effects are not valued in the marketplace. There are several methods that economists have
developed to estimate changes in environmental effects that are not valued in the marketplace
(Freeman, 1993). These include hedonic price analysis, stated preference models (including
contingent valuation, contingent choice, and contingent ranking), and travel cost models.
Hedonic price analysis works by analyzing the way that market prices change when an associated
environmental effect changes. Part of the economic costs imposed by the reduced visibility
caused by PM can be estimated by looking at the differences in sales price between otherwise
identical houses that have different degrees of visibility impairment.
The contingent valuation method (CVM) has been used to determine estimated value
changes in both visibility and ecosystem functions (Hanley and Spash, 1993; Chestnut, 1997).
The CVM determines pollutant-related effects by using carefully structured surveys to estimate
the amount of compensation equivalent to a given change in environmental quality or
equivalently, how much they would be willing to pay to obtain a given change in environmental
quality. There is an extensive scientific literature and body of practice on both this theory and
technique.
Other valuation methods include stated preference models, including contingent choice and
contingent ranking (also known as conjoint analysis), as well as travel cost models (Johnson and
Desvousges, 1997; Hanley and Spash, 1993), However, the primary methods used to date for
valuation of visibility have been the hedonic price and contingent valuation methods (Hanley and
Spash, 1993).
The effects of PM on visibility may differ widely between urban residential and
recreational areas. Separate estimates are needed to account for welfare changes associated with
improvements in visibility in class I areas. Chestnut and Dennis (1997) developed a method for
estimating the value to the public of visibility improvements in class I areas using the results of a
1990 cooperative agreement project jointly funded by the EPA and the National Park Service:
"Preservation Values For Visibility Protection at the National Parks." Using the contingent
valuation method, Chestnut and Davis calculated a household willingness to pay for visibility
March 2001
4-113
DRAFT-DO NOT QUOTE OR CITE
-------
1 improvements in class I areas, capturing both use and nonuse recreational values. This analysis
2 also accounts for geographic variations in the willingness to pay. The results indicate a
3 willingness to pay per deciview improvement in visibility of between $5 and $ 17 per household.
4
5
6 4.4 EFFECTS ON MATERIALS
7 Effects of air pollution on materials are related to both aesthetic appeal and physical
8 damage. Studies have demonstrated that particles, primarily consisting of carbonaceous
9 compounds, cause soiling of commonly used building materials and culturally important items,
10 such as statutes and works of art. Physical damage from the dry deposition of air pollutants, such
11 as PM (especially sulfates and nitrates) and SO2, and the absorption or adsorption of corrosive
12 agents on deposited particles also can result hi the acceleration of naturally occurring weathering
13 processes of man-made building and cultural materials.
14 In the atmosphere, PM may be "primary", existing in the same form in which it was
15 emitted, or "secondary", formed by the chemical reactions of free, absorbed, or dissolved gases.
16 The major constituents of atmospheric PM are sulfate, nitrate, ammonium, and hydrogen ions;
17 particle-bound water; elemental carbon; a great variety of organic compounds; and crustal
18 material. A substantial fraction of the fine particle mass, particularly during the warmer months,
19 is secondary sulfate and nitrate. Sulfates may be formed by the gas-phase conversion of SO2 to
20 H2SO4 by OH radicals and aqueous-phase reactions of SO2 with H2O2, O3, or O2. During the day,
21 NO2 may be converted to nitric acid (HNO3) by reacting with OH radicals. Nitrogen dioxide also
22 can be oxidized to HNO3 by a sequence of reactions initiated by O3. A more detailed discussion
23 of the atmospheric chemistry of PM appears in Chapter 2 of this document.
24 Limited new studies have been published that better define the role of air pollution in
25 materials damage. This section briefly summarizes information on particle and sulfur-containing
26 pollutants (formed by the chemical reactions of SO2 with other atmospheric pollutants) exposure-
27 related effects on materials addressed in the 1996 PM AQCD (U.S. Environmental Protection
28 Agency, 1996a) and presents relevant information published since completion of that document.
29 The effects of nitrates on manmade building materials and naturally occurring cultural materials
30 was discussed in the criteria document on nitrogen oxides (U.S. Environmental Protection
31 Agency, 1993).
March 2001
4-114
DRAFT-DO NOT QUOTE OR CITE
-------
1 4.4.1 Effects of Particles and Sulfur Dioxide on Man-Made Surfaces
2 4.4.1.1 Metals
3 Metals under go natural weathering processes in the absence of environmental pollutants.
4 The additive effect of pollutants on the natural weathering processes will depend on the nature of
5 the pollutant and the deposition rate (the uptake of a pollutant by the material's surface), and the
6 presence of moisture. The influence of the metal protective corrosion film, the presence of other
7 surface electrolytes, the orientation of the metal surface, the presence of surface moisture, and the
8 variability in the electrochemical reactions will also contribute to the affect of pollutant exposure
9 on metal surfaces.
10 Several studies demonstrate the importance of tune of surface wetness (caused by dew and
11 fog condensation and rain) on metals. Surface moisture facilitates the deposition of pollutants,
12 especially SO2, and promotes corrosive electrochemical reactions on metals (Haynie and Upham,
13 1974; Sydberger and Ericsson, 1977). Of critical importance is the formation of hygroscopic
14 salts on the metal that increases the time of surface wetness and, thereby, enhances the corrosion
15 process.
16 Pitchford and McMurry (1994) and Zhang et al. (1993) demonstrated particle size-related
17 effects of relative humidity. The effect of temperature on the rate of corrosion is complex.
18 Under normal temperature conditions, temperature would not have an affect on the rate of
19 corrosion. When the temperature decreases the relative humidity increases and the diffusivity
20 decreases. The corrosion rate decreases as the temperature approaches freezing because ice
21 prohibits the diffusion of SO2 to the metal surface and minimizes electrochemical processes
22 (Haynie, 1980; Biefer, 1981; Sereda, 1974).
23 The metal protective corrosion film (i.e., the rust layer on metal surfaces) provides some
24 protection against further corrosion. The effectiveness of the corrosion film in slowing down the
25 corrosion process is affected by the solubility of the corrosion layer, and the concentration and
26 deposition rate of pollutants. If the metal protective corrosion film is insoluble, it may add some
27 protection against acidic pollutants. An atmospheric corrosion model that considers the
28 formation and dissolution of the corrosion film on galvanized steel was proposed by Spence et al.
29 (1992). The model considers the effects of SO2, rain acidity, and the time of wetness on the rate
30 of corrosion. Although the model does not characterize specifically particle effects, the
31 contribution of particulate sulfate was considered in model development.
March 2001
4-115
DRAFT-DO NOT QUOTE OR CITE
-------
1 Whether suspended particles actually impact on the corrosion of metals is not clear.
2 Several studies suggest that suspended particles will promote the corrosion of metals (Goodwin
3 et al., 1969; Barton, 1958; Sanyal and Singhania, 1956; Baedecker et al., 1991); however, other
4 studies have not demonstrated a correlation between particle exposure and metal corrosion
5 (Mansfeld, 1980; Edney et al., 1989). Walton et al. (1982) suggested that catalytic species within
6 several species in fly ash promote the oxidation of SOX to a corrosive state. Still other
7 researchers indicate that the catalytic effect of particles is not significant, and that the corrosion
8 rate is dependent on the conductance of the thin-film surface electrolytes during periods of
9 wetness. Soluble particles likely increase the solution conductance (Skerry et al., 1988; Askey
10 etal., 1993).
11 The corrosion of most ferrous metals (iron, steel, and steel alloys) is increased by
12 increasing SO2 exposure. Steels are susceptible to corrosion when exposed to SO2 in the absence
13 of protective organic or metallic coatings. Studies on the corrosive effects of SO2 on steel
14 indicate that the rate of corrosion increases with increasing SO2 and is dependent on the
15 deposition rate of the SO2 (Baedecker et al., 1991; Butlin et al., 1992a). The corrosive effects of
16 SO2 on aluminum is exposure-dependent, but appears to be insignificant (Haynie, 1976; Fink
17 et al., 1971; Butlin et al., 1992a). The rate of formation of the patina on copper (protective
18 covering) can take as long as 5 years and is dependent on the SO2 concentration, deposition rate,
19 temperature, and relative humidity (Simpson and Horrobin, 1970). Further corrosion is
20 controlled by the availability of copper to react with deposited pollutants (Graedel et al., 1987).
21 Butlin et al. (1992a), Baedecker et al. (1991), and Cramer et al. (1989) reported an average
22 corrosion rate of 1 //in/year for copper; however, less than a third of the corrosion was attributed
23 to SO2 exposure, suggesting that the rate of patina formation was more dependent on factors
24 other than SO2. A recent report by Strandberg and Johansson (1997) showed relative humidity to
25 be the primary factor in copper corrosion and patina formation. The results of the studies on
26 particles and SO2 corrosion of metals are summarized in Table 4-8.
27
28 4.4.1.2 Painted Finishes
29 Exposure to air pollutants affect the durability of paint finishes by promoting discoloration,
30 chalking, loss of gloss, erosion, blistering, and peeling. Evidence exists that indicates particles
31 can damage painted finishes by serving as carriers for corrosive pollutants (Cowling and Roberts,
March 2001
4-116
DRAFT-DO NOT QUOTE OR CITE
-------
^ .
CB §
CQ U
JJ 0§
tl"^
Average corrosion rate ranged from 0.63 to
1 .33 //rn/y. The highest corrosion was noted i
most industrialized area. However, the corros
rates did not differ significant regardless of th
concentration, suggesting that SO2 exposure n
be the dominant factor in zinc corrosion.
0
illil
•1311
L. on -H --, „
Specimens exposed at 5 sites fo:
Average S02 concentrations ran
15 ±17 ppb (5.2 ±10.4 to 39.3
PM concentrations ranged from
Highest pollutant concentratiom
1 year exposure site.
u
B
N
Is
^ OS
W ^
-------
CO
H*j
^7
H
w
1
fy^
Q
O
Q
1
&
1
1
S
1
1
5*
CO
Q
S
QS
O
U
_y
•s
"S
o
oo
4
H
I
>,
43
CX
ob
2
3
Comments
_o
•3
1
3
VI
a
UJ
u
S
5\
o\
s~~/
"3
ts
•a
C4
ffl
8(5" K
M P t/3 t^
Average corrosion rate for 3- and 5-year exposure
was about 1 /an/year but the soluble portion was 1
than a third of that which could be contributed to
exposure. Dry deposition of S02 was not as
important in patina formation as wet deposition oi
S
=>
g
1
o
•§
Ov
^ — >
"3
t>
Ui
0
^
•§ el
^3
en
Majority of test sites showed a corrosion rate of
1 ± 0.2 //m/year. The corrosion rate was
1.48 /urn/year at the site receiving the most rainfal
The lowest corrosion rate, 0.66 //m/year, was
associated with low rainfall, low S02.
§
'S
ex
1
Sf
ON
^
"3
u
1.8
*5 £3
CQ is
4> C
U O
C/l 0
u
1
u
£^*
t3 O\
CO ^
s°§
11
•fc* *n
OO i-^
^>
S02 had no effect on copper when relative humidi
was «75%. Increasing relative humidity increase:
patina formation in presence of trace S02.
No S02-related effects were noted on copper
specimens exposed to high S02 regardless of the
percent relative humidity.
•2-C »
!!l
®5I
3 P —
Ifi
•§-••§
8* ts '
^ 'T-« _c
ai S O
I'll 7!
tt ""^ -O
*S ^
S r~- '-e
-------
1 1954) or by staining and pitting of the painted surfaces (Fochtman and Langer, 1957; Wolff et al.,
2 1990).
3 The erosion rate of oil-base house paint has been reported to be enhanced by exposure to
4 SO2 and high humidity. In a study by Spence et al. (1975), an erosion rate of 36.71 ±
5 8.03 yum/year was noted for oil-base house paint samples exposed to SO2 (78.6 Aig/m3), O3
6 (156.8 Aig/m3), and NO2 (94 Afg/m3) and low humidity (50%). The erosion rate increased with
7 increased SO2 and humidity. The authors concluded that SO2 and humidity accounted for 61% of
8 the erosion. Acrylic coil coating and vinyl coil coating shows less pollutant-related erosion.
9 Erosion rates range from 0.7 to 1.3 yum/year and 1.4 to 5.3 /zm/year, respectively. Similar
10 findings on SO2-related erosion of oil-base house paints and coil coatings have been reported by
11 other researchers (Davis et al., 1990; Yocom and Grappone, 1976; Yocom and Upham, 1977;
12 Campbell et al., 1974). Several studies suggest that the effect of SO2 is caused by its reaction
13 with extender pigments such as calcium carbonate and zinc oxide (Campbell et al., 1974; Xu and
14 Balik, 1989; Edney, 1989; Edney et al., 1988, 1989). However, Miller et al. (1992) suggested
15 that calcium carbonate acts to protect paint substrates. Another study indicated that exposure to
16 SO2 can increase the drying time of some paints by reacting with certain drying oils and will
17 compete with the auto-oxidative curing mechanism responsible for crosslinking the binder
18 (Holbrow, 1962).
19
20 4.4.1.3 Stone and Concrete
21 Numerous studies suggest that air pollutants can enhance the natural weathering processes
22 on building stone. The development of crusts on stone monuments have been attributed to the
23 interaction of the stone's surface with sulfur-containing pollutants, wet or dry deposition of
24 atmospheric particles, and dry deposition of gypsum particles from the atmosphere. Because of a
25 greater porosity and specific surface, mortars have a greater potential for reacting with
26 environmental pollutants (Zappia et al., 1998). Details on these studies are discussed hi
27 Table 4-9. The stones most susceptible to the deteriorating effects of sulfur-containing pollutants
28 are the calcareous stones (limestone, marble, and carbonated cement). Exposure-related damage
29 to building stones result from the formation of salts in the stone that are subsequently washed
30 away during rain events leaving the stone surface more susceptible to the effects of pollutants.
31 Dry deposition of sulfur-containing pollutants promotes the formation of gypsum on the stone's
March 2001
4-119
DRAFT-DO NOT QUOTE OR CITE
-------
i_i
.2
Exposure Co
C3
tj
S
£«s i o"
x S= ,2 S O
OQ
II
C o
o^
S 2
c •«
i s
tl
a
>
i
O
6 ts
C
-§ a
'
HI
|oal
<; oo o
1 s-si
March 2001
4-120
DRAFT-DO NOT QUOTE OR CITE
-------
E MATTER AND SULFUR DIOXIDE ON STOP
H
; 4-9 (cont'd). EFFECTS OF PARTICULA
w
1
H
>>
•s.
1
1
I
0
Exposure Conditions
ID
§
1
"cS
•s
IS
H O\
^^ ^^
In the absence of moisture, little reaction is seen. SO2 is
oxidized to sulfates in the presence of moisture. The effect
is enhanced in the presence of O3. Massangis Jaune Roche
limestone was the least affected by the pollutant exposure.
Crust lined pores of specimens exposed to SO2.
Samples exposed to S02, N02, and NO at 10 ppmv,
both with and without 03 and under dry (coming to
equilibrium with the 84% RH) or wetted with
C02-equilibrated deionized water conditions.
Exposure was for 30 days.
u
I g g2
to 3 B u
§ i 3 IB
g 1— > CO M
i^ '§>.§ 5! .2
| |||J
«S-S«§ll
o^
OS
OS
CO
•>"
Significant amounts of gypsum were noted on the Portland
stone. Sheltered stones also showed soiling by
carbonaceous particles and other combustion products.
Etch holes and deep etching was noted in some of the
exposed unsheltered samples.
Samples exposed for 2 mo under both sheltered and
unsheltered conditions. Mean daily atmospheric S02
concentration was 68.7 /ug/m3 and several heavy
rainfalls.
!
CO
S3 3s
T3
£ §
§|
*
j_.
'I
O s~**
•§ Os
"a °*
oo C-
Exposure to particles from combustion processes enhanced
sulfation of calcareous materials by S02 because of metal
content of particles.
Sample exposed in laboratory to 3 ppm SO2 and 95%
RH at 25 °C for 150 days. Samples were coated with
three carbonaceous particle samples from combustion
sources, and with activated carbon and graphite.
JU
1 8 8
ail '
UEHft.
a
•§•
Op OS
2 os
S ^j
Carrara marble found to be more reactive with S02 than
Georgia marble possibly because of the compactness of the
Georgia marble. Greater effects noted when samples were
also exposed to N02.
Samples exposed in sheltered ambient environment
for 6, 12, or 20 mo.
0 3H
1 "i
5* -5
a u
U 0
•— '
t>
!| K1
oi O
Pollutant exposed samples showed increased weight gain
over that expected from natural weathering processes.
There was a blackening of stone samples exposed to
carbonaceous rich particulate matter.
Samples exposed for 6 mo (cold and hot conditions)
in ambient environment. PM concentrations ranged
from 57.3 to 1 16.7 yug/m3 (site 1) and 88 to
189.8 /ug/m3 (site 2). Some exposures also were
associated with high SO2, NO, and N02.
J2
i
1
u
s /~s
la
vG
•g a
35 t3
Exposure to fly-ash did not enhance oxidation of SO2 to
sulfates. Mineral oxides in fly ash contributed to sulphation
ofCaC03.
Samples artificially exposed to fly-ash containing
1309.3 Mg/m3 SO2 (0.5 ppm), at 95% RH and 25 °C
for 81 or 140 days. Fly-ash samples from five
different sources were used in study.
I
"w
OH O *^
j | 8
J-§l
•=; s ftn
March 2001
4-121
DRAFT-DO NOT QUOTE OR CITE
-------
04
CTS OF P
E
a
o
u
5
6d
ffi
H
J?
1
3
S
Comments
CO
asure Condition
g-
u
g
I
*C3
'I.?
*!
i)
ji "E.
^O . *. "*"* g
CO *+-» C3
CO c; O co
'•C .S o **3
C O *O '| i
«
O 42 13
oo "E-3
tlj
Q. OT CO
,856Aig/m3(3p;
60, or 90 days;
every 7 days to
oSS
* V. "
Igl
c? <_) u
P<° S3
Isl
fl*
C3 ^^ *^
00 K f
Ib
1 §1
1 c E
g -2 'c
3 (S o
u
o
JO
CO
g S
o eg
M =^ S
^•3-
B '-g .g «*
§• g_ i £*
g §-33 o
cfc g (D
on M« co !2
S "S S §
P Ji « o
Jl^JJ
^3 u 8 S3
"s i-i
iln>
g .g> g)
HI
- £
T3
S
« J§
n
fo
a
l
to
xp
Samples from structures exposed for varying periods of Black layers were found to be primarily comprised of iron Nord and
sited time under ambient air conditions. Samples selected compounds, quartz, silicate, soot, and dirt. Ericsson (1993)
because of black layer on surface.
ented
Limestone
5
§
3
B
sandstone
Calcite-cei
sandstone
Granite
Brick
Samples of ancient grey crust formed between 1 1 80 Crust samples contained calcite, soil dust, carbonaceous Ausset et al.
and 1636 on the Church of Saint Trophime in Arks and particles, and gypsum crystals. (1998)
formed between 1 530 and 1 1 87 on the Palazz
d'Accursio in Bolonga.
ble Samples of the stones and mortars were representative Mortars were more reactive than the stones. Of the Zappia et al.
narble of those used in the past and currently for new mortars, cement and pozzolan mortar were more reactive (1998)
tone construction and restorations. Samples were exposed than the lime mortar. Carrara marble was the least
lestone for 6, 12, and 24 mo under ambient conditions in reactive of the stones. The maximum amount of
r Milan. degradation was found in areas sheltered from rain.
Limestone
Sandstone
ca
1
U
1—
Travertine
W5
1
c
Portland li
CO
Lime mort
1 |
C o
Pozzolan t
Cement m
March 2001
4-122
DRAFT-DO NOT QUOTE OR CITE
-------
rV]
^
O
H
O3
'j^.
LTE MATTER AND SULFUR DIOXIDE O
^
ja
(j
6
^
c^
te
O
«3
H
O
w
is
w
•
^^s
•^_)
a
o
9s
4
2
H
1
_o
1
n
Comments
CA
g
•s
••3
§
u
E
3
1
w
I
CO
"ca
ts
'S
o /*"••*
•S o^
co C-
S
3
. Exposure to environmental pollutants caused the
formation of two separate layers on the mortar: an o
thin surface black crust composed of gypsum and
carbonaceous particles and the inner composed of
products from the dissolution and sulphation of the
carbonate matrix in the mortar.
ts
O
fe
£
'5
2
g
N
e
g
1
S
g
^
C3
2)
Cl_,
O
*p,
CO
1
e
1
3
"w
t3
.2 »
S'o
OQ Ci
t*-,
o
'S 2
Gypsum main component of crust followed by
carbonaceous particles and iron oxides. Estimated r
crust formation was 2-5 ^m/year. Total amount of
gypsum formed over the lifetime of exposure was 5
13 mg/cm2, an estimated 0.2 mg/cm2/year.
1
-^
1 "S
S o
to OS
S 4i
° "3
o *^
3^"
1 o
coK
_o
to
03
1
U
March 2001
4-123
DRAFT-DO NOT QUOTE OR CITE
-------
1 surface. Gypsum is a gray to black crusty material comprised mainly of calcium sulfate
2 dihydrate from the reaction of calcium carbonate (calcite) in the stone with atmospheric SO2 and
3 moisture (relative humidities exceeding 65%). Approximately 99% of the sulfur in gypsum is
4 sulfate because of the sulphation process caused by the deposition of SO2 aerosol. Sulphites also
5 are present in the gypsum layer as an intermediate product (Sabbioni et al., 19961; Ghedini et al.,
6 2000; Gobbi et al., 1998; Zappia et al., 1998). Gypsum is more soluble than calcite and is known
7 to form on limestone, sandstones, and marble when exposed to SO2. Gypsum also has been
8 reported to form on granite stone by replacing silicate minerals with calcite (Schiavon et al.,
9 1995). Gypsum occupies a larger volume than the original stone, causing the stone's surface to
10 become cracked and pitted. The rough surface serves as a site for deposition of airborne
11 particles.
12 The dark colored gypsum is caused by surface deposition of carbonaceous particles
13 (noncarbonate carbon) from combustion processes occurring in the area (Sabbioni, 1995;
14 Saiz-Jimenez, 1993; Ausset et al., 1998), trace metals contained in the stone, dust, and numerous
15 other anthropogenic pollutants. After analyzing damaged layers of several stone monuments,
16 Zappia et al. (1993) found that the dark-colored damaged surfaces contained 70% gypsum and
17 20% noncarbonate carbon. The lighter colored damaged layers were exposed to rain and
18 contained 1 % gypsum and 4% noncarbonate carbon. It is assumed that rain removes reaction
19 products, permitting further pollutant attack of the stone monument, and likely redeposits some
20 of the reaction products at rain runoffs sites on the stone. Following sulfur compounds, carbon
21 was reported to be the next highest element in dark crust on historical monuments in Rome.
22 Elemental carbon and organic carbon accounted for 8 and 39% of the total carbon in the black
23 crust samples. The highest percentage of carbon, carbonate carbon, was caused by the carbonate
24 matrix in the stones. The high ratio of organic carbon to elemental carbon indicates the presence
25 of a carbon source other than combustion processes (Ghedini et al., 2000). Cooke and Gibbs
26 (1994) suggested that stones damaged during times of higher ambient pollution exposure likely
27 would continue to exhibit a higher rate of decay, termed the "memory effect", than newer stones
28 exposed under lower pollution conditions. Increased stone damage also has been associated with
29 the presence of sulfur oxidizing bacteria and fungi on stone surfaces (Garcia-Valles et al., 1998;
30 Young, 1996; Saiz-Jimenez, 1993; Diakumaku et al., 1995).
March 2001
4-124
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Dissolution of gypsum on the stone's surface initiates structural changes in the crust layer.
Garica-Valles et al. (1998) proposed a double mechanism; the dissolution of the gypsum, in the
presence of sufficient moisture, followed by recrystallization inside fissures or pores. In the
event of limited moisture, the gypsum in dissolved and recrystallizes at its original location.
According to the authors, this would explain the gypsum-rich crustal materials on stone surfaces
sheltered from precipitation.
Moisture was found to be the dominant factor in stone deterioration for several sandstones
(Petuskey et al., 1995). Dolske (1995) reported that the deteriorative effects of sulfur-containing
rain events, sulfates, and SO2 on marble were largely dependent on the shape of the monument or
structure rather than the type of marble. The author attributed the increased fluid turbulence over
a nonflat vertical surface versus a flat surface to the increased erosion. Sulfur-containing
particles also have been reported to enhance the reactivity of Carrara marble and Travertine and
Trani stone to SO2 (Sabbioni et al., 1992). Particles with the highest carbon content had the
lowest reactivity.
The rate of stone deterioration is determined by the pollutant and the pollutant
concentration, the stone's permeability and moisture content, and the pollutant deposition
velocity. Dry deposition of SO2 between rain events has been reported to be a major causative
factor in pollutant-related erosion of calcareous stones (Baedecker et al., 1991; Dolske, 1995;
Cooke and Gibbs, 1994; Schuster et al., 1994; Hamilton et al., 1995; Webb et al., 1992). Sulfur
dioxide deposition increases with increasing relative humidity (Spiker et al., 1992), but the
pollutant deposition velocity is dependent on the stone type (Wittenburg and Dannecker, 1992),
the porosity of the stone, and the presence of hygroscopic contaminants.
Although it is clear from the available information that gaseous pollutants, in particular dry
deposition of SO2 will promote the decay of some types of stones under the specific conditions,
carboneous particles (noncarbonate carbon) may help to promote the decay process by aiding in
the transformation of SO2 to a more acidic species (Del Monte and Vittori, 1985). Several
authors have reported enhanced sulfation of calcareous material by SO2 in the presence of
particles containing metal oxides (Sabbioni et al., 1996; Hutchinson et al., 1992).
March 2001
4-125
DRAFT-DO NOT QUOTE OR CITE
-------
1 4.4.2 Soiling and Discoloration of Man-Made Surfaces
2 Ambient particles can cause soiling of man-made surfaces. Soiling has been defined as the
3 deposition of particles of less than 10 //m on surfaces by impingement. Soiling generally is
4 considered an optical effect, that is, soiling changes the reflectance from opaque materials and
5 reduces the transmissions of light through transparent materials. Soiling can represent a
6 significant detrimental effect requiring increased frequency of cleaning of glass windows and
7 concrete structures, washing and repainting of structures, and, in some cases, reduction in the
8 useful life of the object. Particles, in particular carbon, also may help catalyze chemical reactions
9 that result in the deterioration of materials during exposure.
10 It is difficult to determine the accumulated particle levels that cause an increase in soiling;
11 however, soiling is dependent on the particle concentration in the ambient environment, particle
12 size distribution, and the deposition rate and the horizontal or vertical orientation and texture of
13 the surface being exposed (Haynie, 1986). The chemical composition and morphology of the
14 particles and the optical properties of the surface being soiled will determine the time at which
15 soiling is perceived (Nazaroff and Cass, 1991). Carey (1959) reported that the average observer
16 could observe a 0.2% surface coverage of black particles on a white background. A recent study
17 suggest that it would take a 12% surface coverage by black particles before there is 100%
18 accuracy in identifying soiling (Bellan et al., 2000). The rate at which an object is soiled
19 increases linearly with time; however, as the soiling level increases, the rate of soiling decreases.
20 The buildup of particles on a horizontal surface is counterbalanced by an equal and opposite
21 depletion process. The depletion process is based on the scouring and washing effect of wind
22 and rain (Schwar, 1998).
23
24 4.4.2.1 Stones and Concrete
25 Most of the research evaluating the effects of air pollutants on stone structures have
26 concentrated on gaseous pollutants. The deposition of the sulfur-containing pollutants are
27 associated with the formation of gypsum on the stone (see Section 4.4.1.3). The dark color of
28 gypsum is attributed to soiling by carbonaceous particles from nearby combustion processes.
29 A lighter gray colored crust is attributed to soil dust and metal deposits (Ausset et al., 1998;
30 Camuffo, 1995; Moropoulou et al., 1998). Realini et al. (1995) found the formation of a dark
31 gypsum layer and a loss of luminous reflection in Carrara marble structures exposed for 1 year
March 2001
4-126
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
under ambient air conditions. Dark areas of gypsum were found by McGee and Mossitti (1992)
on limestone and marble specimens exposed under ambient air conditions for several years. The
black layers of gypsum were located in areas shielded from rainfall. Particles of dirt were
concentrated around the edges of the gypsum formations. Lorusso et al. (1997) attributed the
need for frequent cleaning and restoration of historic monuments in Rome to exposure to total
suspended particulates. They also concluded that, based on a decrease in brightness (graying),
surfaces are soiled proportionately over time; however, graying is higher on horizontal surfaces
because of sedimented particles. Davidson et al. (2000) evaluated the effects of air pollution
exposure on a limestone structure on the University of Pittsburgh campus using estimated
average TSP levels in the 1930s and 1940s and actual values for the years 1957 to 1997.
Monitored levels of SO2 were available for the years 1980 to 1998. Based on the available data
on pollutant levels and photographs, it was thought that soiling began while the structure was
under construction. With decreasing levels of pollution, the soiled areas have been slowly
washed away, the process taking several decades, leaving a white, eroded surface. Studies
describing the effects of particles on stone surfaces are discussed in Table 4-9.
4.4.2.2 Household and Industrial Paints
Few studies are available that evaluate the soiling effects of particles on painted surfaces.
Particles composed of elemental carbon, tarry acids, and various other constituents are
responsible for soiling of structural painted surfaces. Coarse-mode particles (>2.5 //m) initially
contribute more soiling of horizontal and vertical painted surfaces than do fine-mode particles
(<2.5 /^m), but are more easily removed by rain (Haynie and Lemmons, 1990). The
accumulation of fine particles likely promotes remedial action (i.e., cleaning of the painted
surfaces). Coarse-mode particles are primarily responsible for soiling of horizontal surfaces.
Rain interacts with coarse particles, dissolving the particle and leaving stains on the painted
surface (Creighton et al., 1990; Haynie and Lemmons, 1990). Haynie and Lemmons (1990)
proposed empirical predictive equations for changes in surface reflectance of gloss-painted
surfaces that were exposed protected and unprotected from rain and oriented horizontally and
vertically.
Early studies by Parker (1955) and Spence and Haynie (1972) demonstrated an association
between particle exposure and increased frequency of cleaning of painted surfaces. Particle
March 2001
4-127
DRAFT-DO NOT QUOTE OR CITE
-------
1 exposures also caused physical damage to the painted surface (Parker, 1955). Unsheltered
2 painted surfaces are initially more soiled by particles than sheltered surfaces but the effect is
3 reduced by rain washing. Reflectivity is decreased more rapidly on glossy paint than on flat paint
4 (Haynie and Lemmons, 1990). However, surface chalking of the flat paint was reported during
5 the exposure. The chalking interfered with the reflectance measurements for particle soiling.
6 Particle composition measurements that were taken during exposure of the painted surfaces
7 indicated sulfates to be a large fraction of the fine mode and only a small fraction of the coarse
8 mode. Although no direct measurements were taken, fine mode particles likely also contained
9 large amounts of carbon and possibly nitrogen or hydrogen (Haynie and Lemmons, 1990).
10
11
12 4.5 EFFECTS OF ATMOSPHERIC PARTICIPATE MATTER ON
13 CLIMATE CHANGE PROCESSES AND THEIR POTENTIAL
14 HUMAN HEALTH AND ENVIRONMENTAL IMPACTS
15 Global climate change processes and their potential human health and environmental
16 impacts have been accorded extensive attention during the past several decades, and they still
17 continue to be of broad national and international concern. This is reflected by extensive
18 research and assessment efforts undertaken since the mid-1970s by U.S. Federal Government
19 Agencies (e.g., NOAA, EPA, CDC, etc.) or via U.S. Federal hiteragency programs (e.g., the U.S.
20 Global Climate Change Research Program [USGCRP]) and by analogous extensive research and
21 assessment efforts undertaken by numerous other national governments or international
22 collaborative activities (e.g., those coordinated by the Intergovernmental Panel on Climate
23 Change [IPCC], established in the 1980s under the joint auspices of the World Meteorological
24 Organization [WMO], and the United Nations Environment Programme [UNEP]).
25 Atmospheric particles play important roles in two key types of global climate change
26 processes or phenomena: (1) alterations in the amount of solar radiation in the ultraviolet range
27 (especially UV-B) penetrating through the Earth's atmosphere and reaching its surface, where it
28 can exert a variety of effects on human health, plant and animal biota, and other environmental
29 components; and (2) alterations in the amount of solar radiation in the visible range being
30 transmitted through Earth's atmosphere and either being reflected back into space or absorbed
31 (together with trapping of infrared radiation emitted by the Earth's surface by certain gases),
March 2001
4-128
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28'
29
30
31
which enhances heating of the Earth's surface and lower atmosphere (i.e., the widely-known
"greenhouse effect") and leads to consequent "global warming" impacts on human health and the
environment. Atmospheric particles also play a lesser role by absorbing infrared radiation
emitted by the Earth's surface.
The effects of atmospheric PM on the transmission of electromagnetic radiation emitted by
the sun at ultraviolet and visible wavelengths and by the earth at infrared wavelengths depend on
the radiative properties (extinction efficiency, single scattering albedo, and asymmetry
parameter) of the particles, which are, in turn, dependent on the size and shape of the particles,
the composition of the particles arid the distribution of components within individual particles.
In general, the radiative properties of particles are size and wavelength dependent. In addition,
the extinction cross-section tends to be at a maximum when the particle radius is similar to the
wavelength of the incident radiation. Thus, fine particles present mainly in the accumulation
mode would be expected to exert a greater influence on the transmission of electromagnetic
radiation than would coarse particles. The composition of particles can be crudely summarized
in terms of the broad classes identified in Chapter 6 of the 1996 PM AQCD and recapitulated in
Chapter 2 of this document (e.g., fine particles mainly consisting of nitrate, sulfate, mineral dust,
elemental carbon, organic carbon compounds [e.g., PAHs], and metals derived from high
temperature combustion or smelting processes). The major sources of these components are
shown in Table 2.1 of Chapter 2 in this document.
Knowledge of the factors controlling the transfer of solar radiation in the ultraviolet
spectral region is needed for assessing the potential biological and environmental impacts
associated with exposure to UV-B radiation (290 to 315 nm). Knowledge of the effects of PM
on the transfer of radiation in the visible and infrared spectral regions is needed for assessing the
relation between particles and global warming and its environmental and biological impacts.
Key information regarding important conceptual aspects and factors related to solar ultraviolet
radiation processes and effects is summarized first below and atmospheric PM roles noted,
followed by summarization of global wanning processes, their potential human health and
environmental impacts, and potential relationships to atmospheric PM.
March 2001
4-129
DRAFT-DO NOT QUOTE OR CITE
-------
1 4.5.1 Solar Ultraviolet Radiation Transmission Impacts on Human Health
2 and the Environment: Atmospheric Particulate Matter Effects
3 4.5.1.1 Bases for Concern Regarding Increased Ultraviolet Radiation Transmission
4 The transmission of solar UV-B radiation through the earth's atmosphere is controlled by
5 ozone, clouds, and particles. The depletion of stratospheric ozone caused by the release of
6 anthropogenically produced chlorine (Cl)-and bromine (Br)-containing compounds has resulted
7 in heightened concern over potentially serious increases in the amount of solar UV-B radiation
8 (SUVB) reaching the Earth's surface. SUVB is also responsible for initiating the production of
9 OH radicals that oxidize a wide variety of volatile organic compounds, some of which can
10 deplete stratospheric ozone (e.g., CH3C1, CH3Br), absorb terrestrial infrared radiation (e.g., CH4 ),
11 and contribute to photochemical smog formation (e.g., C2H4 , C5H8 ).
12 Increased penetration of SUVB to the Earth's surface as the result of stratospheric ozone
13 depletion continues to be of much concern because of projections of consequent increased
14 surface-level SUVB exposure and associated potential negative impacts on human health, plant
15 and animal biota, and man-made materials. Several summary overviews (Kripke, 1989; Grant,
16 1989; Kodama and Lee, 1993; Van der Leun et al., 1995,1998) of salient points related to
17 stratospheric ozone depletion processes and bases for concern provide a concise introduction to
18 the subject, as does Figure 4-25. As shown to the left in the figure, stratophospheric ozone
19 depletion results from: (a) anthropogenic production and associated emission into the lower
20 atmosphere of certain trace gases having long atmospheric residence times (e.g.,
21 chlorofluorocarbons [CFCs], carbon tetrachloride [CC14], and Halon 1211 [CF2C1 Br] and 1301
22 [CF3Br], which have atmospheric residence times of 75 to 100 years, 50 years, 25 years, and
23 110 years, respectively); (b) their tropospheric accumulation and gradual transport, over decades,
24 up to the stratosphere, where (c) photodissociation processes release Cl and Br, that (especially
25 under very cold subzero upper atmospheric conditions) catalyze ozone reduction; leading to
26 (d) stratospheric ozone depletion that is most marked over Antarctica during Southern
27 Hemisphere wintertime, to a less marked but still significant extent over the Arctic Polar Region
28 during Northern Hemisphere wintertime, and to a lesser extent over mid-latitude regions during
29 any season.
30 Given the long tune involved in transport of such gases to the stratosphere and their long
31 residence times there, any effects already seen on stratospheric ozone are likely caused by the
March 2001
4-130
DRAFT-DO NOT QUOTE OR CITE
-------
BASES FOR CONCERN ABOUT STRATOSPHERIC OZONE DEPLETION
DUE TO CFC's, HALONS, AND OTHER TRACE GASES
Stratospheric
Ozone Depletion
Cl, Br Catalyze
Ozone Reduction
Photodissociation
Releases Cl and Br
Slow Transport
to Stratosphere
Tropospheric
Accumulation
Air Emissions of
CFC's, Hajons, etc.
OZONE DEPLETION EFFECTS
[CFC's & O3 Column
Reorganization
Climate Changes:
Temp., Winds, eta
Increase of Air
Stagnation Periods
Accumulation of
Tropospheric O3
and Acid Aerosols
Increased UV-B Light
Penetration to Surface
Man's Production
of CFC's, Halons,
Other Trace Gasses
Environmental
Effects: Crop,
Forest Damage
Human
Health
Effects
Infectious
Diseases
Increased
Altered Bio-
geochemical
Cycling
UV-B Radiation Direct
Human Health Impacts
Natural Ecosystem
and Agriculture
Impacts
Skin Damage
(Sunburn)
Damage
to Eye
Skin Cancer
Premature
Skin Aging
Terrestrial
Ecosystem Shifts
Lower Crop Yields
AND
Cataracts
Incidence
Increased
Aquatic
Ecosystem Shifts
Less Plankton &
Seafood
Figure 4-25. Processes involved in stratospheric ozone depletion because of man's
production of CFCs, halons, and other trace gases are shown to the left. The
types of effects caused by stratospheric ozone depletion and consequent
increased UV-B penetration to the Earth's surface are hypothesized to include
both direct effects on human health (e.g., increased cancer rates, immune
suppression, etc.) and other terrestrial and aquatic ecological effects resulting
from increased UV-B alterations of biogeochemical cycles.
Source: Adapted from Grant (1989).
1 atmospheric loadings of trace gases from anthropogenic emissions several decades ago, and those
2 gases already in the atmosphere may continue to exert stratospheric ozone depletion effects well
3 into the 21st century. Shorter lived gases, such as CH3Br, also exert significant ozone depletion
4 effects.
March 2001
4-131
DRAFT-DO NOT QUOTE OR CITE
-------
1 The main types of effects hypothesized as likely to result from stratospheric ozone
2 depletion and consequent increased SUVB penetration through the Earth's atmosphere include
3 the following.
4 (1) Direct Human Health Effects, such as skin damage (sunburn), leading to more rapid aging
5 and increased incidence of skin cancer; ocular effects (retinal damage and increased cataract
6 formation possibly leading to blindness); and suppression of some immune system
7 components (contributing to skin cancer induction and spread to nonirradiated skin areas, as
8 well as possibly increasing susceptibility to certain infectious diseases or decreasing
9 effectiveness of vacinations).
10 (2) Agricultural/Ecological Effects, mediated largely through altered biogeochemical cycling
11 resulting in consequent damaging impacts on terrestrial plants (leading to possible reduced
12 yields of rice, other food crops, and commercially important trees, as well as to biodiversity
13 shifts in natural terrestrial ecosystems); and deleterious effects on aquatic life (including
14 reduced ocean zooplankton and phytoplankton, as important base components of marine
15 food-chains supporting the existence of commercially important, edible fish and other
16 seafood, as well as to other aquatic ecosystem shifts).
17 (3) Indirect Human Health and Ecological Effects, mediated through increased tropospheric
1 g ozone formation (and consequent exacerbation of surface-level, ozone-related health and
19 ecological impacts) and alterations in the concentrations of other important trace species,
20 most notably the hydroxyl radical and acidic aerosols.
21 (4) Other Types of Effects, such as faster rates of polymer weathering because of increased
22 UV-B radiation and other effects on man-made commercial materials and cultural artifacts,
23 secondary to climate change or exacerbation of air pollution problems.
24 Extensive qualitative and quantitative characterizations of stratospheric ozone depletion
25 processes and projections of their likely potential impacts on human health and the environment
26 have been the subjects of periodic (1988,1989,1991,1994, 1998) international assessments
27 carried out under WMO and UNEP auspices since the 1987 signing of the Montreal Protocol on
28 Substances that Deplete the Ozone Layer. The reader is referred for more detailed up-to-date
29 information to the two most recently completed international assessments of processes
30 contributing to stratospheric ozone depletion and the status of progress towards ameliorating the
31 problem (WMO, 1999) and revised qualitative and quantitative projections of likely consequent
March 2001
4-132
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
human health and environmental effects (UNEP, 1998). (See Appendices 4A and 4B for
synopses of key points abstracted from the executive summaries of these assessments).
Of considerable importance is the growing recognition, as reflected in these newer
assessments, of impacts of enhanced solar radiation on biogeochemical cycles (see, for example,
Zepp et al., 1998, and earlier discussions in this chapter [Sections 4.2.2.1 and 4.2.2.2]). As noted
in the Zepp et al. paper, the effects of UV-B radiation (both in magnitude and direction) on trace
gas (e.g., CO) emissions and mineral nutrient cycling are species specific and can affect a variety
of processes. These include, for example, changes in the chemical composition of living plant
tissue, photodegradation of dead plant matter (e.g., ground litter), release of CO from vegetation
previously charred by fire, changes in microbial decomposer communities, and effects on
nitrogen-fixing microorganisms and plants. Also, studies of natural acquatic ecosystems indicate
that organic matter is the primary determinant of UV-B penetration through water. Organic
matter changes, caused by enhanced UV-B penetration and augmented by acidification and
climate change, contribute to clarification of water and changes in light quality that broadly
impact the effects of UV-B on aquatic biogeochemical cycles. Enhanced UV-B levels have both
positive and negative impacts on aquatic ecosystem microbial activities that can affect nutrient
cycling and the uptake or release of greenhouse gases. Thus, there are emerging complex issues
regarding interactions and feedbacks between climate change and changes in terrestrial and
marine biogeochemical cycles because of increased UV-B penetration to the Earth's surface.
As noted in the above detailed assessments, since the signing of the Montreal Protocol,
much progress has been made in reducing emissions of ozone depleting gases, leading to
estimates of the maximum extent of stratospheric ozone depletion as likely having been reached
in the year 2000, to be followed by gradual lessening of the problem and its impacts during the
next half-century. However, the assessments also note that the modeled projections are subject
to considerable uncertainty. The role of atmospheric particles, discussed below, is one of
numerous salient factors complicating modeling efforts.
4.5.1.2 Airborne Particle Impacts on Atmospheric Ultraviolet Radiation Transmission
A given amount of ozone in the lower troposphere has been shown to absorb more solar
radiation than an equal amount of ozone in the stratosphere because of the increase in its
effective optical path produced by Rayleigh scattering in the lower atmosphere (Bruehl and
March 2001
4-133
DRAFT-DO NOT QUOTE OR CITE
-------
1 Crutzen, 1988). The effects of particles are more complex. The impact of particles on the SUVB
2 flux throughout the boundary layer are highly sensitive to the altitude of the particles and to their
3 single scattering albedo. Even the sign of the effect can reverse as the composition of the particle
4 mix changes from scattering to absorbing types (e.g., from sulfate to elemental carbon or PAHs)
5 (Dickerson et al., 1997). In addition, scattering by particles also may increase the effective
6 optical path of absorbing molecules, such as ozone, in the lower atmosphere.
7 The effects of particles present in the lower troposphere on the transmission of SUVB have
8 been examined both by field measurements and by radiative transfer model calculations. The
9 presence of particles in urban areas modifies the spectral distribution of solar irradiance at the
10 surface. Shorter wavelength radiation (i.e., in the ultraviolet) is attenuated more than visible
11 radiation (e.g., Peterson et al., 1978; Jacobson, 1999). Wenny et al. (1998) also found greater
12 attenuation of SUVB than SUVA (315 to 400 nm). However, this effect depends on the nature
13 of the specific particles involved and, therefore, is expected to depend strongly on location.
14 Lorente et al. (1994) observed an attenuation of SUVB ranging from 14 to 37%, for solar zenith
15 angles ranging from about 30° to about 60°, in the total (direct and diffuse) SUVB reaching the
16 surface hi Barcelona during cloudless conditions on very polluted days (aerosol scattering optical
17 depth at 500 nm, 0.46 s TSOO „„, 5 1.15) compared to days on which the turbidity of urban air was
18 similar to that for rural air (i;soo nm £ 0.23). Particle concentrations that can account for these
19 observations can be estimated roughly by combining Koschmeider's relation for expressing
20 visual range in terms of extinction coefficient with one for expressing the mass of PM2 5 particles
21 in terms of visual range (Stevens et al., 1984). By assuming a scale height (i.e., the height at
22 which the concentration of a substance falls off to 1/e of its value at the surface) of 1 km for
23 PM2 5, an upper limit of 30 fig/ m3 can be derived for the clear case and between 60 and
24 150 Aig/m3 for the polluted case. Estupinan et al. (1996) found that summertime haze under clear
25 sky conditions attenuates SUVB between 5 and 23% for a solar zenith angle of 34 °, compared to
26 a clear sky day in autumn. Mims (1996) measured a decrease in SUVB by about 80% downwind
27 of major biomass burning areas in Amazonia in 1995. This decrease in transmission
28 corresponded to optical depths at 340 nm ranging from three to four. Justus and Murphey (1994)
29 found that SUVB reaching the surface decreased by about 10% because of changes in aerosol
30 loading in Atlanta, GA, from 1980 to 1984. Also, higher particle levels in Germany (48 °N) may
March 2001
4-134
DRAFT-DO NOT QUOTE OR CITE
-------
1 be responsible for greater attenuation of SUVB than in New Zealand (Seckmeyer and McKenzie,
2 1992).
3 In a study of the effects of nonurban haze on SUVB transmission, Wenny et al. (1998)
4 derived a very simple regression relation between the measured aerosol optical depth at 312 nm,
5
6 ln( SUVB transmission at solar noon) = -0.1422 T312 „„ - 0.138, R2 = 0.90,
7
8 and the transmission of SUVB to the surface. In principle, values of T3I2 nm could be found from
9 knowledge of the aerosol optical properties and visual range values. Wenny et al. (1998) also
10 found that absorption by particles accounted for 7 to 25% of the total (scattering + absorption)
11 extinction. Relations such as the above one are strongly dependent on local conditions and
12 should not be used in other areas without knowledge of the differences in aerosol properties.
13 Although all of the above studies reinforce the idea that particles play a maj or role in modulating
14 the attenuation of SUVB, none included measurements of ambient PM concentrations, so direct
15 relations between PM levels and SUVB transmission could not be determined.
16 Liu et al. (1991) estimated, roughly, overall effects of increases of anthropogenic airborne
17 particles that have occurred since the beginning of the industrial revolution on atmospheric
18 transmission of SUVB. Based on (a) estimates of the reduction in visibility from about 95 km to
19 about 20 km over nonurban areas in the eastern United States and in Europe, (b) calculations of
20 optical properties of airborne particles found in rural areas to extrapolate the increase in
21 extinction at 550 to 310 nm, and (c) radiative transfer model calculations, Liu et al. concluded
22 that the amount of SUVB reaching Earth's the surface likely has decreased from 5 to 18% since
23 the beginning of the industrial revolution. This was attributed mainly to scattering of SUVB
24 back to space by sulfate containing particles. Radiative transfer model calculations have not
25 been done for urban particles.
26 Although aerosols are expected to decrease the flux of SUVB reaching the surface,
27 scattering by particles is expected to result in an increase in the actinic flux within and above the
28 aerosol layer. However, when the particles significantly absorb SUVB, a decrease in the actinic
29 flux is expected. Actinic flux is the radiant energy integrated over all directions at a given
30 wavelength incident on a point in the atmosphere, and is the quantity needed to calculate rates of
31 photolytic reactions in the atmosphere. Blackburn et al. (1992) measured attenuation of the
March 2001
4-135
DRAFT-DO NOT QUOTE OR CITE
-------
1 photolysis rate of ozone and found that aerosol optical depths near unity at 500 nm reduced
2 ozone photolysis rate by as much as a factor of two. Dickerson et al. (1997) showed that the
3 photolysis rate for NO2 , a key parameter for calculating the overall intensity of photochemical
4 activity, could be increased within and above a scattering aerosol layer extending from the
5 surface, although it would be decreased at the surface. This effect is qualitatively similar to what
6 is seen in clouds, where photolysis rates are increased in the upper layers of a cloud and above
7 the cloud (Madronich, 1987). For a simulation of an ozone episode that occurred during July
8 1995 in the Mid-Atlantic region, Dickerson et al. (1997) calculated ozone increases of up to
9 20 ppb compared to cases that did not include the radiative effects of particles in urban airshed
10 model (UAM-IV) simulations. In contrast, Jacobson (1998) found that particles may have
11 caused a 5 to 8% decrease in O3 levels during the Southern California Air Quality Study in 1987.
12 Absorption by organic compounds and nitrated inorganic compounds was hypothesized to
13 account for the reductions in UV radiation intensity.
14 The photolysis of ozone in the Hartley bands also leads to production of electronically
15 excited oxygen atoms, O('D) that then react with water vapor to form OH radicals. Thus,
16 enhanced photochemical production of ozone is accompanied by the scavenging of species
17 involved in greenhouse warming and stratospheric depletion. However, these effects may be
18 neutralized or even reversed by the presence of absorbing material in the particles. Any
19 evaluation of the effects of particles on photochemical activity therefore will depend on the
20 composition of the particles and also will be location-specific.
21 Also complicating any straightforward evaluation of UV-B penetration to specific areas of
22 the Earth's surface are the influences of clouds, as discussed by Erlick et al. (1998), Frederick
23 et al. (1998), and Soulen and Fredrick (1999). Varying estimations of atmospheric transmission
24 of UV and visible spectrum light are obtained for cloudy atmospheres, depending on presence of
25 aerosols and the extent of their external or internal mixing with cloud droplets. Even in
26 situations of very low atmospheric PM (e.g., over Antarctica), interannual variations in
27 cloudiness over specific areas can be as important as ozone levels in determining UV surface
28 irradiation, with net impacts varying from a month or season to another (Soulen and Fredrick,
29 1999).
30 Given the above considerations, quantitation of projected effects of variations in
31 atmospheric PM on human health or the environment because of particle impacts on transmission
March 2001
4-136
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
of solar UV-B would require location-specific evaluations, taking into account composition,
concentration, and internal structure of the particles; temporal variations in atmospheric mixing
heights and depths of layers containing the particles; and consequent impacts on surface-level
exposures of humans, ecosystem constituents, or man-made materials. The outcome of such
modeling effects would likely vary from location to location in terms of increased or decreased
surface level UV-B exposures because of location-specific changes in atmospheric PM
concentrations or composition. For example, to the extent that any location-specific scattering by
airborne PM were to affect the directional characteristics of UV radiation at ground level, and
thereby enhance radiation incident from low angles (Dickerson, 1997), the biological
effectiveness of resulting ground-level UV-B exposures could be enhanced. Airborne PM also
can reduce the ground-level ratio of photorepairing radiation (UV-A and short-wavelength
visible) to damaging UV-B radiation. Lastly, PM deposition is a major source of PAH in certain
freshwater lakes and coastal areas, and the adverse effects of solar UV are enhanced by uptake of
PAH by aquatic organisms. Thus, although airborne PM may, in general, tend to reduce ground-
level UV-B, its net effect in some locations may be to increase UV damage to certain aquatic and
terrestrial organisms, as discussed by Cullen and Neale (1997).
4.5.2 Global Warming Processes, Human Health and Environmental
Impacts, and Atmospheric Particle Roles
4.5.2.1 Bases for Concern Regarding Global Warming and Climate Change
Various trace gases emitted because of man's activities, including several noted above as
contributing to stratospheric ozone depletion, can act as "greenhouse gases" (GHG). That is, as
their tropospheric concentrations increase, they retard the escape of infrared radiation from the
earth's surface and thereby contribute to the trapping of heat near the surface (the "greenhouse
effect") and, ultimately, to consequent global warming and climate change. Much concern has
evolved,with regard to increases in the naturally very low concentrations in the atmosphere of
some of these gases, especially carbon dioxide (CO2), nitrous oxide (N2O), methane (CH4),
chloroflurocarbons (CFCs), and tropospheric ozone (O3).
Atmospheric processes involved in mediating global warming and its likely consequent
effects have been reviewed extensively previously (United Nations Environment Programme,
1986; World Meteorological Organization, 1988; U.S. Environmental Protection Agency, 1987;
March 2001
4-137
DRAFT-DO NOT QUOTE OR CITE
-------
1 IPCC, 1996, 1998; US GCRP, 2000) and more concisely summarized by others (e.g., Grant
2 1989; Patz et al., 2000a; Patz et al., 2000b). The main focus here is (a) to provide first a very
3 brief summary of key points regarding processes involved and types of effects projected as likely
4 to be associated with global warming and climate change and, then, (b) to undertake discussion
5 of salient considerations regarding potential impacts of atmospheric PM on such processes and
6 effects.
7 All of the above noted assessments and summaries emphasize that estimating likely future
8 global warming trends and associated climate change caused by greenhouse gases is extremely
9 complex, with modeling results being highly dependent on key assumptions about the rates of
10 future increases in various gases and numerous other factors (including particle effects).
11 Modeling of the magnitude of the warming directly associated with radiative forcing by
12 greenhouse gases (without feedback enhancement) projects temperature increases, for example,
13 of about 1.2 °C for a doubling of CO2; another 0.45 °C for a simultaneous doubling of N2O and
14 CH4; and an additional 0.15 °C from a uniform 1-ppb increase in atmospheric concentrations of
15 CFC-11 and CFC-12. Indirect effects (feedbacks) that likely would increase temperatures further
16 are expected to occur. Increased water vapor (trapping heat) and snow and ice melting (reducing
17 reflection of radiation back into space) are two examples of such feedback factors expected to
18 increase temperatures. However, major uncertainties exist with regard to feedbacks between
19 global wanning and clouds, which could either amplify or, perhaps, reduce a temperature rise.
20 Taking assumptions about rates of increase (or decrease) in GHG concentrations, consequent
21 initial warming effects, feedback effects, and accompanying uncertainties into account, numerous
22 modeling efforts have attempted to project likely future trends in global warming. Despite the
23 complexity and uncertainties inherent in such modeling efforts, all typically agree that some
24 global warming has occurred and will continue to occur during the coming decades, but the
25 ranges of quantitative estimates vary considerably depending on specific assumptions
26 incorporated into the models. Thus, for example, "low" scenarios assuming stabilization or
27 reductions in GHG emissions (resulting from implementation of the 1987 Montreal Protocol)
28 project lower temperature changes than other scenarios assuming higher rates of increase in GHG
29 emissions or differing feedback-effect patterns.
3 o Given the wide range of estimates of global warming trends and patterns of associated
31 climate change emerging from modeling efforts, the estimation of likely human health and
March 2001
4-138
DRAFT-DO NOT QUOTE OR CITE
-------
1 ecological effects associated with global warming on any quantitative basis is extremely difficult.
2 The onset of any notable global warming effect is also important, with various analyses
3 indicating that global temperatures for the past century have been rising (and now appear to be
4 beyond average levels within the range of variation seen with cycles of global warming or
5 cooling over the past several centuries before marked anthropogenic emissions of greenhouse
6 gases occurred). Also posing difficulties for the quantitative estimation of human health and
7 other effects are expected wide regional variations in temperature and climate characteristics
8 (e.g., rain and snowfall amounts) that may be projected reasonably to result from various global
9 warming trend scenarios. Lastly, it should be noted that, despite general warming trends in
10 long-term average temperatures, wide extremes in both high and low temperatures also are
11 expected to occur more frequently in some areas.
12 A Special Report of the IPCC Working Group II on Regional Impacts of Climate Change:
13 An Assessment of Vulnerabilities (IPCC, 1998) assesses global wanning processes and identifies
14 several types of vulnerabilities likely to occur because of climate change resulting from global
15 warming. Such general types of vulnerabilities include impacts on terrestrial and aquatic
16 ecosystems, hydrology and water resources, food and fiber production, coastal systems, and
17 human health. Appendix 4C provides excerpts of materials from the executive summary of the
18 IPCC (1998) report that comprise a helpful overview of key points regarding projected global
19 warming processes, likely climate change patterns, and their consequent impacts in terms of the
20 types of vulnerabilities noted above.
21 The IPCC (1998) report notes that human activities resulting in emissions of long-lived
22 GHCs are projected by General Circulation Models (GCMs) to lead to global and regional
23 changes in temperature, precipitation and other climate variables—resulting in increases in
24 global mean sea level; prospects for more extreme weather events, floods, and droughts in some
25 areas; and consequent changes in soil moisture. Based on various scenarios of current and
26 plausible future emissions of GHGs and aerosols and the range of sensitivities of climate change
27 to atmospheric levels (and residence time) of GHGs, GCMs project mean annual global surface
28 temperature increases in the range of 1 to 3.5 °C by 2100, a global mean sea level rise of 15 to
29 95 cm, and significant changes in spatial and temporal patterns of precipitation. The average rate
30 of warming will be more rapid than any seen hi the past 10,000 years, although regional changes
31 could differ substantially from mean global rates.
March 2001
4-139
DRAFT-DO NOT QUOTE OR CITE
-------
1 Human health, ecosystems, and socioeconomic sectors (e.g., hydrology and water
2 resources, food and fiber production, etc.) are projected to be vulnerable to the magnitude and
3 rate of climate change, as well as increased climate variability. Wide variations in the courses
4 and net impacts of climate change in different geographic areas can be expected, and, although
5 many regions are likely to experience severe adverse impacts (some possibly irreversible) of
6 climate change, some climate change impacts may be locally beneficial in some regions.
7 In general, projected climate change impacts can be expected to represent additional stresses on
8 those natural ecosystems and human societal systems already impacted by increasing resource
9 demands, unsustainable resource management practices, and pollution, with wide variation likely
10 across regions and nations in their ability to cope with consequent alterations in ecological
11 balances, in availability of adequate food, water, and clean air, and in human health and safety.
12 Appendix 4C also includes excerpts from the executive summary of the IPCC 1998 special report
13 regarding the assessment of different types of vulnerabilities to climate change projected for each
14 of 10 different geographic regions of the Earth, with emphasis being placed in Appendix 4C on
15 those projected for two regions (North America and Polar) of most relevance to the continental
16 United States and Alaska.
17 Appendix 4C notes that (a) the characteristics of subregions and sectors of North America
18 suggest that neither impacts of climate change nor response options will be uniform, and (b)
19 many systems of North America are moderately to highly sensitive to climate change, with the
20 range of estimated effects including the potential for substantial damage or, conversely, the
21 potential for some beneficial outcomes. The most vulnerable continental United States sectors
22 and regions include long-lived natural forest ecosystems in the East and ulterior West, water
23 resources in the southern plains, agriculture in the Southeast and southern plains, northern
24 ecosystems and habitats, estuaries and beaches in developed areas, and low-latitude cool and cold
25 water fisheries. Other sectors or subregions may benefit from warmer temperatures or increased
26 CO2 fertilization (e.g., west coast coniferous forests; some western rangelands; reduced energy
27 costs for heating in northern latitudes; reduced road salting and snow-clearance costs; longer
28 open-water seasons in norther channels and ports; and agriculture in the northern latitudes, the
29 interior West, and the west coast). For Alaska, substantial shifts in ecosystems (with possible
30 major declines or loss of some sensitive species like bear and caribou or of other ice-dependent
31 animals) may occur in parallel to beneficial effects such as opening of ice-bound water
March 2001
4-140
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
transportation routes or possible expanded agricultural viability secondary to longer growing
seasons. On the other hand, for North America, the potential for mainly deleterious direct or
indirect effects on human health is likely to increase (e.g., increased mortality directly linked to
temperature extremes, increases in incidence and spread of vector-borne infectious diseases,
impacts secondary to sea-level rise, and impacts secondary to increased tropospheric air pollution
[as depicted in Figure 4-26]).
More detailed evaluations of possible global climate change impacts on various U.S.
geographic areas are being conducted by the United States Global Change Research Program
(USGCRP). An overview report on the assessment results and key findings from a series of
workshops convened by the USGCRP National Assessment Synthesis team (NAST) has been
prepared (USGCRP, 2000). Selected highly salient key points from the report and subsidiary
regional assessments are presented in Appendix 4D. Overall key findings from the USGCRP
(2000) report are noted below.
(1) Increased Warming. Assuming continued growth in world GHG emissions, the primary
climate models used in the USGCRP assessment project that temperatures in the United
States will rise by 5 to 10 °F (3 to 6 °C) on average during the next 100 years.
(2) Differing Regional Impacts. Climate change will vary widely across the United States.
Temperature increases will vary somewhat from region to region. Heavy and extreme
precipitation events are likely to become more frequent, yet some regions will get drier.
The potential impacts of climate change will vary widely across the nation.
(3) Vulnerable Ecosystems. Many ecosystems are highly vulnerable to the projected rate and
magnitude of climate change. A few, such as alpine meadows in the Rocky Mountains and
some barrier islands, are likely to disappear entirely in some areas, with others, such as
some forests of the Southeast, being likely to experience major species shifts or break up.
Goods and services lost through disappearance or fragmentation of certain ecosystems are
likely to be costly or impossible to replace.
(4) Widespread Water Concerns. Water is an issue in every region, but the nature of the
vulnerabilities varies, with different nuances in each. Drought is an important concern in
every region. Floods and water quality are concerns in many regions. Snowpack changes
are especially important in the West, the Pacific Northwest, and Alaska.
March 2001
4-141
DRAFT-DO NOT QUOTE OR CITE
-------
BASES FOR CONCERN ABOUT GLOBAL WARMING AND CLIMATE
CHANGE EFFECTS ON THE ENVIRONMENT AND HUMAN HEALTH
GLOBAL WARMING
AND CLIMATE CHANGE
Geographic Variations in
Temperature Increases
(Avg. & Extremes), Rainfall,
I Sea-Level Rise I
/ Increased Frequency \
of
\ Air Stagnation Periods /
1
f >
Elderly (> 65 yrs) Most at
Risk — Also Infants
Key: Acclimatization
Geographic Distribution &
Abundance of Vectors/Hosts
Depend on Temperature,
Moisture, Habitat effects
/ Y / \
HEAT-STRESS
MORTALITY
Lower
threshold
temperatures
in North than
in South
If only partial
acclimatization
then heat
deaths rise
COLD-INDUCED
MORTALITY
Higher
threshold
temperatures
in South
(>0°C) than
North (<0'C)
Prob. drop in
cold-related
mortality
TICK-BORNE
DISEASES
USA
EXAMPLES:
Lyme Disease
Rocky
Mountain
Spotted
Fever
MOSQUITO-
BORNE
DISEASES
USA
EXAMPLES:
Malaria
Dengue Fever
Arbovirus-
Related
Encephalitis
Near-term: Storm Surges,
Costal Flooding
Long Term: Inland Advance
of Saltwater Oceans/Seas
/
HUMAN
HEALTH
IMPACTS
Loss of Life
Nutrition
Vector-Borne
Diseases
Other
Communicable
Diseases
\
OTHER
TYPES OF
IMPACTS
Damage to:
Industries
Agriculture
Aquatic and
Land
Ecosystems
i
Increased Tropospheric
Air Pollution
(PM, O3, CO, etc)
/
HUMAN
HEALTH
IMPACTS
Acute Pulmon.
Function
Decrements
Impaired Lung
Defenses
Increased
Respiratory
Disease
Susceptibility
\
OTHER
TYPES OF
IMPACTS
Forest and
Agriculture
Damage
Ecosystem
Effects
Materials
Damage
Figure 4-26. Bases for concern about global warming and climate change effects on the
environment and human health. Types of hypothesized likely human health
effects include (1) increases in mortality directly linked to temperature
extremes, (2) increases in incidence and spread of vector-borne infectious
diseases, (3) impacts secondary to projected sea-level rise, and (4) impacts
secondary to increased tropospheric air pollution. Additional impacts can be
expected because of shifting agricultural sustainability in various U.S. regions
consequent to extreme weather patterns leading to inland flooding or
droughts.
Source: Adapted from Grant (1989).
1 (5) Secure Food Supply. At the national level, the U.S. agriculture sector is likely to be able to
2 adapt to climate change. Overall, U.S. crop productivity is very likely to increase over the
3 next few decades, but the gains will not be uniform across the nation. Falling prices and
4 competitive pressures are very likely to stress some farmers, while benefiting consumers.
March 2001
4-142
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
(6) Near-Term Increases in Forest Growth. Forest productivity is likely to increase over the
next several decades in some areas as trees respond to higher CO2 levels. Over the longer
term, changes in larger scale processes such as fire, insects, droughts, and disease will
possibly decrease forest productivity. Also, climate change is likely to cause long-term
shifts in forest species (e.g., distribution of sugar maple stands more northward, out of the
United States).
(7) Increased Damage in Coastal and Permafrost Areas. Climate change and the resulting rise
in sea level are likely to exacerbate threats to building, roads, powerlines, and other
infrastructure in climatically sensitive places, such as low-lying coastlines and the
permafrost regions of Alaska.
(8) Other Stresses Magnified by Climate Change. Climate change will very likely magnify the
cumulative impacts of other stresses, such as air and water pollution and habitat destruction
caused by human development patterns. For some systems, such as coral reefs, the
combined effects of climate change and other stresses are very likely to exceed a critical
threshold, bringing large, possibly irreversible impacts.
(9) Surprises Expected. It is likely that some aspects and impacts of climate change will be
totally unanticipated as complex systems respond to ongoing climate change in
unforeseeable ways.
(10) Uncertainties Remain. Significant uncertainties remain in the science underlying regional
climate changes and their impacts. Further research is needed to improve understanding
and predictive ability about societal and ecosystem impacts and to provide the public with
additional useful information about adaptation strategies.
The selected findings highlighted in Appendix 4D from the USGCRP (2000) report and
subsidiary regional reports illustrate well the considerable uncertainties and difficulties in
projecting likely climate change impacts on regional or local scales. The findings presented in
Appendix 4D also reflect well the mixed nature of projected potential climate change impacts
(combinations of mostly deleterious, but other possible beneficial effects) for U.S. regions and
their variation across the different regions. Difficulties in assessing regional-specific potential
impacts also can be illustrated by discussion below of determinants of the potential extent of
direct or indirect impacts of global warming on human health, as abstracted from various
published assessments cited above or alluded to in Appendices 4C, 4D, and 4E.
March 2001
4-143
DRAFT-DO NOT QUOTE OR CITE
-------
1 Modeling efforts and published analyses by Kalkstein and others have helped to identify
2 important factors that affect the magnitude of temperature-dependent mortality and provide
3 bases for projecting future temperature-related mortality trends (see Appendix 4E for recently
4 published projections for U.S. cities by Kalkstein and Greene, 1997). Examples of key
5 determinants of temperature-related mortality include (1) weather-sensitive mortality occurs
6 mainly as a function of extremes of temperature beyond certain threshold points (for increasing
7 or decreasing temperatures) that are characteristic of any particular city; (2) the extent of the
8 mortality is generally more dependent on the duration of the periods (days) during which
9 threshold points are exceeded than on maximum temperatures and also varies as a function of
10 combined relative humidity, temperature, and barometric pressure conditions that constitute
11 "oppressive" weather events that vary for different locales; and (3) the major population segment
12 typically most severely affected are the elderly (2:65 years old).
13 Threshold temperature findings for summer and winter in U.S. cities suggest that weather
14 effects on mortality are relative (i.e., they vary in relation to the typical conditions to which local
15 residents have become acclimatized). Thus, the highest summer threshold temperatures for
16 mortality are found for the South and Southeast and the lowest hi the Pacific and Northeast U.S.
17 regions. Conversely, lowest threshold temperatures for winter mortality are found for cities in
18 the coldest regions, whereas notably higher thresholds for cold-associated deaths occur for
19 warmer region cities, with threshold values for some being well above the freezing point. Also,
20 the total accumulated times of occurrence in the season of particular oppressive weather events
21 are important determinants of mortality levels (e.g., hot conditions early in the spring and
22 summer have a larger impact than similar conditions later in the summer, and length of a
23 heat-stress period also has a larger impact than maximum temperatures reached).
24 Acclimatization is a key determinant of weather-related mortality, and the greatest initial
25 increases in heat-related mortality might be expected in cities where temperatures are normally
26 cooler or in areas where global-warming-induced climate changes lead to increased frequency
27 and durations of high-temperature episodes. Eventual acclimatization may occur over the years,
28 however, when higher-than-usual temperatures become the new norm for cities in currently
29 cooler, more northern regions. As for cold-associated deaths, if average winter temperatures
30 were not to drop as low as usual in various regions, then whiter mortality might generally
31 decrease because of fewer days falling below existing winter threshold levels for many cities.
March 2001
4-144
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
But, if acclimatization occurs to higher average winter temperatures and wider variations in
temperature extremes occur in some areas because of global-warming-induced weather changes,
then those periods of lower maximum temperatures (especially of several days duration) could
cause even higher than past mortality rates previously observed with comparable winter
conditions. More sophisticated modeling also is needed to take into account combined effects of
temperature extremes and weather-related increases in air pollutants as possible mortality
determinants; for example, increased mortality or morbidity effects because of temperature
extremes may be exacerbated or added to by higher surface level atmospheric PM derived from
increased coal or oil combustion to generate more heat (in winter) or electricity (in summer for
air conditioning) during extreme temperature periods.
In addition to concern about possible mortality increases because of temperature extremes,
global warming, and consequent climate change also may impact human health through increases
in some infectious diseases. For many parts of the world, infectious diseases remain among the
leading causes of death, as occurred earlier in industrialized or "developed" countries (where
diseases such as influenza, pneumonia, and tuberculosis were among the leading causes of death
in 1900). Since then, the incidence and associated mortality for these and other infectious
diseases such as diphtheria, typhus, and polio have been reduced dramatically in industrialized
countries, hi developed countries, it is not clear to what extent global-warming-induced climate
change may cause general increases in the incidence of such diseases, unless serious disruptions
of social structures occur or, in some coastal areas, breakdowns in sanitation systems happen as a
consequence of sea-level rise. The spread of infectious diseases is likely of greater concern for
many less developed countries, where inadequate medical care systems, immunization programs,
housing conditions, and nutrition make them more vulnerable to the spreading of such diseases.
Of particular shared concern for both developed and less developed countries with regard to
potential global warming impacts are infectious diseases spread by climate-dependent vectors.
Vector-borne diseases are those for which the infectious microbial agent is transmitted to humans
via another agent (the vector), such as the flea, tick, or mosquito. Well known examples of
vector-borne diseases are malaria (transmitted to humans via mosquitos) and bubonic plague
(transmitted via fleas or, at times, via animals directly to man as a respiratory disease). Climate
change can affect vector-borne diseases by various direct impacts on the infectious agent, the
vector, or intermediate hosts through variations in temperature, humidity, rainfall, or storm
March 2001
4-145
DRAFT-DO NOT QUOTE OR CITE
-------
1 patterns that alter (1) the multiplication rates of the infectious agent or the vector, (2) the biting
2 rate of the vector, (3) the geographic distribution of the intermediate animal hosts, or (4) the
3 amount of time that intermediate hosts or human hosts are exposed to the vector. Climate change
4 also can affect indirectly the rates or incidences of vector-borne diseases via impacts on
5 agricultural practices, ecosystem mixes (of grasses, trees, underbrush, etc.), surface water levels,
6 or other factors that determine intermediate host or vector distribution or survival. The variety of
7 vector-borne diseases is considerable, with some being of more concern than others for particular
8 countries, depending on specific climatic conditions and existing pools of infected hosts (both
9 human and intermediate animal hosts). Examples of vector-borne diseases illustrative of
10 concerns that apply to the United States for potential spread of vector-borne diseases include
11 Lyme disease, Rocky Mountain spotted fever, dengue fever, malaria, and viral encephalitis.
12 Lyme disease (initially recognized in Lyme, CT) is an inflammatory disease caused by a
13 spirochete, Borrelia burgdorferi, transmitted by several subspecies ofLcodes racinus ticks.
14 Numerous species of birds and mammals can be hosts for various subspecies of the tick vector,
15 with varying geographic distributions. Lyme disease has four major U.S. foci, is spreading
16 rapidly, and has been found in Europe (Germany, Switzerland, France, and Austria). The U.S.
17 distribution of human cases of the disease tends to match areas where the tick vector is abundant,
18 and deer populations, along with factors such as temperature, humidity and local vegetation,
19 represent key determinants of tick abundance. The precise impact of global warming and climate
20 change on the distribution of Lyme disease is difficult to estimate. Lengthening of warm weather
21 periods and shortening of winter weather could enhance tick vector abundance and its potential
22 spread into adjoining areas if the weather changes (temperature, precipitation, etc.) were to favor
23 wider distribution of deer or other animal or bird hosts. Shifts of human populations into or out
24 of affected areas in response to changes in local climate also would help determine location-
25 specific alterations in Lyme disease rates.
26 Rocky Mountain spotted fever (initially identified in western mountain areas but actually
27 much more prevalent in southeastern U.S. states) is a highly fatal disease if not promptly
28 diagnosed and treated. Caused by the occobacillus, Rickettsiae rickettsii, the disease is spread by
29 ticks and is also known as tick fever, with analogous diseases occurring in many other countries.
30 The main North America vectors are the dog tick, D. variabilis, and the wood ticks, D. andersoni
31 and D. occidentalis, with varying geographic distributions. Geographic tick distributions parallel
March 2001
4-146
DRAFT-DO NOT QUOTE OR CITE
-------
1 closely the typical U.S. distribution of disease cases (highest incidence across the South).
2 Crucial for the spread of Rocky Mountain spotted fever is the wide variety of intermediate hosts
3 available to the ticks (i.e., many woodland mammals and birds) and temperature. Certain
4 optimum ranges of high temperatures (24 to 30 °C) likely speed the rickettsial growth in the
5 ticks, and ambient temperatures are important in determining tick breeding season length and
6 cycles, as well as their activity levels and biting rates. Each are enhanced by higher temperatures,
7 and the abundance of the vector is held in check, in part, by frequency and length of time that
8 winter temperatures drop well below freezing, thus killing overwintering adults. Lastly, relative
9 humidity conditions and rainfall are important as well, hi that hot dry weather results in
10 desiccation of ticks and their eggs, reducing reproduction rates. Global warming-induced climate
11 change might increase the range of Rocky Mountain spotted fever tick vectors into more
12 northward U.S. areas and, possibly, into Canada, assuming the climate change includes sufficient
13 rainfall to sustain adequate habitats for host species and adequate moisture for survival of ticks
14 and eggs. Hot, dry periods caused by any prolonged drought conditions in the United States or
15 Canada predicted by some global warming scenarios, conversely, would not be conducive to
16 increased incidence of the disease in drought-affected areas.
17 Malaria, once widespread in the southern United States, remains endemic in many areas of
18 the world and is caused by four agents: (1) Plasmodium vivax, (2) P. malariae, (3) P. ovale, and
19 (4) P. falciparum. The agents cause clinical syndromes of varying severity, the most serious
20 being caused by P. falciparum, which can progress to death (>10% fatality in untreated children
21 and nonimmune adults). The other forms, although less severe, are still debilitating and are
22 typified by recurring episodes of fever, chills, and sweating. Malarial agents are transmitted from
23 infected humans, as the main host pool, by the bite of various subspecies of anopheles mosquitos.
24 Ambient temperatures of at least 15 to 18 °C are crucial for development of the malarial agents
25 within the mosquitos, and ambient temperature levels determine breeding season length and
26 survival rates (higher tropical temperatures being most favorable). Man's agricultural activities,
27 in providing irrigation ditches and more stagnant water habitats, has contributed to spread and
28 abundance of the anopheles mosquito in many areas of the world. Malaria is now rarely
29 endogenously transmitted in the United States, the pool of infected humans as hosts having been
30 reduced very substantially, owing to mosquito eradication programs. Prior to such programs, the
31 disease was endemic in widespread southern U.S. areas up to the 1940s, but, since then,
March 2001
4-147
DRAFT-DO NOT QUOTE OR CITE
-------
1 outbreaks mainly have occurred because of infected immigrants entering the country or U.S.
2 military veterans returning from overseas endemic areas. Global warming leading to higher
3 temperatures in more northerly U.S. areas and Europe could enhance conditions for the spread of
4 the disease. Both the range and abundance of competent vectors (various anopheles subspecies)
5 likely would be increased, especially if increased irrigation were required to support agriculture,
6 owing to higher temperatures. Also, higher temperatures in more northerly areas could extend
7 the range of adequate temperatures (>15 to 18 °C) needed for development of malarial agents in
8 the mosquitos. The remaining key factor in determining the likelihood of the spread of malaria,
9 however, is the infected host pool, with numbers of infected human hosts moving into or out of
10 areas of enhanced vulnerability being of crucial importance, as emphasized by Longstreth (1999).
11 Dengue fever, another mosquito-borne disease, is caused by four serotypes of a Group B
12 arbovirus. Fever, general muscle ache, severe headache, and retroorbital pain typify dengue fever
13 (usually not fatal); but it can progress to dengue haemorrhagic fever or dengue shock syndrome
14 (often fatal). Once endemic along the U.S. Gulf and South Atlantic coasts, dengue fever is now
15 rarely endogenously transmitted in the United States. The Aedis aegypti mosquito is the primary
16 vector, with wide southern U.S. distribution. The breeding season of the A. aegypti mosquito is
17 temperature-dependent, with breeding year-round in southern Florida, nearly year-round
18 elsewhere in Florida and along the Gulf Coast, and much shorter for successively more
19 northward bands of geographic distribution. Another potential vector, Aedes triseriatus, is
20 endogenous to states east of the Mississippi, and Aedes albopictus, a proven dengue vector
21 introduced from northern Asia, has been found hi scattered U.S. sites. Higher temperatures are
22 also crucial for dengue transmission; transmission of dengue occurred experimentally only if
23 A. aegypti mosquitos were kept at 30 °C, and the incubation period for the virus to develop in the
24 mosquitos was shortened at 32 to 35 °C. Consistent with this, cases of dengue haemorrhagic
25 fever increased at non-U.S. sites when daily mean temperatures were 28 to 30 °C during hot
26 seasons, but decreased at the sites during cooler seasons with 25 to 28 °C temperatures.
27 Temperature increases in temperate ozone areas with A. Aegypti or A. albopictus present would
28 tend to expand the range of these dengue fever vectors, including potential spread especially of
29 A. albopictus farther north in the United States and, perhaps, into Canada, in view of its
30 adaptation to cold weather as well. Whether or not increases in dengue will actually occur,
31 however, likely will depend on the distribution of rainfall and moisture content, the effects of
March 2001
4-148
DRAFT-DO NOT QUOTE OR CITE
-------
1 agricultural practices (e.g., increased irrigation), and movements of infected human hosts into or
2 out of areas with increased vector density.
3 Arbovirus-induced encephalitis syndromes vary in severity but include several that can be
4 highly fatal and are related to several other types of arbovirus-related syndromes (e.g., yellow
5 fever, dengue and other haemorrhagic fevers, hepatitis, arthritis, rashes, various tropical fevers).
6 Different types of mosquitos that serve as competent vectors for various types of
7 arbovirus-induced encephalitis of concern for the United States display different patterns of
8 distribution and differentially infect other hosts besides man (e.g., birds and large vertebrates
9 [horses, etc.] for some, birds and swine for another, and small woodland animals for others).
10 All have temperature-dependent components involved in development or transmission of the
11 viruses, but specific effects vary for different types. For example, the maximum temperatures
12 allowing the western equine encephalitis (WEE) vector to transmit the virus effectively are below
13 25 °C, and this allows for earlier spread of the disease in warm periods and the possible more
14 northern spread of the disease. In contrast, St. Louis encephalitis (SLE) arbovirus development
15 and transmission are markedly enhanced by temperatures exceeding 25 °C. Rainfall and
16 moisture patterns are also important, with most vectors (e.g., Cx tarsalis) benefitting from higher
17 rainfall; but at least one (Cxpipiens) is enhanced by less rainfall, with outbreaks of its
18 encephalitis syndrome being more common during high-temperature drought periods. Thus,
19 effects of global warming and climate change on the incidence and spread of arbovirus-related
20 encephalitis syndromes are difficult to predict. However, it generally appears that higher
21 temperatures should enhance the abundance and wider U.S. geographic distribution of most of
22 the competent mosquito vectors. All of this again assumes that higher temperatures and rainfall
23 patterns will be such to allow adequate habitats for other hosts besides humans in the potential
24 new range areas. Lastly, as noted before for the other infectious diseases discussed, the
25 movement of populations into or out of the affected areas also will be important in determining
26 any location-specific increased (or decreased) incidence of arbovirus-related encephalitises.
27 Of special concern would be the introduction of any new arboviruses not now currently endemic
28 in the United States (e.g., Japanese B encephalitis [VBE], not currently found hi the United
29 States but closely related to SLE in terms of involving Culex mosquitos and birds, with several
30 Culex subspecies in the United States found to be effective vectors for the virus).
March 2001
4-149
DRAFT-DO NOT QUOTE OR CITE
-------
1 The above discussion of potential effects of global warming and climate change on the
2 incidence and spread of infectious diseases is further complicated by considerations of possible
3 impacts of expected sea-level rise in response to global warming. Some low-lying coastal areas
4 now serving as excellent habitats for certain mosquitos, for example, might come to be inundated
5 by seawater and no longer be available breeding areas. Or, increased storm surges or expansion
6 of marshy areas reaching farther inland might contribute to creation of conditions in some areas
7 more favourable to enhance mosquito breeding. Also of concern is the potential for disruption of
8 sanitation systems. The spread of infectious diseases, besides the vector-borne types discussed
9 above, could be increased because of flooding of coastal cities secondary to heavy precipitation
10 events (e.g., hurricanes). Inundation of sewage treatment facilities and sewage lines might not
11 only result in immediate spread of disease-containing fecal or other material, but damage to such
12 sanitation system components could result in longer term disruption of waste-removal
13 capabilities and the spread of disease.
14 Lastly, another concern with climate-induced heavy precipitation events or sea-level rise is
15 the potential for flooding of inland or coastal waste disposal sites. This could result in increased
16 spread of waterborne infectious diseases, depending on the specific materials present in such
17 dumps and the extent of their dispersal caused by flooding. The flooding of dump sites
18 containing hazardous chemical wastes represents yet another potential concern associated with
19 sea-level rise. The spread of various toxic chemicals from such waste disposal sites could carry
20 with it increased threats of many types of possible health effects, as well as potential
21 environmental effects (natural vegetation and ecosystem damage, contamination of crop lands by
22 toxic chemicals, etc.).
23 Difficulties in projecting region-specific climate change impacts are complicated further by
24 the need to evaluate potential effects of local- or regional-scale changes in key air pollutants not
25 only on global scale temperature trends but also in terms of potentially more local- or regional-
26 scale impacts on temperature and precipitation patterns. Of much importance for this are varying
27 roles played by atmospheric particles.
28
29 4.5.2.2 Airborne Particle Relationships to Global Warming and Climate Change
30 Atmospheric particles both scatter and absorb incoming solar radiation at visible light
31 wavelengths. The scattering of solar radiation back to space leads to a decrease in transmission
March 2001
4-150
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
of visible radiation to the Earth's surface and, hence, to a decrease in the heating rate of the
surface and the atmosphere. The absorption of either incoming solar radiation or outgoing
terrestrial infrared radiation by atmospheric particles results in heating of the lower atmosphere.
Interactions of atmospheric particles with electromagnetic radiation from the visible through the
infrared spectral regions are responsible for their direct effects on climate, which are the result of
the same physical processes responsible for visibility degradation. Visibility reduction is caused
by particle scattering in all directions, whereas climate effects result mainly from scattering in the
upward direction. The net effect of the above processes can be expressed as a radiative forcing,
which is the change in the average net radiation at the top of the troposphere because of a change
in solar (shortwave, or visible) or terrestrial (longwave, or infrared) radiation (Houghton et al.,
1990). The radiative forcing drives the climate to respond, but because of uncertainties in a
number of feedback mechanisms involving climate response, radiative forcing is used as a first-
order estimate of the potential importance of various substances. Sulfate particles scatter solar
radiation effectively and do not absorb at visible wavelengths, whereas they absorb weakly at
infrared wavelengths (IPCC, 1995). Nitrate particles exhibit grossly similar properties. The
effects of mineral dust particles are complex; they weakly absorb solar radiation but their overall
effect on solar radiation depends on particle size and the reflectivity of the underlying surface.
They absorb infrared radiation and thus contribute to greenhouse warming (Tegen et al., 1996).
Organic carbon particles mainly reflect solar radiation, whereas elemental carbon and other black
carbon particles (e.g., PAHs with H:C ratios of <0.3) are strong absorbers of solar radiation
(IPCC, 1995). However, the optical properties of carbonaceous particles are modified if they
become coated with water or sulfuric acid. Particles containing black carbon also can exert a
direct effect after deposition onto surfaces that are more reflective (e.g., snow and ice). In this
case, additional solar radiation is absorbed by the surface; conversely, more reflective particles
deposited on a dark surface result in additional solar radiation being reflected back to space.
Anthropogenic (Twomey, 1974; Twomey, 1977) and biogenic (Charlson et al., 1987)
sulfate particles also exert indirect effects on climate by serving as cloud condensation nuclei,
which results in changes in the size distribution of cloud droplets by producing more particles
with smaller sizes. The same mass of liquid water in smaller particles leads to an increase in
amount of solar radiation that clouds reflect back to space because the total surface area of the
cloud droplets is increased. This has been supported by satellite observations indicating that the
March 2001
4-151
DRAFT-DO NOT QUOTE OR CITE
-------
1 effective radius of cloud droplets is smaller in the Northern Hemisphere than in the Southern
2 Hemisphere (Han et al., 1994). Smaller cloud droplets also have a lower probability of
3 precipitating and, thus, have a longer lifetime than larger ones. Although the effects of sulfate
4 have been considered most widely, interactions with other aerosol components also may be
5 important. Novakov and Penner (1993) have provided evidence that carbonaceous particles can
6 modify the nucleation properties of sulfate particles.
7 The amount of solar radiation incident on the earth-atmosphere system, or the solar
8 constant, is 1370 W m'2, or 342.5 W rn2 on a globally averaged basis (calculated by dividing the
9 solar constant by 4). The addition of sulfate and organic carbon as airborne PM results in
10 enhanced scattering and net cooling, whereas the addition of particles containing elemental
11 carbon results in absorption of solar and terrestrial radiation and net heating. The estimated
12 raditive forcing because of the scattering of solar radiation back to space caused mainly by
13 sulfate particles is - 0.4 W m"2 (IPCC, 1995), with an uncertainty range of a factor of two. The
14 uncertainty range reflects uncertainties in the emissions of SO2, the amount of SO2 that is
15 oxidized to sulfate, the atmospheric lifetime of sulfate, and the optical properties of the sulfate
16 particles. These values may be compared to the radiative forcing exerted by greenhouse gases of
17 about + 2.4 W m'2, with an uncertainty factor of 1.15 from the preindustrial era (ca. 1800) to
18 1994. Since the latter part of the 19th century, the mean surface temperature of the earth has
19 increased from 0.3 to 0.6 °C according to the IPCC (1995) assessment. Estimates of the indirect
20 effects of particles range from 0 to -1.5 W m"2 (IPCC, 1995). Because of a lack of quantitative
21 knowledge, no central value could be given. Therefore, on a globally averaged basis, the direct
22 and indirect effects of anthropogenic sulfate particles likely have offset partially the warming
23 effects caused by increases in levels of greenhouse gases (Charlson et al., 1992).
24 Much of the work investigating the effects of particles on climate has focused on sulfate
25 particles. However, particles containing elemental carbon (EC) from fossil fuel combustion and
26 biomass burning or mineral dust may exert radiative forcing, with spatial distributions very
27 different than for sulfate. Tegen et al. (1996) and Tegen and Lacis (1996) used a global scale
28 three-dimensional model to evaluate the radiative forcing caused by mineral dust particles.
29 Tegen and Lacis (1996) found that the sign and the magnitude of the radiative forcing depends on
30 the height distribution of the dust and the effective radius of the particles. In particular, for a dust
31 layer extending from 0 km to 3 km, positive radiative forcing at visible wavelengths is found for
March 2001
4-152
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
particle radii greater than 1.8 ^m, whereas negative forcing is found for smaller particles. They
calculated a global mean radiative forcing caused by mineral dust from all sources of 0.14 W m"2
and from mineral dust from lands disturbed by human activity of 0.09 W m"2. This value
represents a near cancellation between a much larger solar forcing of-0.25 W m"2 and a thermal
forcing of 0.34 W m"2. Uncertainty factors could not be estimated for these calculations because
they were judged to be largely unknown. Haywood and Shine (1995) estimated a global mean
radiative forcing of 0.1 W m"2, with an uncertainty factor >3, caused by the absorption of solar
radiation by EC released by fossil fuel combustion. The IPCC (1995) estimated a global mean
radiative forcing of - 0.1 W m"2 caused by particles produced by biomass burning, with an
uncertainty factor of three. The global mean radiative forcing exerted by particles would then be
-0.5 W m'2, with an uncertainty of about a factor of 2.4. Figure 4-27 summarizes estimates of
global mean radiative forcing exerted by greenhouse gases and various types of particles.
Deviations from the global mean values can be very large on the regional scale. For
instance, Tegen et al. (1996) found that local radiative forcing exerted by dust raised from
disturbed lands ranges from -2.1 W m"2 to 5.5 W m"2 over desert areas and their adjacent seas.
The largest regional values of radiative forcing caused by anthropogenic sulfate are about
-3 W m'2 in the eastern United States, south central Europe, and eastern China (Kiehl and
Briegleb, 1993). These regional maxima in aerosol forcing are at least a factor of 10 greater than
their global mean values shown in Figure 4-27. By comparison, regional maxima in forcing by
the well-mixed greenhouse gases are only about 50% greater than their global mean value (Kiehl
and Briegleb, 1993). Thus, the estimates of local radiative forcing by particles also are large
enough to completely cancel the effects of greenhouse gases in many regions and to cause a
number of changes in the dynamic structure of the atmosphere that still need to be evaluated.
A number of anthropogenic pollutants whose distributions are highly variable are also effective
greenhouse absorbers. These gases include O3 and, possibly, HNO3, C2H4 , NH3, and SO2, all of
which are not commonly considered in radiative forcing calculations (Wang et al. 1976). High
ozone values are found downwind of urban areas and areas where there is biomass burning.
However, Van Borland et al. (1997) found that there may not be much cancellation between the
radiative effects for ozone and for sulfate, because both species have different seasonal cycles
and show significant differences in their spatial distribution.
March 2001
4-153
DRAFT-DO NOT QUOTE OR CITE
-------
adiative forcing (W m"2)
J _* M W
III III
c.
CO
f "
15-1-
JQ
0
o -
T
1
JL.
•
— Halocarbons
— N2O
*~~" CH^
— c°2 Sulfate
T ^
T /\
I Tropospheric -.-
Ozone 1
Direct
Effect
In
E
Fossil Solar
Fuel Biomass Mineral Variability
Soot Burning QUSt "p
v .x, _ m
T
direct
!ffect
Confidence level
High Low Low Low Very Very Very Very Very
low low low low low
Figure 4-27. Estimated global mean radiative forcing exerted by gas and various particle
phase species from 1850 to 1950.
Source: Adapted from IPCC (1995) and Tegen and Lacis (1996).
1 Observational evidence for the climatic effects of particles is sparse. Haywood et al. (1999)
2 found that the inclusion of anthropogenic aerosols results in a significant improvement between
3 calculations of reflected sunlight at the top of the atmosphere and satellite observations in
4 oceanic regions close to sources of anthropogenic PM.
5 Uncertainties in calculating the direct effect of airborne particles arise from a lack of
6 knowledge of their vertical and horizontal variability, their size distribution, chemical
7 composition and the distribution of components within individual particles. For instance,
8 gas-phase sulfur species may be oxidized to form a layer of sulfate around existing particles in
9 continental environments, or they may be incorporated in sea-salt particles (e.g., Li-Jones and
10 Prospero, 1998). In either case, the radiative effects of a given mass of the sulfate will be much
March 2001
4-154
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
lower than if pure sulfate particles were formed. It also must be stressed that the overall radiative
effect of particles at a given location is not simply determined by the sum of effects caused by
individual classes of particles because of interactions between particles with different radiative
characteristics and with gases.
Calculations of the indirect effects of particles on climate are subject to much larger
uncertainties than are calculations of their direct effects, reflecting uncertainties in a large
number of chemical and microphysical processes in describing the effects of sulfate on the size
distribution and number of droplets within a cloud. A complete assessment of the radiative
effects of PM will require supercomputer calculations that incorporate the spatial and temporal
behavior of particles of varying composition that have been emitted or formed from precursors
emitted from different sources. Refining values of model input parameters (such as improving
emissions estimates) may be as important as improving the models per se in calculations of direct
radiative forcing (Pan et al., 1997) and indirect radiative forcing (Pan et al., 1998) caused by
sulfate. However, uncertainties associated with the calculation of radiative effects of particles
likely will remain much larger than those associated with well-mixed greenhouse gases.
This means that, although on a global scale atmospheric particles likely exert an overall net
effect of slowing global warming, much uncertainty would apply to any modeling efforts aimed
at projecting net effects on global warming processes, resulting climate change, and any
consequent human health or environmental effects because of location-specific increases or
decreases in anthropogenic emissions of atmospheric particles or their precursors. For example,
any net impacts of regional sulfates in reducing global-climate-change-induced increases in local
temperatures may well be offset partially by local surface level heating because of carbonaceous
particles from diesel emissions or coal combustion energy generation being deposited on snow or
ice covered surfaces or contributing to more rapid evaporation or rainout of water from overhead
clouds.
March 2001
4-155
DRAFT-DO NOT QUOTE OR CITE
-------
1 4.6 SUMMARY
2 4.6.1 Particulate Matter Effects on Vegetation and Ecosystems
3 Human existence on this planet depends on ecosystems and the services and products they
4 provide. Both ecosystem structure and function play an essential role in providing societal
5 benefits. Society derives two types of benefits from the structural aspects of an ecosystem:
6 (1) products with market value such as fish, minerals, forage, forest products, biomass fuels,
7 natural fiber, and many pharmaceuticals, and the genetic resources of valuable species (e.g.,
8 plants for crops and timber and animals for domestication); and (2) the use and appreciation of
9 ecosystem for recreation, aesthetic enjoyment, and study.
10 Ecosystem functions that maintain clean water, pure air, a green earth, and a balance of
11 creatures, are functions that enable humans to survive. They are the dynamics of ecosystems.
12 The benefits they impart include absorption and breakdown of pollutants, cycling of nutrients,
13 binding of soil, degradation of organic waste, maintenance of a balance of gases in the air,
14 regulation of radiation balance, climate, and the fixation of solar energy. Concern has risen in
15 recent years concerning the integrity of ecosystems because there are few ecosystems on the
16 Earth today that are not influenced by humans. For this reason, the deposition of PM and its
17 impact on vegetation and ecosystems is of great importance.
18 The PM whose effects on vegetation and ecosystems are considered in this chapter is not a
19 single pollutant but represents a heterogeneous mixture of particles differing in origin, size, and
20 chemical constituents. The effects of exposure to a given mass concentration of PM of particular
21 size (measured as PM10; PM2 5, etc.) may, depending on the particular mix of deposited particles,
22 lead to widely differing phytotoxic responses. This has not been characterized adequately.
23 Atmospheric deposition of particles to ecosystems takes place via both wet and dry
24 processes through the three major routes indicated below.
25 (1) Precipitation scavenging, in which particles are deposited in rain and snow
26 (2) Fog, cloud water, and mist interception
27 (3) Dry deposition, a much slower, yet more continuous removal to surfaces
28 Deposition of heavy metal particles to ecosystems occurs by wet and dry processes. Dry
29 deposition is considered more effective for coarse particles of natural origin and elements such as
30 iron and manganese, whereas wet deposition generally is more effective for fine particles of
March 2001
4-156
DRAFT-DO NOT QUOTE OR CITE
-------
1 atmospheric origin and elements such as cadmium, chromium, lead, nickel, and vanadium. The
2 actual importance of wet versus dry deposition, however, is highly variable, depending on the
3 type of ecosystem, location, and elevation.
4 Deposition of PM on above-ground plant parts can have either a physical and or chemical
5 impact, or both. Particles transferred from the atmosphere to plant surfaces may cause direct
6 effects if they (1) reside on the leaf, twig, or bark surface for an extended period; (2) be taken up
7 through the leaf surface; or (3) are removed from the plant via resuspension to the atmosphere,
8 washing by rainfall, or litter-fall with subsequent transfer to the soil.
9 Chemical effects include excessive alkalinity or acidity. The effects of "inert" PM are
10 mainly physical, whereas the effects of toxic particles are both chemical and physical. The
11 effects of dust deposited on plant surfaces or on soil are more likely to be associated with their
12 chemistry than with the mass of deposited particles and are usually of more importance than any
13 physical effects. The majority of the easily identifiable direct and indirect effects, other than
14 climate-change impacts, occur in severely polluted areas around heavily industrialized point
15 sources such as limestone quarries; cement kilns; and iron; lead, and various smelting factories.
16 Studies of the direct effects of chemical additions to foliage in particulate deposition have found
17 little or no effects of PM on foliar processes; however, both conifers and deciduous species have
18 shown significant effects on leaf surface structures after exposure to simulated acid rain or mist
19 at pH 3.5. Many experimental studies indicate that epicuticular waxes (which function to prevent
20 water loss from plant leaves) can be destroyed by acid rain in a few weeks. This function is
21 particularly crucial in conifers because of the longevity of evergreen foliage.
22 Though there has been no direct evidence of a physiological association between tree injury
23 and exposure to metals, heavy metals have been implicated because their deposition pattern is
24 correlated with forest decline. The role of heavy metals has been indicated by phytochelatin
25 measurements. Phytochelatins are intracellular metal-binding peptides that act as indicator of
26 metal stress. Because they are produced by plants as a response to sublethal concentrations of
27 heavy metals, they can be used to indicate that heavy metals are involved in forest decline.
28 Concentrations of the phytochelatins increased with altitude, as did forest decline, and they also
29 increased across regions showing increased levels of forest injury.
30 Secondary organics formed in the atmosphere have been referred to under the following
31 terms: toxic substances, pesticides, hazardous air pollutants (HAPS), air toxics, semivolatile
March 2001
4-157
DRAFT-DO NOT QUOTE OR CITE
-------
1 organic compounds (SOCs), and persistent organic pollutants (POPS). The chemical substances
2 listed under the above headings are not criteria pollutants controlled by NAAQS as cited under
3 CAA Sections 108 and 109 (U.S. Code, 1991), but rather are controlled under CAA Sect. 112,
4 Hazardous Air Pollutants. Their possible effects in the environment on humans and ecosystems
5 are discussed in many other government documents and publications. They are mentioned in this
6 chapter because, in the atmosphere many of the chemical compounds are partitioned between gas
7 and particle phases and are deposited as particulate matter. As particles, they become airborne
8 and can be distributed over a wide area and impact remote ecosystems. Some of the chemical
9 compounds are of concern to humans because they may reach toxic levels in food chains of both
10 animals and humans, whereas others tend to decrease or maintain the same toxicity as they move
11 through the food chain.
12 An important characteristic of fine particles is their ability to affect the flux of solar
13 radiation passing through the atmosphere directly, by scattering and absorbing solar radiation,
14 and indirectly, by acting as cloud condensation nuclei that, in turn, influence the optical
15 properties of clouds. Regional haze has been estimated to diminish surface solar visible radiation
16 by approximately 8%. Crop yields have been reported as being sensitive to the amount of
17 sunlight received, and crop losses have been attributed to increased airborne particle levels in
18 some areas of the world.
19 The transmission of solar UV-B radiation through the Earth's atmosphere is controlled by
20 ozone, clouds, and particles. The depletion of stratospheric ozone caused by the release of
21 chlorofluorcarbons and other ozone-depleting substances has resulted in heightened concern
22 regarding potentially serious increases in the amount of solar UV-B (SUVB) radiation reaching
23 the Earth's surface. Plant species vary enormously in their response to UV-B exposures, and
24 large differences in response also occur among different genotypes within a species. In general,
25 dicotyledonous plants are more sensitive than monocotyledons from similar environments.
26 In addition, plant responses may differ depending on stage of development. Because plants
27 evolved under'the selective pressure of ambient UV-B radiation in sunlight, they have developed
28 adaptive mechanisms. Although inhibition of photosynthesis is a detrimental growth effect,
29 flavonoid synthesis represents acclimation. Plants growing under full light have been shown to
30 be protected against UV-B effects but not when growing under weak visible light. A common
March 2001
4-158
DRAFT-DO NOT QUOTE OR CITE
-------
1 adaptation is alteration in leaf transmission properties, which results in attenuation of UV-B in
2 the epidermis before it can reach the leaf interior.
3 Indirect effects of PM on plants are usually the most significant because they can alter
4 nutrient cycling in ecosystems and inhibit plant uptake of nutrients and, therefore, have a great
5 impact on ecosystem biodiversity. Indirect effects occur through the soil and result from the
6 deposition of heavy metals, nitrates, sulfates, or acidic precipitation and their impact on the soil
7 microbial community. The soil environment is one of the most dynamic sites of biological
8 interaction in nature. Bacteria in the soil are essential components of the nitrogen and sulfur
9 cycles that make these elements available for plant uptake. Fungi form mycorrhizae,
10 a mutualistic symbiotic relationship, that is integral in mediating plant uptake of mineral
11 nutrients. Changes in the soil environment that influence the role of the bacteria and fungi in
12 nutrient cycling and availability determine plant and ecosystem response.
13 Major impacts of PM on soil environments occur through deposition of nitrates and sulfates
14 and the acidifying effect of the ET ion associated with these compounds in wet and dry
15 deposition. Although the soils of most of North American forest ecosystems are nitrogen
16 limited, there are some forests that exhibit severe symptoms of nitrogen saturation. They include
17 the high-elevation, spruce-fir ecosystems in the Appalachian Mountains; the eastern hardwood
18 watersheds at the Femow Experimental Forest near Parsons, WV; the mixed conifer forest and
19 chaparral watershed with high smog exposure in the Los Angeles Air Basin; the high-elevation
20 alpine watersheds in the Colorado Front Range; and a deciduous forest in Ontario, Canada.
21 Nitrogen saturation results when additions to soil background nitrogen (nitrogen loading)
22 exceed the capacity of plants and soil microorganisms to utilize and retain nitrogen. An
23 ecosystem no longer functions as a sink under these circumstances. Possible ecosystem
24 responses to nitrate saturation, as postulated by Aber and his coworkers, include (1) a permanent
25 increase in foliar nitrogen and reduced foliar phosphorus and lignin because of the lower
26 availability of carbon, phosphorus, and water; (2) reduced productivity in conifer stands caused
27 by disruptions of physiological function; (3) decreased root biomass and increased nitrification
28 and nitrate leaching; (4) reduced soil fertility, the results of increased cation leaching, increased
29 nitrate and aluminum concentrations in streams, and decreased water quality. Saturation implies
30 that some resource other than nitrogen is limiting biotic function. Water and phosphorus for
31 plants and carbon for microorganisms are the resources most likely to be the secondary limiting
March 2001
4-159
DRAFT-DO NOT QUOTE OR CITE
-------
1 factors. The appearance of nitrogen in soil solution is an early symptom of excess nitrogen. In
2 the final stage, disruption of forest structure becomes visible.
3 Changes in nitrogen supply can have a considerable impact on an ecosystem's nutrient
4 balance. Increases in soil nitrogen play a selective role. Plant succession patterns and
5 biodiversity are affected significantly by chronic nitrogen additions in some ecosystems.
6 Long-term nitrogen fertilization studies in both New England and Europe suggest that some
7 forests receiving chronic inputs of nitrogen may decline in productivity and experience greater
8 mortality. For example, long-term fertilization experiments at Mount Ascutney, VT, suggest that
9 declining coniferous forest stands with slow nitrogen cycling may be replaced by deciduous
10 fast-growing forests that cycle nitrogen rapidly. Excess nitrogen inputs to unmanaged heathlands
11 in the Netherlands also have been found to result in nitrophilous grass species replacing slower
12 growing heath species. Over the past several decades, the composition of plants in the forest
13 herb layers had been shifting toward species commonly found on nitrogen-rich areas. It also was
14 observed that the fruiting bodies of mycorrhizal fungi had decreased in number.
15 Notable impacts of excess nitrogen deposition also have been observed with regard to
16 aquatic systems. For example, atmospheric nitrogen deposition into soils in watershed areas
17 feeding into estuarine sound complexes (e.g., the Pamlico Sound of North Carolina) appear to
18 contribute to excess nitrogen flows in runoff (especially during and after heavy rainfall events
19 such as hurricanes). Together with excess nitrogen runoff from agricultural practices or other
20 uses (e.g., fertilization of lawns or gardens), massive influxes of such nitrogen into watersheds
21 and sounds can lead to dramatic decreases in water oxygen and increases in algae blooms that
22 can cause extensive fish kills and damage to commercial fish and sea food harvesting.
23 Acidic deposition has played a major role in soil acidification in some areas of Sweden,
24 elsewhere in Europe, and in eastern North America. Soil acidification and its effects result from
25 deposition of nitrates, sulfates, and associated H+ ion. A major concern is that soil acidity will
26 lead to nutrient deficiency. Growth of tree species can be affected when high aluminum-to-
27 nutrient ratios limit uptake of calcium and magnesium and create a nutrient deficiency. Calcium
28 is essential in the formation of wood and the maintenance of cells (the primary plant tissues
29 necessary for tree growth), and it must be dissolved in soil water to be taken up by plants. Acidic
30 deposition can increase aluminum concentrations in soil water by lowering the pH in aluminum-
31 rich soils through dissolution and ion-exchange processes. Aluminum in soil can then be taken
March 2001
4-160
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
up by roots more readily than calcium because of its greater affinity for negatively charged
surfaces. Tree species can be adversely affected if altered Ca/Al ratios impair Ca or Mg uptake.
Overall, then, PM produced by human activities has the potential to cause the loss of
ecosystem biodiversity in ways that reduces the ability of ecosystems to provide the services that
society requires to sustain life. The major impacts of PM on ecosystems are the indirect effects
that occur through the soil and affect plant growth, vigor, and reproduction. Mineral nutrient
cycling can be altered by the deposition of heavy metals. The deposition of nitrogen and sulfur
and the acidifying effects of the two in association with the H+ ion in precipitation also alter
biogeochemical cycling, cause soil acidification, alter the Ca/Al ratio, and impact the growth of
vegetation and forest trees, in particular. Leaching of nitrates and other minerals through runoff
can impact coastal and aquatic wetlands and, thus, influence their ability to produce the products
and services necessary for human society.
4.6.2 Particulate Matter-Related Effects on Visibility
Visibility is defined as the degree to which the atmosphere is transparent to visible light and
the clarity and color fidelity of the atmosphere. Visual range is the farthest distance a black
object can be distinquished against the horizontal sky. Visibility impairment is any humanly
perceptible change in visibility. For regulatory purposes, visibility impairment, characterized by
light extinction, visual range, contrast, and coloration, is classified into two principal forms:
(1) "reasonably attributable" impairment, attributable to a single source or small group of
sources, and (2) regional haze, any perceivable change in visibility caused by a combination of
many sources over a wide geographical area.
Visibility is measured by human observation, light scattering by particles, the light
extinction-coefficient and parameters related to the light-extinction coefficient (visual range and
deciview scale), the light scattering coefficient, and fine PM concentrations. The air quality
within a sight path will affect the illumination of the sight path by scattering or absorbing solar
radiation before it reaches the Earth's surface. The rate of energy loss with distance from a beam
of light is the light extinction coefficient. The light extinction coefficient is the sum of the
coefficients for light absorption by gases (oag), light scattering by gases (asg), light absorption by
particles (aap), and light scattering by particles (asp). Atmospheric particles are frequently divided
into fine and coarse particles. Corresponding coefficients for light scattering and absorption by
March 2001 4-161 DRAFT-DO NOT QUOTE OR CITE
-------
1 fine and coarse particles are osfp and oafp and oscp and oacp, respectively. Visibility within a sight
2 path longer than approximately 100 km (60 mi) is affected by change in the optical properties of
3 the atmosphere over the length of the sight path.
4 Visibility impairment is associated with airborne particle properties, including size
5 distributions (i.e., fine particles in the 0.1- to 1.0-/zm size range) and aerosol chemical
6 composition, and with relative humidity. With increasing relative humidity, the amount of
7 moisture available for absorption by particles increases, thus causing the particles to increase in
8 both size and volume. As the particles increase in size and volume, the light scattering potential
9 of the particles also generally increases. Visibility impairment is greatest in the eastern United
10 States and Southern California. In the eastern United States, visibility impairment is caused
11 primarily by light scattering by sulfate aerosols and, to a lesser extent, by nitrate particles and
12 organic aerosols, carbon soot, and crustal dust. Haziness in the southeastern United States,
13 caused by increased atmospheric sulfate, has increased by ca. 80% since the 1950s and is greatest
14 in the summer months, followed by the spring and fall, and winter. Light scattering by nitrate
15 aerosols is the major cause of visibility impairment in Southern California. Nitrates contribute
16 about 40% to the total light extinction in Southern California and accounts for 10 to 20% of the
17 total extinction in other U.S. areas.
18 Organic particles are the second largest contributors to light extinction in most U.S. areas.
19 Organic carbon is the greatest cause of light extinction in the Pacific Northwest, Oregon, Idaho,
20 and Montana, accounting for 40 to 45% of the total extinction. Also, organic carbon contributes
21 between 15 to 20% to the total extinction in most of the western United States and 20 to 30% in
22 the remaining U.S. areas.
23 Coarse mass and soil, primarily considered "natural extinction", is responsible for some of
24 the visibility impairment in northern California and Nevada, Oregon, southern Idaho, and
25 western, Wyoming. Dust transported from Southern California and the subtropics has been
26 associated with regional haze in the Grand Canyon and other southwestern U.S. class I areas.
27
28 4.6.3 Particulate Matter-Related Effects on Materials
29 Building materials (metals, stones, cements, and paints) undergo natural weathering
30 processes from exposure to environmental elements (wind, moisture, temperature fluctuations,
31 sun light, etc.). Metals form a protective film that protects against environmentally induced
March 2001
4-162
DRAFT-DO NOT QUOTE OR CITE
-------
1 corrosion. The natural process of metal corrosion from exposure to natural environmental
2 elements is enhanced by exposure to anthropogenic pollutants, in particular SO2, rendering the
3 protective film less effective.
4 Dry deposition of SO2 enhances the effects of environmental elements on calcereous stones
5 (limestone, marble, and cement) by converting calcium carbonate (calcite) to calcium sulfate
6 dihydrate (gypsum). The rate of deterioration is determined by the SO2 concentration, the stone's
7 permeability and moisture content, and the deposition rate; however, the extent of the damage to
8 stones produced by the pollutant species apart from the natural weathering processes is uncertain.
9 Sulfur dioxide also has been found to limit the life expectancy of paints by causing discoloration
10 and loss of gloss and thickness of the paint film layer.
11 A significant detrimental effect of particle pollution is the soiling of painted surfaces and
12 other building materials. Soiling changes the reflectance of a material from opaque and reduces
13 the transmission of light through transparent materials. Soiling is a degradation process that
14 requires remediation by cleaning or washing, and, depending on the soiled surface, repainting.
15 Available data on pollution exposure indicates that particles can result in increased cleaning
16 frequency of the exposed surface and may reduce the life usefulness of the material soiled.
17 Attempts have been made to quantify the pollutants exposure levels at which materials damage
18 and soiling have been perceived. However, to date, insufficient data are available to advance our
19 knowledge regarding perception thresholds with respect to pollutant concentration, particle size,
20 and chemical composition.
21
22 4.6.4 Effects of Participate Matter on the Transmission of Solar Ultraviolet
23 Radiation and Global Warming Processes
24 Extensive potential future impacts on human health and the environment are projected to
25 occur because of increased transmission of solar ultraviolet radiation (UV-B) through the Earth's
26 atmosphere, secondary to stratospheric ozone depletion resulting from anthropogenic emissions
27 of chlorofluorcarbons (CFCs), halons, and certain other gases. However, the estimation of the
28 likely future extent of detrimental effects caused by increased penetration of solar UV-B to the
29 Earth's surface is complicated by atmospheric particle effects, which vary depending on size and
30 composition of particles that can differ substantially over different geographic areas and from
31 season to season over the same area. Also, atmospheric particles greatly complicate projections
March 2001
4-163
DRAFT-DO NOT QUOTE OR CITE
-------
1 of future trends in global warming processes because of emissions of greenhouse gases;
2 consequent increases in global mean temperature, and resulting changes in regional and local
3 weather patterns; and mainly deleterious (but some beneficial) location-specific human health
4 and environmental impacts.
5 The physical processes (i.e., scattering and absorption) responsible for airborne particle
6 effects on transmission of solar ultraviolet and visible radiation are the same as those responsible
7 for visibility degradation. Scattering of solar radiation back to space and absorption of solar
8 radiation determine the effects of an aerosol layer on solar radiation. The transmission of solar
9 UV-B radiation is affected strongly by atmospheric particles. Measured attenuations of UV-B
10 under hazy conditions range up to 37% of the incoming solar radiation. Measurements relating
11 variations in PM mass directly to UV-B transmission are lacking. Particles also can affect the
12 rates of photochemical reactions occurring in the atmosphere. Depending on the amount of
13 absorbing substances in the particles, photolysis rates either can be increased or decreased.
14 In addition to direct climate effects through the scattering and absorption of solar radiation,
15 particles also exert indirect effects on climate by serving as cloud condensation nuclei, thus
16 affecting the abundance and vertical distribution of clouds. The direct and indirect effects of
17 particles appear to have significantly offset the global warming effects caused by the buildup of
18 greenhouse gases because the onset of the Industrial Revolution, on a globally averaged basis.
19 However, because the lifetime of particles is much shorter than that required for complete mixing
20 within the Northern Hemisphere, the climate effects of particles generally are felt much less
21 homogeneously than are the effects of long-lived greenhouse gases.
22 Any effort to model the impacts of local alterations in particle concentrations on projected
23 global climate change or consequent local and regional weather patterns would be subject to
24 considerable uncertainty. This also would be the case for any projections of impacts of location-
25 specific airborne PM alterations on potential human health or environmental effects associated
26 with either increased atmospheric transmission of solar UV radiation or global warming
27 secondary to accumulation of stratospheric ozone-deleting substances or "greenhouse gases."
March 2001
4-164
DRAFT-DO NOT QUOTE OR CITE
-------
1
REFERENCES
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Abdullah, U. M.; Iqbal, M. Z. (1991) Response of automobile, stone and cement particulate matters on stomatal
clogging of plants. Geobios (Jodhpur, India) 18:196-202.
Aber, J. D.; Nadelhoffer, K. J.; Steudler, P.; Melillo, J. M. (1989) Nitrogen saturation in northern forest ecosystems:
excess nitrogen from fossil fuel combustion may stress the biosphere. Bioscience 39: 378-386.
Aber, J.; McDowell, W.; Nadelhoffer, K.; Magill, A.; Berntson, G.; Kamakea, M.; McNulty, S.; Currie, W.;
Rustad, L.; Fernandez, I. (1998) Nitrogen saturation in temperate forest ecosystems. BioScience 48: 921-934.
Adams, K. M.; Kavis, L. I., Jr.; Japar, S. M.; Pierson, W. R. (1989) Real-time, in situ measurements of atmospheic
optical absorption in the visible via photoacoustic spectroscopy - II. validation for atmospheric elemental
carbon aerosol. Atmos. Environ. 23: 693-700.
Alexander, M. (1977) Introduction to soil microbiology. 2nd ed. New York, NY: John Wiley & Sons- pp 225-230-
239-246; 272-286; 355-356.
Allen, M. F. (1991) The ecology of mycorrhizae. Cambridge, MA: Cambridge University Press.
Allen, E. R.; Cabrera, N.; Kim, J.-C. (1994) Atmospheric deposition studies in rural forested areas of the
southeastern United States. Presented at: 87th annual meeting & exhibition of the Air & Waste Management
Association; June; Cincinnati, OH. Pittsburgh, PA: Air & Waste Management Association- paper no
94-WP88.02.
Altshuller, A. P.; Linthurst, R. A., eds. (1984) Acidic deposition phenomenon and its effects: critical assessment
review papers; volume I, atmospheric sciences; volume II, effects sciences. Washington, DC: U.S.
Environmental Protection Agency, Office of Research and Development; EPA report nos.
EPA/600/8-83/016aF-bF. Available from: NTIS, Springfield, VA; PB85-100030 and PB85-100048.
Andersen, C. P.; Rygiewicz, P. T. (1991) Stress interactions and mycorrhizal plant response: understanding carbon
allocation priorities. Environ. Pollut. 73: 217-244.
Antonovics, J.; Bradshaw, A. D.; Turner, R. G. (1971) Heavy metal tolerance in plants. In: Cragg, J. B., ed.
Advances in ecological research. Volume 7. London, United Kingdom: Academic Press; pp. 1-85.
April, R.; Newton, R. (1992) Mineralogy and mineral weathering. In: Johnson, D. W.; Lindberg, S. E., eds.
Atmospheric deposition and forest nutrient cycling: a synthesis of the integrated forest study. New York, NY:
Springer-Verlag; pp. 378-425. (Billings, W. D.; Golley, F.; Lange, O. L.; Olson, J. S.; Remmert, H., eds.
Ecological studies: analysis and synthesis: v. 91).
Askey, A.; Lyon, S. B.; Thompson, G. E.; Johnson, J. B.; Wood, G. C.; Sage, P. W.; Cooke, M. J. (1993) The effect
of fly-ash particulates on the atmospheric corrosion of zinc and mild steel. Corros. Sci. 34: 1055-1081.
Ausset, P.; Bannery, F.; Del Monte, M.; Lefevre, R. A. (1998) Recording of pre-industrial atmospheric environment
by ancient crusts on stone monuments. Atmos. Environ. 32: 2859-2863.
Ayensu, E.; Van R. Claasen, D.; Collins, M.; Dealing, A.; Fresco, L.; Gadgil, M.; Gitay, H.; Glaser, G.; Juma, C.;
Krebs, J.; lenton, R.; Lubchenco, J.; McNeely, J. A.; Mooney, H. A.; Pinstrup-Andersen, P.; Ramos, M.;
Raven, P.; Reid, W. V.; Samper, C.; Sarukhan, J.; Schei, P.; Tundisi, J. G.; Watson, R. T.; Guanhua, X.;
Zakri, A. H. (1999) International ecosystem assessment. Science 286: 685-686.
Babich, H.; Stotzky, G. (1978) Effects of cadmium on the biota: influence of environmental factors. In: Perlman, D.,
ed. Advances in applied microbiology, volume 23. New York, NY: Academic Press; pp. 55-117.
Baedecker, P. A.; Edney, E. O.; Moran, P. J.; Simpson, T. C.; Williams, R. S. (1991) Effects of acidic deposition on
materials. In: Irving, P. M., ed. Acidic deposition: state of science and technology, volume III: terrestrial,
materials, health and visibility effects. Washington, DC: The U.S. National Acid Precipitation Assessment
Program. (State of science and technology report no. 19).
Baedecker, P. A.; Reddy, M. M.; Reimann, K. J.; Sciammarella, C. A. (1992) Effects of acidic deposition on the
erosion of carbonate stone-experimental results from the U.S. National Acid Precipitation Assessment
Program (NAPAP). Atmos. Environ. Part B 26: 147-158.
Barone, J. B.; Cahill, T. A.; Eldred, R. A.; Flocchini, R. G.; Shadoan, D. J.; Dietz, T. M. (1978) A multivariate
statistical analysis of visibility degradation at four California cities. Atmos. Environ. 12: 2213-2221.
Barton, K. (1958) The influence of dust on atmospheric corrosion of metals. Werkst. Korros. 8/9: 547-549.
Bellan, L. M.; Salmon, L. G.; Cass, G. R. (2000) A study on the human ability to detect soot deposition onto works
of art. Environ. Sci. Technol. 34: 1946-1952.
Bewley, R. J. F.; Parkinson, D. (1984) Effects of sulphur dioxide pollution on forest soil microorganisms. Can J
Microbiol. 30: 179-185.
Biefer, G. J. (1981) Atmospheric corrosion of steel in the Canadian Arctic. Mater. Perform. 20: 16-19.
March 2001
4-165
DRAFT-DO NOT QUOTE OR CITE
-------
1 Billings, W. D. (1978) Plants and the ecosystem. 3rd ed. Belmont, CA: Wadsworth Publishing Company, Inc.;
2 pp. 1-62,83-108.
3 Binenko, V. I.; Harshvardhan, H. (1993) Aerosol effects in radiation transfer. In: Jennings, S. G., ed.
4 Aerosol effects on climate. Tucson, AZ: University of Arizona Press; pp. 190-232.
5 Binkley, D.; Valentine, D.; Wells, C.; Valentine, U. (1989) An empirical analysis of the factors contributing to
6 20-year decrease in soil pH in an old-field plantation of loblolly pine. Biogeochemistry 8: 39-54.
7 Bjom, L. O. (1996) Effects of ozone depletion and increased UV-B on terrestrial ecosystems. Int. J. Environ. Stud.
8 51:217-243.
9 Blackburn, T. E. (1992) Solar photolysis of ozone to singlet D oxygen atoms. J. Geophys. Res. 97: 10,109-10,117.
10 Blandford, J. C. (1994) Transmissometer data reduction and validation (IMPROVE protocol). Fort Collins, CO:
11 Air Resource Specialists, Inc.; technical instruction no. 4400-5000.
12 Blaschke, H. (1990) Mycorrhizal populations and fine root development on Norway spruce exposed to controlled
13 doses of gaseous pollutants and simulated acidic rain treatments. Environ. Pollut. 68: 409-418.
14 Bobbink, R. (1998) Impacts of tropospheric ozone and airborne nitrogenous pollutants on natural and semi-natural
15 ecosystems: a commentary. New Phytol. 139:161-168.
16 Bobbink, R.; Homung, M; Roelofs, J. G. M. (1998) The effects of air-borne nitrogen pollutants on species diversity
17 in natural and semi-natural European vegetation. J. Ecol. 86: 717-738.
18 Bolle, H.-J.; Seiler, W.; Bolin, B. (1986) Other greenhouse gases and aerosols: assessing their role for atmospheric
19 radiative transfer. In: Bolin, B.; Doos, B. R.; Jager, J.; Warrick, R. A., eds. The greenhouse effect, climatic
20 change, and ecosystems. Chichester, United Kingdom: John Wiley and Sons; pp. 157-203. (SCOPE 29).
21 Bondietti, E. A.; McLaughlin, S. B. (1992) Evidence of historical influences of acidic deposition on wood and soil
22 chemistry. In: Johnson, D. W.; Lindberg, S. E., eds. Atmospheric deposition and forest nutrient cycling:
23 a synthesis of the integrated forest study. New York, NY: Springer-Verlag: pp. 358-377. (Billings, W. D.;
24 Golley, F.; Lange, O. L.; Olson, J. S.; Remmert, H., eds. Ecological studies analysis and synthesis: v. 91).
25 Bowden, R. D.; Geballe, G. T.; Bowden, W. B. (1989) Foliar uptake of 15N from simulated cloud water by red
26 spruce (Picea rubens) seedlings. Can. J. For. Res. 19: 382-386.
27 Bradley, R.; Burt, A. J.; Read, D. J. (1981) Mycorrhizal infection and resistance to heavy metal toxicity in Calluna
28 vulgaris. Nature (London) 292: 335-337.
29 Brandt, C. J.; Rhoades, R. W. (1972) Effects of limestone dust accumulation on composition of a forest community.
30 Environ. Pollut. 3:217-225.
31 Brandt, C. J.; Rhoades, R. W. (1973) Effects of limestone dust accumulation on lateral growth of forest trees.
32 Environ. Pollut. 4: 207-213.
33 Briihl, C.; Crutzen, P. J. (1988) On the disproportionate role of tropospheric ozone as a filter against solar UV-B
34 radiation. Geophys. Res. Lett. 16: 703-706.
35 Bruhn, J. N. (1980) Effects of oxidant air pollution on Ponderosa and Jeffrey pine foliage decomposition
36 [dissertation]. Berkeley, CA: University of California; Dissertation Abstracts International v. 42-07B.
37 Brumme, R.; Leimcke, U.; Matzner, E. (1992) Interception and uptake of NH,, and NO3 from wet deposition by
38 above-ground parts of young beech (Fagus silvatica L.) trees. Plant Soil 142: 273-279.
39 Brunei, J.; Diekmann, M.; Falkengren-Grerup, U. (1998) Effects of nitrogen deposition on field layer vegetation in
40 • south Swedish oak forests. Environ. Pollut. (Oxford, U.K.) 102(suppl. 1): 35-40.
41 Bugini, R.; Tabasso, M. L.; Realini, M. (2000) Rate of formation of black crusts on marble. A case study. J. Cult.
42 Heritage 1:111-116.
43 Butlin, R. N.; Coote, A. T.; Devenish, M.; Hughes, I. S. C.; Hutchens, C. M.; Irwin, J. G.; Lloyd, G. O.; Massey,
44 S. W.; Webb, A. H.; Yates, T. J. S. (1992a) Preliminary results from the analysis of metal samples from the
45 National Materials Exposure Programme (NMEP). Atmos. Environ. Part B 26: 199-206.
46 Butlin, R. N.; Coote, A. T.; Devenish, M.; Hughes, I. S. C.; Hutchens, C. M.; Irwin, J. G.; Lloyd, G. O.; Massey,
47 s. W.; Webb, A. H.; Yates, T. J. S. (1992b) Preliminary results from the analysis of stone tablets from the
48 National Materials Exposure Programme (NMEP). Atmos. Environ. Part B 26:189-198.
49 Bytnerowicz, A.; Fenn, M. E. (1996) Nitrogen deposition in California forests: a review. Environ. Pollut.
50 92:127-146.
51 Cahill, T. A.; Eldred, R. A.; Wakabayashi, P. H. (1996) Trends in fine particle concentrations at Great Smoky
52 Mountains National Park. For presentation at: Annual meeting of the Air & Waste Management Association;
53 June; Nashville, TN. Pittsburgh, PA: Air & Waste Management Association; submitted paper no.
54 96-MP1A.05.
March 2001
4-166
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Campbell, G. G.; Schurr, G. G.; Slawikowski, D. E.; Spence, J. W. (1974) Assessing air pollution damage to
coatings. J. Paint Technol. 46: 59-71.
Camuffo, D. (1995) Physical weathering of stones. In: Saiz-Jimenez, C., ed. The deterioration of monuments:
proceedings of the 2nd international symposium on biodeterioration and biodegradation; January 1994;
Sevilla, Spain. Sci. Total Environ. 167: 1-14.
Cape, J. N. (1993) Direct damage to vegetation caused by acid rain and polluted cloud: definition of critical levels
for forest trees. Environ. Pollut. 82: 167-180.
Cape, J. N.; Sheppard, L. J.; Fowler, D.; Harrison, A. F.; Parkinson, J. A.; Dao, P.; Paterson, I. S. (1992)
Contribution of canopy leaching to sulphate deposition in a Scots pine forest. Environ. Pollut. 75: 229-236.
Carey, W. F. (1959) Atmospheric deposits in Britain: a study of dinginess. Int. J. Air Pollut. 2: 1-26.
Ceccotti, S. P.; Messick, D. L. (1997) A global review of crop requirements, supply, and environmental impact on
nutrient sulphur balance. In: Cram, W. J.; De Kok, L. J.; Stulen, I.; Brunold, C.; Rennenberg, H., eds.
Sulphur metabolism in higher plants. Leiden, The Netherlands: Backhuys Publishers; pp. 155-163.
Chameides, W. L.; Yu, H.; Liu, S. C.; Bergin, M.; Zhou, X.; Meams, L.; Wang, G.; Kiang, C. S.; Saylor, R. D.;
Luo, C.; Huang, Y.; Steiner, A.; Giorgi, F. (1999) Case study of the effects of atmospheric aerosols and
regional haze on agriculture: an opportunity to enhance crop yields in China through emission controls?
Proc. Natl. Acad. Sci. U. S. A. 96: 13626-13633.
Chapin, F. S., III. (1991) Integrated responses of plants to stress: a centralized system of physiological responses.
Bioscience 41: 29-36.
Chapin, F. S., Ill; Bloom,.A. J.; Field, C. B.; Waring, R. H. (1987) Plant responses to multiple environmental
factors. BioScience 37: 49-57.
Chapin, F. S., Ill; Sala, O. E.; Burke, I. C.; Grime, J. P.; Hooper, D. U.; Lauenroth, W. K.; Lombard, A.; Mooney,
H. A.; Mosier, A. R.; Naeem, S.; Pacala, S. W.; Roy, J.; Steffen, W. L.; Tilman, D. (1998) Ecosystem
consequences of changing biodiversity. BioScience 48:45-52.
Charlson, R. J.; Lovelock, J. E.; Andreae, M. O.; Warren, S. G. (1987) Oceanic phytoplankton, atmospheric
sulfphur, cloud albedo and climate. Nature (London) 326: 655-661.
Charlson, R. J.; Schwartz, S. E.; Hales, J. M.; Cess, R. D.; Coakley, J. A., Jr.; Hansen, J. E.; Hofmann, D. J. (1992)
Climate forcing by anthropogenic aerosols. Science (Washington, DC) 255: 423-430.
Chestnut, L. (1997) Methodology for estimating values for changes in visibility at national parks [draft
memorandum]. Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards; contract no. 68-D-98-001; EPA air docket A-96-56, document no. VI-B-09-(ooo).
Chestnut, L. G.; Dennis, R. L. (1997) Economic benefits of improvements in visibility: acid rain. Provisions of the
1990 clean air act amendments. J. Air Waste Manage. Assoc. 47: 395-402.
Chevone, B. I.; Herzfeld, D. E.; Krupa, S. V.; Chappelka, A. H. (1986) Direct effects of atmospheric sulfate
deposition on vegetation. J. Air Pollut. Control Assoc. 36: 813-815.
Choudhury, B. J. (1987) Relationships between vegetation indices, radiation absorption, and net photosynthesis
evaluated by a sensitivity analysis. Remote Sens. Environ. 22: 209-233.
Clarkson, D. T.; Sanderson, J. (1971) Inhibition of the uptake and long-distance transport of calcium by aluminum
and other polyvalent cations. J. Exp. Bot. 22: 837-851.
Clevenger, T. E.; Saiwan, C.; Koirtyohann, S. R. (1991) Lead speciation of particles on air filters collected in the
vicinity of a lead smelter. Environ. Sci. Technol. 25: 1128-1133.
Cole, D. W.; Johnson, D. W. (1977) Atmospheric sulfate additions and cation leaching in a Douglas fir ecosystem.
Water Resour. Res. 13: 313-317.
Cooke, R. U.; Gibbs, G. B. (1994) Crumbling heritage? Studies of stone weathering in polluted atmospheres.
Atmos. Environ. 28: 1355-1356.
Costanza, R.; d'Arge, R.; De Groot, R.; Farber, S.; Grasso, M.; Hannon, B.; Limburg, K.; Naeem, S.; O'Neill, R. V.;
Paruelo, J.; Raskin, R. G.; Sutton, P.; Van Den Belt, M. (1997) The value of the world's ecosystem services
and natural capital. Nature (London) 387: 253-259.
Cotrufo, M. F.; De Santo, A. V.; Alfani, A.; Bartoli, G.; De Cristofaro, A. (1995) Effects of urban heavy metal
pollution on organic matter decomposition in Quercus ilex L. woods. Environ. Pollut. 89: 81-87.
Cowling, J. E.; Roberts, M. E. (1954) Paints, varnishes, enamels, and lacquers. In: Greathouse, G. A.; Wessel, C. J.,
eds. Deterioration of materials: causes and preventive techniques. New York, NY: Reinhold Publishing
Corporation; pp. 596-645.
March 2001
4-167
DRAFT-DO NOT QUOTE OR CITE
-------
1 Cramer, S. D.; McDonald, L. G.; Bhagia, G.; Flinn, D. R.; Linstrom, P. J.; Carter, J. P. (1989) Effects of acid
2 deposition on the atmospheric corrosion of structural metals. Research Triangle Park, NC: U.S.
3 Environmental Protection Agency; internal report.
4 Creighton, P. J.; Lioy, P. J.; Haynie, F. H.; Lemmons, T. J.; Miller, J. L.; Gerhart, J. (1990) Soiling by atmospheric
5 aerosols in an urban industrial area. J. Air Waste Manage. Assoc. 40: 1285-1289.
6 Cronan, C. S.; Grigal, D. F. (1995) Use of calcium/aluminum ratios as indicators of stress in forest ecosystems.
7 J. Environ. Qual. 24:209-226.
8 Cullen, J. J.; Neale, P. J. (1997) Biological weighting functions for describing the effects of ultraviolet radiation on
9 aquatic ecosystems. In: Hader, D.-P., ed. The effects of ozone depletion on aquatic ecosystems. Austin, TX:
10 Academic Press, pp. 97-118.
11 Dahlgren, R. A. (1994) Soil acidification and nitrogen saturation from weathering of ammonium-bearing rock.
12 Nature 368: 838-841.
13 Daily, G. C. (1997) Introduction: what are ecosystem services? In: Daily, G. C., ed. Nature's services: societal
14 dependence on natural ecosystems. Washington, DC: Island Press; pp. 1-10.
15 Daily, G. C.; Ehrlich, P. R. (1999) Managing Earth's ecosystems: an interdisciplinary challenge. Ecosystems
16 2:277-280.
17 Darley, E. F. (1966) Studies on the effect of cement-kiln dust on vegetation. J. Air Pollut. Control Assoc.
18 16:145-150.
19 Dassler, H.-G.; Ranft, H.; Rehn, K.-H. (1972) Zur Widerstandsfahigkeit von Geholzen gegeniiber
20 Fluorverbindungen und Schwefeldioxid [The susceptibility of woody plants exposed to fluorine compounds
21 and SOJ. Flora (Jena) 161:289-302.
22 Dattner, S. L. (1995) Urban visibility in the El Paso del Norte airshed. Presented at: 88th annual meeting &
23 exhibition of the Air & Waste Management Association, volume 3B. Basic sciences: air quality and noise;
24 June; San Antonio, TX. Pittsburgh, PA: Air & Waste Management Association; paper no. 95-WA76.02.
25 Davidson, C. L; Tang, W.; Finger, S.; Etyemezian, V.; Striegel, M. F.; Sherwood, S. I. (2000) Soiling patterns on a
26 tall limestone building: changes over 60 years. Environ. Sci. Technol. 34: 560-565.
27 Davis, G. D.; Shaw, B. A.; Arah, C. O.; Fritz, T. L.; Moshier, W. C.; Simpson, T. C.; Moran, P. J.; Zankel, K. L.
28 (1990) Effects of SO2 deposition on painted steel surfaces. Surf. Interface Anal. 15: 107-112.
29 Day, D.; Malm, W. C.; Kreidenweis, S. M. (1996) Aerosol light scattering measurements as a function of relative
30 humidity at Great Smoky Mountains National Park. In: 89th annual meeting & exhibition of the Air & Waste
31 Management Association; June; Nashville, TN. Pittsburgh, PA: Air & Waste Management Association;
32 paper no. 96-TA1B.06.
33 Day, D. E.; Malm, W. C.; Kreidenweis, S. M. (2000) Aerosol light scattering measurements as a function of relative
34 humidity. J. Air Waste Manage. Assoc. 50: 710-716.
35 Del Monte, M.; Vittori, O. (1985) Air pollution and stone decay: the case of Venice. Endeavour 9: 117-122.
36 Diakumaku, E.; Gorbushina, A. A.; Krumbein, W. E.; Panina, L.; Soukharjevski, S. (1995) Black fungi in marble
37 and limestones-an aesthetical, chemical and physical problem for the conservation of monuments.
38 In: Saiz-Jimenez, C., ed. The deterioration of monuments: proceedings of the 2nd international symposium
39 on biodeterioration and biodegradation; January 1994; Sevilla, Spain. Sci. Total Environ. 167: 295-304.
40 Dickerson, R. R.; Kondragunta, S.; Stenchikov, G.; Civerolo, K. L.; Doddridge, B. G.; Holben, B. N. (1997)
41 The impact of aerosols on solar ultraviolet radiation and photochemical smog. Science (Washington, DC)
42 278:827-830.
43 Dobbins, R. A.; Mulholland, G. W.; Bryner, N. P. (1994) Comparison of a fractal smoke optics model with light
44 extinction measurements. Atmos. Environ. 28: 889-897.
45 Dolske, D. A. (1995) Deposition of atmospheric pollutants to monuments, statues, and buildings. Sci. Total
46 Environ. 167: 15-31.
47 Dreisinger, B. R.; McGovern, P. C. (1970) Monitoring atmospheric sulphur dioxide and correlating its effects on
48 crops and forests in the Sudbury area. In: Linzon, S. N., ed. Impact of air pollution on vegetation conference;
49 April; Toronto, ON, Canada. Pittsburgh, PA: Air Pollution Control Association.
50 Driscoll, C. T.; Van Dreason, R. (1993) Seasonal and long-term temporal patterns in the chemistry of Adirondack
51 lakes. Water Air Soil Pollut. 67: 319-344.
52 Driscoll, C. T.; Wyskowski, B. J.; DeStaffan, P.; Newton, R. M. (1989) Chemistry and transfer of aluminum in a
53 forested watershed in the Adirondack region of New York, USA. In: Lewis, T. E., ed. Environmental
54 chemistry and toxicology of aluminum. Chelsea, MI: Lewis Publishers, Inc.; pp. 83-105.
March 2001
4-168
DRAFT-DO NOT QUOTE OR CITE
-------
1 Driscoll, C. T.; Newton, R. M.; Gubala, C. P.; Baker, J. P.; Christensen, S. W. (1991) Adirondack mountains.
2 In: Charles, D. F., ed. Acidic deposition and aquatic ecosystems: regional case studies. New York, NY:
3 Springer-Verlag; pp. 133-202.
4 Edney, E. O. (1989) Paint coatings: controlled field and chamber experiments. Research Triangle Park, NC:
5 U.S. Environmental Protection Agency, Atmospheric Research and Exposure Assessment Laboratory; report
6 no. EPA/600/3-89/032. Available from: NTIS, Springfield, VA; PB89-189849.
7 Edney, E. O.; Stiles, D. C.; Corse, E. W.; Wheeler, M. L.; Spence, J. W.; Haynie, F. H.; Wilson, W. E., Jr. (1988)
8 Field study to determine the impact of air pollutants on the corrosion of galvanized steel. Mater. Perform.
9 27:47-50.
10 Edney, E. O.; Cheek, S. F.; Corse, E. W.; Spence, J. W.; Haynie, F. H. (1989) Atmospheric weathering caused by
11 dry deposition of acidic compounds. J. Environ. Sci. Health Part A 24: 439-457.
12 Eldering, A.; Cass, G. R. (1996) Source-oriented model for air pollutant effects on visibility. J. Geophys. Res.
13 [Atmos.]: 101:19,343-19,369.
14 Eldering, A.; Larson, S. M.; Hall, J. R.; Hussey, K. J.; Cass, G. R. (1993) Development of an improved image
15 processing based visibility model. Environ. Sci. Technol. 27: 626-635.
16 Eldering, A.; Cass, G. R.; Moon, K. C. (1994) An air monitoring network using continuous particle size distribution
17 monitors: connecting pollutant properties to visibility via Mie scattering calculations. Atmos. Environ.
18 28:2733-2749.
19 Eldering, A.; Hall, J. R.; Hussey, K. J.; Cass, G. R. (1996) Visibility model based on satellite-generated landscape
20 data. Environ. Sci. Tech. 30: 361 -370.
21 Eldred, R. A.; Cahill, T. A. (1994) Trends in elemental concentrations of fine particles at remote sites in the
22 United States of America. Atmos. Environ. 28: 1009-1019.
23 Eldred, R. A.; Cahill, T. A.; Feeney, P. J. (1993) Comparison of independent measurements of sulfur and sulfate in
24 the IMPROVE network. Presented at: 86th annual meeting and exhibition of the Air & Waste Management
25 Association; June; Denver, CO. Pittsburgh, PA: Air & Waste Management Association; paper no.
26 93-RA-110.02.
27 Eldred, R. A.; Cahill, T. A.; Flocchini, R. G. (1997) Composition of PM25 and PM10 aerosols in the IMPROVE
28 network. In: Special issue on aerosols and atmospheric optics-part I, [published papers from the international
29 specialty conference; September 1994; Snowbird, UT]. J. Air Waste Manage. Assoc. 47: 194-203.
30 Ellenberg, H. (1987) Floristic changes due to eutrophication. In: Asman, W. A. H.; Diederen, S. M. A., eds.
31 Ammonia and acidification: proceedings of a symposium of the European Association for the Science of Air
32 Pollution (EURASAP); April; Bilthoven, The Netherlands. European Association for the Science of Air
33 Pollution; pp. 301-308.
34 Eller, B. M. (1977) Road dust induced increase of leaf temperature. Environ. Pollut. 13: 99-107.
35 Eriksson, P.; Johansson, L.-G.; Strandberg, H. (1993) Initial stages of copper corrosion in humid air containing SO2
36 and NO2. J. Electrochem. Soc. 140: 53-59.
37 Erlick, C.; Frederick, J. E.; Saxena, V. K.; Wenny, B. N. (1998) Atmospheric transmission in the ultraviolet and
38 visible: aerosols in cloudy atmospheres. J. Geophys. Res. 103: 31,541-31,556.
39 Estupinan, J. B.; Raman, S.; Crescenti, G. H.; Streicher, J. J.; Barnard, W. F. (1996) Effects of clouds and haze on
40 UV-B radiation. J. Geophys. Res. 101: 16,807-16,816.
41 Evans, E. G.; Pitchford, M. (1991) Federal visibility monitoring in the eastern United States. Presented at:
42 84th annual meeting and exhibition of the Air & Waste Management Association; June; Vancouver, BC,
43 Canada. Pittsburgh, PA: Air & Waste Management Association; paper no. 91-49.6.
44 Falkengren-Grerup, U. (1987) Long-term changes in pH of forest soils in southern Sweden. Environ. Pollut.
45 43:79-90.
46 Falkengren-Grerup, U. (1998) Nitrogen response of herbs and graminoids in experiments with simulated acid soil
47 solution. Environ. Pollut. 102(suppl. 1): 93-99.
48 Farber, R. J.; Welsing, P. R.; Rozzi, C. (1994) PM10 and ozone control strategy to improve visibility in the
49 Los Angeles Basin. Atmos. Environ. 28: 3277-3283.
50 Farmer, A. M. (1993) The effects of dust on vegetation-a review. Environ. Pollut. 79: 63-75.
51 Fenn, M. E.; Poth, M. A.; Johnson, D. W. (1996) Evidence for nitrogen saturation in the San Bernardino Mountains
52 in southern California. For. Ecol. Manage. 82: 211-230
53 Fenn, M. E.; Poth, M. A.; Aber, J. D.; Baron, J. S.; Bormann, B. T.; Johnson, D. W.; Lemly, A. D.; McNulty, S. G.;
54 Ryan, D. F.; Stottlemyer, R. (1998) Nitrogen excess in North American ecosystems: predisposing factors,
55 ecosystem responses, and management strategies. Ecol. Appl. 8: 706-733.
March 2001
4-169
DRAFT-DO NOT QUOTE OR CITE
-------
1 Fink, F. W.; Buttner, F. H.; Boyd, W. K. (1971) Final report on technical-economic evaluation of air-pollution
2 corrosion costs on metals in the U.S. Columbus, OH: Battelle Memorial Institute, Columbus Laboratories;
3 EPA contract no. CPA 70-86; EPA report no. APTD-0654. Available from: NTIS, Springfield, VA;
4 PB-198-453.
5 Flowers, E. C.; McCormick, R. A.; Kurfis, K. R. (1969) Atmospheric turbidity over the United States, 1961-1966.
6 J. Appl. Meteorol. 8: 955-962.
7 Fochtman, E. G.; Langer, G. (1957) Automobile paint damaged by airborne iron particles. J. Air Pollut. Control
8 Assoc. 6: 243-247.
9 Foster, J. R. (1991) Effects of organic chemicals in the atmosphere on terrestrial plants. In: Moser, T. J.; Barker,
10 J. R.; Tingey, D. T., eds. Ecological exposure and effects of airborne toxic chemicals: an overview. Corvallis,
11 OR: U.S. Environmental Protection Agency, Environmental Research Laboratory; pp. 60-89; EPA report no.
12 EPA-600/3-91-001. Available from: NTIS, Springfield, VA; PB91-148460.
13 Foster, N. W.; Nicolson, J. A.; Hazlett, P. W. (1989) Temporal variation in nitrate and nutrient cations in drainage
14 waters from a deciduous forest. J. Environ. Qual. 18: 238-244.
15 Fox, D.; McDonald, K.; Zannetti, P.; Nejedley, Z. (1997) Impact of north-western emission changes on visibility in
16 the Rocky Mountain parks. Presented at: Air & Waste Management Association's 90th annual meeting &
17 exhibition; June; Toronto, Ontario, Canada. Pittsburgh, PA: Air & Waste Management Association; paper
18 no. 97-WA70.03.
19 Fox, D. G.; Malm, W. C.; Mitchell, B.; Fisher, R. W. (1999) Where there's fire, there's smoke: fine particulate and
20 regional haze. EM November: 15-24.
21 Frederick, J. E.; Qu, Z.; Booth, C. R. (1998) Ultraviolet radiation at sites on the Antarctic coast. J. Photochem.
22 Photobiol. 668:183-190.
23 Frcedman, B.; Hutchinson, T. C. (1981) Sources of metal and elemental contamination of terrestrial environments.
24 In: Lepp, N. W., ed. Effect of heavy metal pollution on plants. Volume 2: metals in the environment. London,
25 United Kingdom: Applied Science Publishers; pp. 35-94.
26 Freeman, A. M., III. (1993) The measurement of environmental and resource values: theory and methods.
27 Washington, DC: Resources for the Future; pp. 1-27.
28 Freudenburg, W. R.; Alario, M. (1999) What ecologists can learn from nuclear scientists. Ecosystems 2: 286-291.
29 Fuller, K. A.; Malm, W. C.; Kreidenweis, S. M. (1999) Effects of mixing on extinction by carbonaceous particles.
30 J. Geophys. Res. 104:15,941-15,954.
31 Furst, P.; Krause, G. H. M.; Hein, D.; Delschen, T.; Wilmers, K. (1993) PCDD/PCDF in cow's milk in relation to
32 their levels in grass and soil. Chemosphere 27: 1349-1357.
33 Gabruk, R. S.; Sykes, R. I.; Seigneur, C.; Pai, P.; Gillespie, P.; Bergstrom, R. W.; Saxena, P. (1999) Evaluation of
34 the reactive and optics model of emissions (ROME). Atmos. Environ. 33: 383-399.
35 Gaggi, C.; Bacci, E.; Calamari, D.; Fanelli, R. (1985) Chlorinated hydrocarbons in plant foliage: an indication of the
36 tropospheric contamination level. Chemosphere 14:1673-1686.
37 Garcia-Valles, M.; Vendrell-Saz, M.; Molera, J.; Blazquez, F. (1998) Interaction of rock and atmosphere: patinas on
38 Mediterranean monuments. Environ. Geol. 36(1-2): 137-149.
39 Garner, J. H. B. (1991) Ozone exposure and nitrogen loading in the forests of eastern North America. In: Berglund,
40 R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment: papers from an
41 international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management
42 Association; pp. 289-310. (A&WMA transactions series no. TR-19).
43 Garner, J. H. B. (1994) Nitrogen oxides, plant metabolism, and forest ecosystem response. In: Alscher, R. G.;
44 Wellburn, A. R., eds. Plant responses to the gaseous environment: molecular, metabolic and physiological
45 aspects, [3rd international symposium on air pollutants and plant metabolism]; June 1992; Blacksburg, VA.
46 London, United Kingdom: Chapman & Hall; pp. 301 -314.
47 Gamer, J. H. B.; Pagano, T.; Cowling, E. B. (1989) An evaluation of the role of ozone, acid deposition, and other
48 airborne pollutants in the forests of eastern North America. Asheville, NC: U.S. Department of Agriculture,
49 Forest Service, Southeastern Forest Experiment Station; general technical report SE-59.
50 Garten, C. T., Jr. (1988) Fate and distribution of sulfur-35 in yellow poplar and red maple trees. Oecologia
51 76:43-50.
52 Garten, C. T., Jr.; Hanson, P. J. (1990) Foliar retention of 15N-nitrate and 15N-ammonium by red maple
53 (Acer rubrum) and white oak (Quercus alba) leaves from simulated rain. Environ. Exp. Bot. 30: 333-342.
54 Gawel, J. E.; Ahner, B. A.; Friedland, A. J.; Morel, F. M. M. (1996) Role for heavy metals in forest decline
55 indicated by phytochelatin measurements. Nature 381: 64-65.
March 2001
4-170
DRAFT-DO NOT QUOTE OR CITE
-------
1 Ghedini, N.; Gobbi, G.; Sabbioni, C.; Zappia, G. (2000) Determination of elemental and organic carbon on
2 damaged stone monuments. Atmos. Environ. 34: 4383-4391.
3 . Gildon, A.; Tinker, P. B. (1983) Interactions of vesicular-arbuscular mycorrhizal infection and heavy metals in
4 plants: I. the effects of heavy metals on the development of vesicular-arbuscular mycorrhizas. New Phytol.
5 95:247-261.
6 Gilliam, F. S.; Adams, M. B.; Yurish, B. M. (1996) Ecosystem nutrient responses to chronic nitrogen inputs at
7 Fernow Experimental Forest, West Virginia. Can. J. For. Res. 26: 196-205.
8 Gobbi, G.; Zappia, G.; Sabbioni, C. (1998) Sulphite quantification on damaged stones and mortars. Atmos. Environ.
9 32:783-789.
10 Goodwin, J. E.; Sage, W.; Tilly, G. P. (1969) Study of erosion by solid particles. Proc. Inst. Mech. Eng.
11 184:279-292.
12 Graedel, T. E.; Nassau, K.; Franey, J. P. (1987) Copper patinas formed in the atmosphere—I. introduction.
13 Corros. Sci. 27: 639-657.
14 Grand Canyon Visibility Transport Commission. (1996) Report of the Grand Canyon Visibility Transport
15 Commission to the United States Environmental Protection Agency.
16 Grant, L. D. (1989) Health effects issues associated with regional and global air pollution problems. In: Proceedings
17 of the changing atmosphere: implications for global security; June, 1988; Toronto, Canada. Geneva,
18 Switzerland: World Meteorological Organization; report no. WMO-No. 710.
19 Gray, H. A.; Cass, G. R. (1998) Source contributions to atmospheric fine carbon particle concentrations.
20 Atmos. Environ. 32: 3805-3825.
21 Greenberg, B. M.; Wilson, M. I.; Huang, X.-D.; Duxbury, C. L.; Gerhardt, K. E.; Gensemer, R. W. (1997)
22 The effects of ultraviolet-B radiation on higher plants. In: Wang, W.; Gorsuch, J. W.; Hughes, J. S., eds.
23 Plants for Environmental Studies. New York, NY: Lewis Publishers; pp. 1-35.
24 Guderian, R. (1977) Accumulation of pollutants in plant organs. In: Air pollution: phytotoxicity of acidic gases and
25 its significance in air pollution control. Berlin, Germany: Springer-Verlag; pp. 66-74. (Ecological studies:
26 v. 22).
27 Guderian, R., ed. (1985) Air pollution by photochemical oxidants: formation, transport, control, and effects on
28 plants. New York, NY: Springer-Verlag. (Billings, W. D.; Golley, F.; Lange, O. L.; Olson, J. S.;
29 Remmert, H., eds. Ecological studies: analysis and synthesis: v. 52).
30 Guderian, R. (1986) Terrestrial ecosystems: particulate deposition. In: Legge, A. H.; Krupa, S. V., eds.
31 Air pollutants and their effects on the terrestrial ecosystem. New York, NY: John Wiley & Sons; pp.
32 339-363. (Advances in environmental science and technology: v. 18).
33 Gundersen, P.; Callesen, I.; De Vries, W. (1998a) Nitrate leaching in forest ecosystems is related to forest floor C/N
34 ratios. Environ. Pollut. 102(suppl. 1): 403-407.
35 Gundersen, P.; Emmett, B. A.; Kjonaas, O. J.; Koopmans, C. J.; Tietema, A. (1998b) Impact of nitrogen deposition
36 on nitrogen cycling in forests: a synthesis of NITREX data. For. Ecol. Manage. 101: 37-55.
37 Hamilton, R. S.; Revitt, D. M.; Vincent, K. J.; Butlin, R. N. (1995) Sulphur and nitrogen particulate pollutant
38 deposition on to building surfaces. In: Saiz-Jimenez, C., ed. The deterioration of monuments: proceedings of
39 the 2nd international symposium on biodeterioration and biodegradation; January 1994; Sevilla, Spain. Sci.
40 Total Environ. 167: 57-66.
41 Han, Q.; Rossow, W. B.; Lacis, A. A. (1994) Near-global survey of effective droplet radii in liquid water clouds
42 using ISCCP data. J. Clim. 7: 465-497.
43 Haneef, S. J.; Johnson, J. B.; Thompson, G. E.; Wood, G. C. (1993) Simulation of the degradation of limestones
44 and dolomitic sandstone under dry deposition conditions. In: Corrosion control for low-cost reliability:
45 preceedings [of the] 12th international corrosion congress. Volume 2: process industries plant operation.
46 Houston, TX: National Association of Corrosion Engineers; pp. 700-710.
47 Hanley, N.; Spash, C. L. (1993) Cost-benefit analysis and the environment. Hants, United Kingdom: Edward Elgar.
48 Harwell, M. A.; Myers, V.; Young, T.; Bartuska, A.; Gassman, N.; Gentile, J. H.; Harwell, C. C.; Appelbaum, S.;
49 Barko, J.; Causey, B.; Johnson, C.; McLean, A.; Smola, R.; Templet, P.; Tosini, S. (1999) A framework for
50 an ecosystem integrity report card. BioScience 49: 543-556.
51 Hauhs, M. (1989) Lange Bramke: an ecosystem study of a forested catchment. In: Adriano, D. C.; Havas, M., eds.
52 Acidic precipitation: volume 1, case studies. New York, NY: Springer-Verlag; pp. 275-305. (Adriano, D. C.;
53 Salomons, W., eds. Advances in environmental science).
March 2001
4-171
DRAFT-DO NOT QUOTE OR CITE
-------
1 Haynie, F. H. (1976) Air pollution effects on stress induced intergranular corrosion of 7005-T53 aluminum alloy.
2 In: Stress corrosion-new approaches. Philadelphia, PA: American Society for Testing and Materials;
3 pp. 32-43. (ASTM special technical publication 610).
4 Haynie, F. H. (1980) Theoretical air pollution and climate effects on materials confirmed by zinc corrosion data.
5 In: Sereda, P. J.; Litvan, G. G., eds. Durability of building materials and components: proceedings of the
6 1 st international conference; August 1978; Ottawa, ON, Canada. Philadelphia, PA: American Society for
7 Testing and Materials; pp. 157-175. (ASTM special technical publication 691).
8 Haynie, F. H. (1986) Environmental factors affecting corrosion of weathering steel. In: Baboian, R., ed. Materials
9 degradation caused by acid rain: developed from the 20th state-of-the-art symposium of the American
10 Chemical Society; June 1985; Arlington, VA. Washington, DC: American Chemical Society; pp. 163-171.
11 (ACS symposium series 318).
12 Haynie, F. H.; Lemmons, T. J. (1990) Particulate matter soiling of exterior paints at a rural site. Aerosol Sci.
13 Technol. 13: 356-367.
14 Haynie, F. H.; Upham, J. B. (1974) Correlation between corrosion behavior of steel and atmospheric pollution data.
15 In: Cobum, S. K., ed. Corrosion in natural environments: presented at the 76th annual meeting American
16 Society for Testing and Materials; June 1973; Philadelphia, PA. Philadelphia, PA: American Society for
17 Testing and Materials; pp. 33-51. (ASTM special technical publication 558).
18 Haywood, J. M.; Shine, K. P. (1995) The effect of anthropogenic sulfate and soot aerosol on the clear sky planetary
19 radiation budget. Geophys. Res. Lett. 22: 603-606.
20 Haywood, J. M.; Ramaswamy, V.; Soden, B. J. (1999) Tropospheric aerosol climate forcing in clear-sky satellite
21 observations over the oceans. Science (Washington, DC) 283: 1299-1303.
22 Heal, G. (2000) Valuing ecosystem services. Ecosystems 3: 24-30.
23 Hedin, L. O.; Granat, L.; Likens, G. E.; Buishand, T. A.; Galloway, J. N.; Butler, T. J.; Rodhe, H. (1994)
24 Steep declines in atmospheric base cations in regions of Europe and North America. Nature (London)
25 367:351-354.
26 Heinsdorf, D. (1993) The role of nitrogen in declining Scots pine forests (Pinus sylvestris) in the lowland of
27 East Germany. Water Air Soil Pollut. 69: 21-35.
28 Herbert, D. A.; Rastetter, E. B.; Shaver, G. R.; Agren, G. I. (1999) Effects of plant growth characteristics on
29 biogeochemistry and community composition in a changing climate. Ecosystems 2: 367-382.
30 Hicks, B. B. (1986) Differences in wet and dry particle deposition parameters between North America and Europe.
31 In: Lee, S. D.; Schneider, T.; Grant, L. D.; Verkerk, P. J., eds. Aerosols: research, risk assessment and
32 control strategies; proceedings of the second U.S.-Dutch international symposium; May 1985; Williamsburg,
33 VA. Chelsea, MI: Lewis Publishers; pp. 973-982.
34 Hicks, B. B.; Baldocchi, D. D.; Meyers, T. P.; Hosker, R. P., Jr.; Matt, D. R. (1987) A preliminary multiple
35 resistance routine for deriving dry deposition velocities from measured quantities. Water Air Soil Pollut.
36 36:311-330.
37 Hinckley, T.; Ford, D.; Segura, G.; Sprugel, D. (1992) Key processes from tree to stand level. In: Wall, G., ed.
38 Implications of climate change for Pacific Northwest forest management. Waterloo, ON, Canada: University
39 of Waterloo, Department of Geography; pp. 33-43. (Department of Geography publication series: occasional
40 paper no. 15).
41 Hogan, G. D.; Rennenberg, H.; Fink, S. (1998) Role and effect of sulfur in tree biology. In: Maynard, D. G., ed.
42 Sulfur in the environment. New York, NY: Marcel Dekker, Inc.; pp. 173-217. (Books in soils, plants, and the
43 environment: v. 67).
44 Holbrow, G. L. (1962) Atmospheric pollution: its measurement and some effects on paint. J. Oil Colour Chem.
45 Assoc. 45:701-718.
46 Holling, C. S. (1986) The resilience of terrestrial ecosystems: local surprise and global change. In: Clark, W. C.;
47 Munn, R. E., eds. Sustainable development of the biosphere. Cambridge, UK: Cambridge University;
48 pp. 292-317.
49 Hooper, D. U.; Vitousek, P. M. (1997) The effects of plant composition and diversity on ecosystem processes.
50 Science 277:1302-1305.
51 Hornung, M.; Langan, S. J. (1999) Nitrogen deposition, sources, impacts and responses in natural and semi-natural
52 ecosystems. In: Bangan, S. J., ed. Impact of nitrogen deposition on natural ecosystems and semi-natural
53 ecosystems. Dordrect, Netherlands: Kluwer Academic Publishers; pp. 1-14. [Environmental Pollution, no.3].
54 Horvath, H. (1993) Atmospheric light absorption~a review. Atmos. Environ. Part A 27: 293-317.
March 2001
4-172
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Houghton, J. T.; Jenkins, G. J.; Ephraums, J. J., eds. (1990) Climate change: the IPCC scientific assessment.
Cambridge, MA: Cambridge University Press; p. 55.
Hoyt, D. V. (1978) Amodel for the calculation of solar global insolation. Sol. Energy 21: 27-35.
Hughes, M. K. (1981) Cycling of trace metals in ecosystems. In: Lepp, N. W., ed. Effect of heavy metal pollution
on plants. Volume 2: metals in the environment. London, United Kingdom: Applied Science Publishers-
pp. 95-118.
Hughes, J. B.; Daily, G. C.; Ehrlich, P. R. (1997) The loss of population diversity and why it matters. In: Raven,
P. H., ed. Nature and Human Society: the quest for a sustainable world. Washington, DC: National Academy
Press; pp. 71-83.
Humphreys, F. R.; Lambert, M. J.; Kelly, J. (1975) The occurrence of sulphur deficiency in forests. In: McLachlan,
K. D., ed. Sulphur in Australasian agriculture. Sydney, Australia: Sydney University Press; pp. 154-162.
Huntington, T.G. (2000) The potential for calcium depletion in forest ecosystems of southeastern United States:
review and analysis. Global Biogeochem. Cycles 14: 623-638.
Husar, R. B.; Wilson, W. E. (1993) Haze and sulfur emission trends in the eastern United States Environ Sci
Technol. 27: 12-16.
Husar, R. B.; Holloway, J. M.; Patterson, D. E.; Wilson, W. E. (1981) Spatial and temporal pattern of eastern U.S.
haziness: a summary. Atmos. Environ. 15: 1919-1928.
Husar, R. B.; Elkins, J. B.; Wilson, W. E. (1994) U.S. visibility trends, 1960-1992, regional and national. Presented
at: 87th annual meeting of the Air & Waste Management Association; June; Cincinnati, OH. Pittsburgh, PA:
Air & Waste Management Association; paper no. 94-MP3.05.
Husar, R. B.; Schichtel, B. A.; Falke, S, R.; Wilson, W. E. (2000) Haze trends over the United States, 1980-1995.
J. Geophys. Res. Lett.: In Press.
Hutchinson, A. J.; Johnson, J. B.; Thompson, G. E.; Wood, G. C.; Sage, P. W.; Cooke, M. J. (1992) The role of
fly-ash particulate material and oxide catalysts in stone degradation. Atmos. Environ. 26A: 2795-2803.
Huttunen, S. (1994) Effects of air pollutants on epicuticular wax structure. In: Percy, K. E.; Cape, J. N.; Jagels, R.;
Simpson, C. J., eds. Air pollutants and the leaf cuticle. New York, NY: Springer-Verlag; pp 81-96
(NATO ASI series).
Intergovernmental Panel on Climate Change (IPCC). (1995) Climate change 1994: radiative forcing of climate
change and an evaluation of the IPCC IS92 emission scenarios. Cambridge, United Kingdom: University
Press.
Intergovernmental Panel on Climate Change (IPCC). (1996) Climate change 1996: contribution of working group
I to the second assessment of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom:
Cambridge University Press; p. 572. [IPCC second assessment report].
Intergovernmental Panel on Climate Change (IPCC). (1998) The regional impacts of climate change: an assessment
of vulnerability. Cambridge, United Kingdom: Cambridge University Press.
Irving, P. M., ed. (1991) Acidic deposition: state of science and technology: volumes I-IV. Washington, DC:
The U.S. National Acid Precipitation Assessment Program.
Iyer, H.; Patterson, P.; Malm, W. C. (2000) Trends in the extremes of sulfur concentration distributions. J. Air
Waste Manage. Assoc. 50: 802-808.
Jacobson, M. Z. (1998) Studying the effects of aerosols on vertical photolysis rate coefficient and temperature
profiles over an urban airshed. J. Geophys. Res. [Atmos.] 103: 10,593-10,604.
Jacobson, M. Z. (1999) Isolating nitrated and aromatic aerosols and nitrated aromatic gases as sources of ultraviolet
light absorption. J. Geophys. Res. 104 (D3): 3527-3542.
Janzen, J. (1980) Extinction of light by highly nonspherical strongly absorbing colloidal particles:
spectrophotometric determination of volume distributions for carbon blacks. Appl. Opt. 19: 2977-2985.
Japar, S. M.; Szkarlat, A. C.; Pierson, W. R. (1984) The determination of the optical properties of airborne particle
emissions from diesel vehicles. Sci. Total Environ. 36: 121-130.
Japar, S. M.; Brachaczek, W. W.; Gorse, R. A., Jr.; Norbeck, J. M.; Pierson, W. R. (1986) The contribution of
elemental carbon to the optical properties of rural atmospheric aerosols. Atmos. Environ. 20: 1281-1289.
Jensen, V. (1974) Decomposition of angiosperm tree leaf litter. In: Dickinson, C. H.; Pugh, G. J. F., eds. Biology of
plant litter decomposition: v. I. London, United Kingdom: Academic Press; pp. 69-104.
Johnson, D. W. (1984) Sulfur cycling in forests. Biogeochemistry 1: 29-43.
March 2001
4-173
DRAFT-DO NOT QUOTE OR CITE
-------
1 Johnson, D. W. (1992) Base cation distribution and cycling. In: Johnson, D. W.; Lindberg, S. E., eds. Atmospheric
2 deposition and forest nutrient cycling: a synthesis of the integrated forest study. New York, NY:
3 Springer-Verlag, Inc.; pp. 275-333. (Billings, W. D.; Golley, F.; Lange, O. L.; Olson, J. S.; Remmert, H., eds.
4 Ecological studies, analysis and synthesis: v. 91).
5 Johnson, F. R.; Desvousges, W. H. (1997) Estimating stated preferences with rated-pair data: environmental, health,
6 and employment effects of energy programs. J. Environ. Econ. Manage. 34: 79-99.
7 Johnson, D. W.; Lindberg, S. E., eds. (1992a) Atmospheric deposition and forest nutrient cycling: a synthesis of the
8 integrated forest study. New York, NY: Springer-Verlag, Inc. (Billings, W. D.; Golley, F.; Lange, O. L.;
9 Olson, J. S.; Remmert, H., eds. Ecological studies: analysis and synthesis: v. 91).
10 Johnson, D. W.; Lindberg, S. E., eds. (1992b) Nitrogen chemistry, deposition, and cycling in forests. In: Johnson,
H D. W.; Lindberg, S. E., eds. Atmospheric deposition and forest nutrient cycling: a synthesis of the integrated
12 forest study. New York, NY: Springer-Verlag, Inc.; pp. 150-213. (Billings, W. D.; Golley, F.; Lange, O. L.;
13 Olson, J. S.; Remmert, H., eds. Ecological studies: analysis and synthesis: v. 91).
14 Johnson, D. W.; Mitchell, M. J. (1988) Responses of forest ecosystems to changing sulfur inputs. In: Maynard,
15 D. G., ed. Sulfur in the environment. New York, NY: Marcel Dekker, Inc.; pp. 219-262. [Books in Soils,
16 Plants, and the Environment, v. 67].
17 Johnson, D. W.; Taylor, G. E. (1989) Role of air pollution in forest decline in eastern North America. Water Air
18 SoilPollut.48:21-43.
19 Johnson, D. W.; Todd, D. E., Jr. (1998) Harvesting effects on long-term changes in nutrient pools of mixed oak
20 forest. Soil Sci. Soc. Amer. J. 62: 1725-1735.
21 Johnson, C. D.; Latimer, D. A.; Bergstrom, R. W.; Hogo, H. (1980) User's manual for the plume visibility model
22 (PLUVUE). Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality
23 Planning and Standards; report no. EPA-450/4-80-032. Available from: NTIS, Springfield, VA;
24 PB81-163297.
25 Johnson, D. W.; Henderson, G. S.; Huff, D. D.; Lindberg, S. E.; Richter, D. D.; Shriner, D. S.; Todd, D. E.;
26 Turner, J. (1982a) Cycling of organic and inorganic sulphur in a chestnut oak forest. Oecologia 54: 141-148.
27 Johnson, D. W.; Cole, D. W.; Bledsoe, C. S.; Cromack, K.; Edmonds, R. L.; Gessel, S. P.; Grier, C. C.; Richards,
28 B. N.; Vogt, K. A. (1982b) Nutrient cycling in forests of the Pacific Northwest. In: Edmonds, R. L., ed.
29 Analysis of coniferous forest ecosystems in the Western United States. Stroudsburg, PA: Hutchinson Ross
30 Publishing Company; pp. 186-232. (US/IBP synthesis series: v. 14).
31 Johnson, D. W.; Van Miegroet, H.; Lindberg, S. E.; Todd, D. E.; Harrison, R. B. (1991a) Nutrient cycling in red
32 spruce forests of the Great Smoky Mountains. Can. J. For. Res. 21: 769-787.
33 Johnson, D. W.; Cresser, M. S.; Nilsson, S. I.; Turner, J.; Ulrich, B.; Binkley, D.; Cole, D. W. (199 Ib) Soil changes
34 in forest ecosystems: evidence for and probable causes. Proc. R. Soc. Edinburgh Sect. B: Biol. Sci.
35 97B:81-116.
36 Johnson, D. W.; Swank, W. T.; Vose, J. M. (1993) Simulated effects of atmospheric sulfur deposition on nutrient
37 cycling in a mixed deciduous forest. Biogeochemistry 23:169-196.
38 Johnson, A. H.; Friedland, A. J.; Miller, E. K.; Siccama, T. G. (1994a) Acid rain and soils of the Adirondacks. III.
39 Rates of soil acidification in a montane spruce-fir forest at Whiteface Mountain, New York. Can. J. For. Res.
40 24:663-669.
41 Johnson, A. H.; Andersen, S. B.; Siccama, T. G. (1994b) Acid rain and soils of the Adirondacks. I. Changes in pH
42 and available calcium, 1930-1984. Can. J. For. Res. 24: 39-45.
43 Johnson, D. W.; Susfalk, R. B.; Brewer, P. F. (1996) Simulated responses of red spruce forest soils to reduced
44 sulfur and nitrogen deposition. J. Environ. Qual. 25: 1300-1309.
45 Johnson, D.W.; Susfalk, R.B.; Brewer, P.,F.; Swank, W.T. (1999) Simulated effects of reduced sulfur, nitrogen, and
46 base cation deposition on soils and solutions in southern Appalachian forests. J. Environ. Qual
47 28:1336-1346.
48 Joslin, J. D.; Wolfe, M. H. (1992) Red spruce soil chemistry and root distribution across a cloud water deposition
49 'gradient. Can. J. For. Res. 22: 893-904.
50 Joslin, J. D.; Wolfe, M. H. (1994) Foliar deficiencies of mature Southern Appalachian red spruce determined from
51 fertilizer trials. Soil Sci. Soc. Am. J. 58:1572-1579.
52 Joslin, J. D.; Kelly, J. M.; Van Miegroet, H. (1992) Soil chemistry and nutrition of North American spruce-fir
53 stands: evidence for recent change. J. Environ. Qual. 21: 12-30.
54 Justus, C. G.; Murphey, B. B. (1994) Temporal trends in surface irradiance at ultraviolet wavelengths. J. Geophys.
55 Res. 99 (Dl): 1389-1394.
March 2001
4-174
DRAFT-DO NOT QUOTE OR CITE
-------
1 Kalkstein, L. S.; Greene, J. S. (1997) An evaluation of climate/mortality relationships in large U.S. cities and the
2 possible impacts of a climate change. Environ. Health Perspect. 105: 84-93.
3 Karmoker, J. L.; Clarkson, D. T.; Saker, L. R.; Rooney, J. M.; Purves, J. V. (1991) Sulphate deprivation depresses
4 the transport of nitrogen to the xylem and the hydraulic conductivity of barley (Hordeum vulgare L.) roots.
5 Planta 185: 269-278.
6 Katsev, I. L.; Zege, E. P. (1994) The modern theory of black object visibility and meteorological visibility range.
7 Atmos. Environ. 28: 763-767.
8 Kelly, J.; Lambert, M. J. (1972) The relationship between sulphur and nitrogen in the foliage ofPinus radiata.
9 Plant Soil 37: 395-407.
10 Kenk, G.; Fischer, H. (1988) Evidence from nitrogen fertilisation in the forests of Germany. Environ. Pollut.
11 54:199-218.
12 Kerker, M. (1969) The scattering of light and other electromagnetic radiation. New York, NY: Academic Press, Inc.
13 (Loebl, E. M., ed. Physical chemistry: a series of monographs, no. 16).
14 Kiehl, J. T.; Briegleb, B. P. (1993) The relative roles of sulfate aerosols and greenhouse gases in climate forcing.
15 Science (Washington, DC) 260: 311-314.
16 Kodama, Y.; Lee, S. D., eds. (1994) The 13th UOEH international symposium and the 2nd pan Pacific cooperative
17 symposium on impact of increased UV-B exposure on human health and ecosystem; October 1993;
18 Kitakyushu, Japan. Kitakyushu, Japan: University of Occupational and Environmental Health.
19 Koschmieder, H. (1924) Theorie der horizontalen Sichtweite II: Kontrast und Sichtweite [Theory of horizontal
20 visibility]. Beitr. Phys. Atmos. 12: 171-181.
21 Koslow, E. E.; Smith, W. H.; Staskawicz, B. J. (1977) Lead-containing particles on urban leaf surfaces.
22 Environ. Sci. Technol. 11: 1019-1021.
23 Krajickova, A.; Mejstfik, V. (1984) The effect of fly-ash particles on the plugging of stomata. Environ. Pollut.
24 36:83-93.
25 Kripke, M. L. (1989) Health effects of stratospheric ozone depletion: an overview. In: Schneider, T.; Lee, S. D.;
26 Wolters, G. J. R.; Grant, L. D. eds. Atmospheric ozone research and its policy implications: proceedings of
27 the Third US-Dutch International Symposium; May, 1988; Nijmegen, The Netherlands. Amsterdam, The
28 Netherlands: Elsevier, pp. 795-802. [Studies in Environmental Science 35].
29 Krupa, S. V.; Legge, A. H. (1986) Single or joint effects of coarse and fine particulate sulfur aerosols and ozone on
30 vegetation. In: Lee, S. D.; Schneider, T.; Grant, L. D.; Verkerk, P. J., eds. Aerosols: research, risk assessment
31 and control strategies, proceedings of the second U.S.-Dutch international symposium; May 1985;
32 Williamsburg, VA. Chelsea, MI: Lewis Publishers, Inc.; pp. 879-887.
33 Krupa, S. V.; Legge, A. H. (1998) Sulphur dioxide, particulate sulphur and its impacts on a boreal forest ecosystem.
34 In: Ambasht, R. S., ed. Modern trends in ecology and environment. Leiden, The Netherlands: Backhuys
35 Publishers; pp. 285-306.
36 Kylin, H.; Grimvall, E.; Ostman, C. (1994) Environmental monitoring of polychlorinated biphenyls using pine
37 needles as passive samplers. Environ. Sci. Technol. 28: 1320-1324.
38 Lamersdorf, N. P.; Meyer, M. (1993) Nutrient cycling and acidification of a northwest German forest site with high
39 atmospheric nitrogen deposition. For. Ecol. Manage. 62: 323-354.
40 Latimer, D. A. (1988) Plume perceptibility thresholds. Presented at: 81st annual meeting of the Air Pollution
41 Control Association; June; Dallas, TX. Pittsburgh, PA: Air Pollution Control Association; paper no. 88-56.6.
42 Latimer, D. A.; Bergstrom, R. W.; Hayes, S. R.; Liu, M.-K.; Seinfeld, J. H.; Whitten, G. Z.; Wojcik, M. A.; Hillyer,
43 M. J. (1978) The development of mathematical models for the prediction of anthropogenic visibility
44 impairment: volumes I-III. Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of
45 Air Quality Planning and Standards; EPA report no. EPA-450/3-78-11 Oa-c. Available from: NTIS,
46 Springfield, VA; PB-293118-SET.
47 Laurenroth, W. K.; Milchunas, D. G. (1985) SO2 effects on plant community function. In: Winner, W. E.; Mooney,
48 H. A.; Goldstein, R. A., eds. Sulfur dioxide and vegetation: physiology, ecology, and policy issues. Stanford,
49 CA: Stanford University Press; pp. 454-477.
50 Lawrence, G. B.; Huntington, T. G. (1999) Soil-Calcium depletion linked to acid rain and forest growth in the
51 eastern United States. Reston, VA: U.S. Geological Survey. Available: http://btdqs.usgs.gov/acidrain [1999,
52 November 23].
53 Lawrence, G. B.; David, M. B.; Shortle, W. C. (1995) A new mechanism for calcium loss in forest-floor soils.
54 Nature 378: 162-164.
March 2001
4-175
DRAFT-DO NOT QUOTE OR CITE
-------
1 Legge, A. H. (1980) Primary productivity, sulfur dioxide, and the forest ecosystem: an overview of a case study.
2 In: Miller, P. R., ed. Proceedings of symposium on effects of air pollutants on Mediterranean and temperate
3 forest ecosystems; June; Riverside, CA. Berkeley, CA: U.S. Department of Agriculture, Pacific Southwest
4 Forest and Range Experiment Station; pp. 51-62; Forest Service general technical report no. PSW-43.
5 Available from: NTIS, Springfield, VA; PB81-133720.
6 Lcrman, S. L.; Darley, E. F. (1975) Particulates. In: Mudd, J. B.; Kozlowski, T. T., eds. Responses of plants to air
7 pollution. New York, NY: Academic Press; pp. 141-158.
8 Levin, S. A. (1998) Ecosystems and the biosphere as complex adaptive systems. Ecosystems 1: 431-436.
9 Li-Jones, X.; Prospero, J. M. (1998) Variations in the size distribution of non-sea-salt sulfate aerosol in the marine
10 boundary layer at Barbados: impact of African dust. J. Geophys. Res. 103: 16,073-16,084.
11 Likens, G. E.; Driscoll, C. T.; Buso, D. C. (1996) Long-term effects of acid rain: response and recovery of a forest
12 ecosystem. Science (Washington, DC) 272: 244-246.
13 Lindberg, S. E.; Harriss, R. C. (1981) The role of atmospheric deposition in an eastern U.S. deciduous forest.
14 Water Air Soil Pollut. 16:13-31.
15 Lindberg, S. E.; Lovett, G. M.; Meiwes, K.-J. (1987) Deposition and forest canopy interactions of airborne nitrate.
16 In: Hutchinson, T. C.; Meema, K. M., eds. Effects of atmospheric pollutants on forests, wetlands and
17 agricultural ecosystems. Berlin, Federal Republic of Germany: Springer-Verlag; pp. 117-130. (NATO as
18 1 series: v. Gl 6).
19 Liu, S. C.; McKeen, S. A.; Madronich, S. (1991) Effect of anthropogenic aerosols on biologically active ultraviolet
20 radiation. Geophys. Res. Lett. 18:2265-2268.
21 Lokke, H.; Bak, J.; Falkengren-Grerup, U.; Finlay, R. D.; Ilvesniemi, H.; Nygaard, P. H.; Starr, M. (1996)
22 Critical loads of acidic deposition for forest soils: is the current approach adequate. Ambio 25: 510-516.
23 Longstreth, J. (1999) Public health consequences of global climate change in the United States—some regions may
24 suffer disproportionately. Environ. Health Perspect. 107(suppl. 1): 169-179.
25 Lorente, J.; Redano, A.; de Cabo, X. (1994) Influence of urban aerosol on spectral solar irradiance. J. Appl.
26 Meteorol. 33:406-415.
27 Lorusso, S.; Marabelli, M.; Troili, M. (1997) Air pollution and the deterioration of historic monuments. J. Environ.
28 Pathol. Toxicol. Oncol. 16: 171-173.
29 Lovett, G. M. (1984) Atmospheric deposition to forests. In: Breece, L.; Hasbrouck, S., eds. Forest-responses to
30 acidic deposition: proceedings of the U. S.-Canadian conference; August 1983; Orono, ME. Orono, ME:
31 University of Maine, Orono, Land and Water Resources Center; pp. 7-18.
32 Lovett, G. M.; Lindberg, S. E. (1984) Dry deposition and canopy exchange in a mixed oak forest as determined by
33 analysis of throughfall. J. Appl. Ecol. 21: 1013-1027.
34 Lovett, G. M.; Lindberg, S. E. (1986) Dry deposition of nitrate to a deciduous forest. Biogeochemistry 2: 137-148.
35 Lovett, G. M.; Lindberg, S. E. (1993) Atmospheric deposition and canopy interactions of nitrogen in forests. Can. J.
36 For. Res. 23: 1603-1616.
37 Lowenthal, D. H.; Rogers, C. F.; Saxena, P.; Watson, J. G.; Chow, J. C. (1995) Sensitivity of estimated light
38 extinction coefficients to model assumptions and measurement errors. Atmos. Environ. 29: 751-766.
39 Madronich, S. (1987) Photodissociation in the Atmosphere. 1. Actinic flux and the effects of ground reflections and
40 clouds. J. Geophys. Res. [Atmos.] 92: 9740-9752.
41 Mahadev, S.; Henry, R. C. (1999) Application of a color-appearance model to vision through atmospheric haze.
42 Color Res. Appl. 24:112-120.
43 Malm, W. C.; Pitchford, M. L. (1997) Comparison of calculated sulfate scattering efficiencies as estimated from
44 size-resolved particle measurements at three national locations. Atmos. Environ. 31: 1315-1325.
45 Malm, W. C.; Sisler, J. F.; Huffman, D.; Eldred, R. A.; Cahill, T. A. (1994) Spatial and seasonal trends in particle
46 concentration and optical extinction in the United States. J. Geophys. Res. [Atmos.] 99: 1347-1370.
47 Malm, W. C.; Molenar, J. V.; Eldred, R. A.; Sisler, J. F. (1996) Examining the relationship among atmospheric
48 aerosols and light scattering and extinction in the Grand Canyon area. J. Geophys. Res. [Atmos.]
49 101:19,251-19,265.
50 Malm, W. C.; Day, D.; Kreidenweis, S. M. (1997) Comparison of measured and reconstructed scattering during an
51 intensive field study at Great Smoky Mountains National Park. Presented at: 90th annual meeting and
52 exhibition of the Air & Waste Management Association; June; Toronto, ON, CA. Pittsburgh, PA: Air &
53 Waste Management Association; 97-WA70.02.
March 2001
4-176
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
Mansfeld, F. (1980) Regional air pollution study: effects of airborne sulfur pollutants on materials. Research
Triangle Park, NC: U.S. Environmental Protection Agency, Environmental Sciences Research Laboratory;
EPA report no. EPA-600/4-80-007. Available from: NTIS, Springfield, VA; PB81-126351.
Marchwinska, E.; Kucharski, R. (1987) The combined influence of SO2 and heavy metal-containing particulates on
beans, carrots and parsley. Environ. Monit. Assess. 8: 11-25.
Markewitz, D.; Richter, D.D.; Allen, L. H.; Urrego, J. B. (1998) Three decades of observed soil acidification in the
Calhoun Experimental Forest: Has acid rain made a difference? Soil Sci. Soc. Amer. J. 62: 1428-1439.
Marschner, H. (1995) Sulfur. In: Mineral nutrition of higher plants. 2nd ed. London, United Kingdom: Academic
Press; pp. 255-265.
Martin, M. H.; Coughtrey, P. J. (1981) Impact of metals on ecosystem function and productivity. In: Lepp, N. W.,
ed. Effect of heavy metal pollution on plants: volume 2, metals in the environment. Barking; United
Kingdom: Applied Science Publishers; pp. 119-158. (Mellanby, K., ed. Pollution monitoring series).
Martin, C. E.; Gravatt, D. A.; Loeschen, V. S. (1992) Photosynthetic responses of three species to acute exposures
of nitrate- and sulphate-containing aerosols. Atmos. Environ. Part A 26: 381-391.
Materna, J. (1984) Impact of atmospheric pollution on natural ecosystems. In: Treshow, M., ed. Air pollution and
plant life. New York, NY: John Wiley & Sons; pp. 397-416.
Mathai, C. V. (1995) The Grand Canyon visibility. EM (December): 20-31.
Matson, P. A.; Parton, W. J.; Power, A. G.; Swift, M. J. (1997) Agricultural intensification and ecosystem
properties. Science (Washington, DC) 277: 504-509.
May, P. F.; Till, A. R.; Gumming, M. J. (1972) Systems analysis of sulphur kinetics in pastures grazed by sheep.
J. Appl. Ecol. 9: 25-49.
McGee, E. S.; Mossotti, V. G. (1992) Gypsum accumulation on carbonate stone. Atmos. Environ. Part B
26: 249-253.
McGowan, T. F.; Lipinski, G. E.; Santoleri, J. J. (1993) New rules affect the handling of waste fuels. Chem. Eng
(March): 122-128.
McGrattan, K. B.; Walton, W. D.; Putori, A. D., Jr.; Twilley, W. H.; McElroy, J. A.; Evans, D. D. (1995) Smoke
plume trajectory from in situ buring of crude oil in Alaska: field experiments. Gaithersburg, MD: U.S.
National Institute of Standards and Technology; report no. NISTIR 5764. Available from: NTIS, Springfield
VA;PB96-131560.
McGrattan, K. B.; Trelles, J.; Baum, H. R.; Rehm, R. G. (1996) Smoke plume trajectory over two-dimensional
terrain. In: Beall, K., ed. Annual conference on fire research: book of abstracts; October; Gaithersburg, MD.
Gaithersburg, MD: United States Department of Commerce, National Institute of Standards and Technology;
report no. NISTIR 5904; pp. 65-66. Available from: NTIS, Springfield, VA; PB97-153514.
McLachlan, M. S. (1996a) Bioaccumulation of hydrophobic chemicals in agricultural food chains. Environ Sci •
Tech. 30: 252-259.
McLachlan, M. S. (1996b) Biological uptake and transfer of polychlorinated dibenzo-p-dioxins and dibenzofurans.
In: Hester, R. E.; Harrison, R. M., eds. Chlorinated organic micropollutants. Cambridge, United Kingdom:
Royal Society of Chemistry; pp. 31 -52. (Issues in environmental science and technology).
McLachlan, M. S. (1999) Framework for the interpretation of measurements of SOCs in plants. Environ. Sci
Technol. 33:1799-1804.
McMurry, P. H.; Stolzenburg, M. R. (1989) On the sensitivity of particle size to relative humidity for Los Angeles
aerosols. Atmos. Environ. 23: 497-507.
McMurry, P. H.; Zhang, X.; Lee, C.-T. (1996) Issues in aerosol measurement for optics assessments. J. Geophys
Res. [Atmos.] 101: 19,189-19,197.
Meiwes, K. J.; Khanna, P. K. (1981) Distribution and cycling of sulphur in the vegetation of two forest ecosystems
in an acid rain environment. Plant Soil 60: 369-375.
Meng, R. Z.; Karamchandani, P.; Seigneur, C. (2000) Simulation of stack plume opacity. J. Air Waste Manage
Assoc. 50: 869-874.
Mercer, G. S. (1994) Transmissometer data reporting (IMPROVE protocol). Fort Collins, CO: Air Resource
Specialists, Inc.; technical instruction no. 4500-5100.
Mid-Atlantic Regional Assessment (MARA) Team. (2000) Preparing for a changing climate: the potential
conswquences of climate variability and change (Mid-Atlantic overview). Washington, DC: U.S. Global
Change Research Program (USGCRP).
March 2001
4-177
DRAFT-DO NOT QUOTE OR CITE
-------
1 Middleton, P. (1997) DAQM-simulated spatial and temporal differences among visibility, PM, and other air quality
2 concerns under realistic emission change scenarios. In: Special issue on aerosols and atmospheric optics-part
3 II, [published papers from the international specialty conference; September 1994; Snowbird, UT]. J. Air
4 Waste Manage. Assoc. 47: 302-316.
5 Mie, G. (1908) Beitrage zur Optik truber Medien, speziell kolloidaler Metallosungen [Optics of cloudy media,
6 especially colloidal metal solutions]. Ann. Phys. (Leipzig) 25: 377-445.
7 Millar, C. S. (1974) Decomposition of coniferous leaf litter. In: Dickinson, C. H.; Pugh, G. J. F., eds. Biology of
8 plant litter decomposition: volume I. New York, NY: Academic Press; pp. 105-128.
9 Miller, P. R.; McBride, J. R., eds. (1999) Oxidant air pollution impacts in the Montane forests of southern
10 California: a case study of the San Bernardino Mountains. New York, NY: Springer-Verlag.
11 (Ecological Studies: v. 134.)
12 Miller, W. C.; Fornes, R. E.; Gilbert, R. D.; Speer, A.; Spence, J. (1992) Removal of CaCO3 extender in residential
13 coatings by atmospheric acidic deposition. In: Measurement of toxic and related air pollutants: proceedings
14 of the 1992 U.S. EPA/A&WMA international symposium. Pittsburgh, PA: Air & Waste Management
15 Association; pp. 129-134. (A&WMA publication VIP-25).
16 Miller, E. K.; Panek, J. A.; Friedland, A. J.; Kadlecek, J.; Mohnen, V. A. (1993) Atmospheric deposition to a
17 high-elevation forest at Whiteface Mountain, New York, USA. Tellus Ser. B 45:209-227.
18 Mims, F. M., III. (1996) Significant reduction of UVB caused by smoke from biomass burning in Brazil.
19 Photochem. Photobiol. 64: 814-816.
20 Mitchell, M. J.; Harrison, R. B.; Fitzgerald, J. W.; Johnson, D. W.; Lindberg, S. E.; Zhang, Y.; Autry, A. (1992a)
21 Sulfur distribution and cycling in forest ecosystems. In: Johnson, D. W.; Lindberg, S. E., eds. Atmospheric
22 deposition and forest nutrient cycling: a synthesis of the integrated forest study. New York, NY:
23 Springer-Verlag; pp. 90-129. (Billings, W. D.; Golley, F.; Lange, O. L.; Olson, J. S.; Remmert, H., eds.
24 Ecological studies: analysis and synthesis: v. 91).
25. Mitchell, M. J.; Foster, N. W.; Shepard, J. P.; Morrison, I. K. (1992b) Nutrient cycling in Huntington Forest and
26 Turkey Lakes deciduous stands: nitrogen and sulfur. Can. J. For. Res. 22:457-464.
27 Molenar, J. V.; Persha, G.; Malm, W. C. (1990) Long path transmissometer for measuring ambient atmospheric
28 extinction. In: Mathai, C. V., ed. Visibility and fine particles: an A&WMA/EPA international specialty
29 conference; October 1989; Estes Park, CO. Pittsburgh PA: Air & Waste Management Association;
30 pp. 293-304. (AWMA transactions series no. TR-17).
31 Molenar, J. V.; Cismoski, D. S.; Tree, R. M. (1992) Intercomparison of ambient optical monitoring techniques.
32 Presented at: 85th annual meeting & exhibition of the Air & Waste Management Association; June;
33 Kansas City, MO. Pittsburgh, PA: Air & Waste Management Association; paper no. 92-60.09.
34 Molenar, J. V.; Malm, W. C.; Johnson, C. E. (1994) Visual air quality simulation techniques. Atmos. Environ.
35 28:1055-1063.
36 Moran, M. D.; Pielke, R. A. (1994) Delayed shear enhancement in mesoscale atmospheric dispersion. Presented at:
37 8th joint conference on applications of air pollution meteorology with A&WMA; January; Nashville, TN.
38 Boston, MA: Meteorological Society; pp. 96-103.
39 Moropoulou, A.; Bisbikou, K.; Torfs, K.; Van Grieken, R.; Zezza, F.; Macri, F. (1998) Origin and growth of
40 weathering crusts on ancient marbles in industrial atmosphere. Atmos. Environ. 32: 967-982.
41 Naeem, S.; Thompson, L. J.; Lawler, S. P.; Lawton, J. H.; Woodfin, R. M. (1994) Declining biodiversity can alter
42 the performance of ecosystems. Nature 368: 734-737.
43 Nash, T. H., III. (1975) Influence of effluents from a zinc factory on lichens. Ecol. Monogr. 45: 183-198.
44 National Acid Precipitation Assessment Program. (1991) Acidic deposition: state of science and technology.
45 Summary report of the National Acid Precipitation Assessment Program. Washington, DC: U.S. National
46 Acid Precipitation Assessment Program.
47 National Oceanic and Atmospheric Administration. (1992) ASOS (Automated Surface Observing System) user's
48 guide. Washington, DC: National Oceanic and Atmospheric Administration.
49 National Park Service. (1998) IMPROVE program goals and objectives. Washington, DC: National Park Service.
50 Available: www2.nature.nps.gov/ard/impr/index.htm [1999, November 24].
51 National Research Council. (1993) Protecting visibility in national parks and wilderness areas. Washington, DC:
52 National Academy Press. 3v.
53
March 2001
4-178
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
National Science and Technology Council. (1998) National acid precipitation assessment program biennial report to
Congress: an integrated assessment; executive summary. Silver Spring, MD: U.S. Department of Commerce,
National Oceanic and Atmospheric Administration. Available:
www.nnic.noaa.gov/CENR/NAPAP/NAPAP_96.htm [1999, November 24].
National Weather Service. (1998) Visibility. Washington, DC: National Weather Service, National Oceanic and
Atmospheric Administration. Available: www.nws.noaa.gov/asos/vsby.htm [1999, November 24].
Nazaroff, W. W.; Cass, G. R. (1991) Protecting museum collections from soiling due to the deposition of airborne
particles. Atmos. Environ. Part A 25: 841-852.
Neff, W. D. (1997) The Denver brown cloud studies from the perspective of model assessment needs and the role of
meteorology. J. Air Waste Manage. Assoc.
Nielsen, K. E.; Ladekarl, U. L.; Nornberg, P. (1999) Dynamic soil processes on heathland due to changes in
vegetation to oak and Sitka spruce. For. Ecol. Manage. 114: 107-116.
Nihlgard, B. (1985) The ammonium hypothesis-an additional explanation to the forest dieback in Europe. Ambio
14: 2-8.
Niklihska, M.; Laskowski, R.; Maryaiislci, M. (1998) Effect of heavy metals and storage time on two types of forest
litter: basal respiration rate and exchangeable metals. Ecotoxicol. Environ. Safety 41: 8-18.
Noble, I. R.; Dirzo, R. (1997) Forests as human-dominated ecosystems. Science (Washington, DC) 277: 522-525.
Nord, A. G.; Ericsson, T. (1993) Chemical analysis of thin black layers on building stone. Stud. Conserv. 38: 25-35.
Norton, R. A.; Childers, N. F. (1954) Experiments with urea sprays on the peach. Proc. Am. Soc. Hortic. Sci.
63:23-31.
Novakov, T.; Penner, J. E. (1993) Large contribution of organic aerosols to cloud-condensation-nuclei
concentrations. Nature (London) 365: 823-826.
Ockenden, W. A.; Stinnes, E.; Parker, C.; Jones, K. C. (1998) Observations on persistent organic pollutants in
plants: implications for their use as passive air samplers and for POP cycling. Environ. Sci. Technol.
32: 2721-2726.
Odum, E. P. (1969) The strategy of ecosystem development: an understanding of ecological succession provides a
basis for resolving man's conflict with nature. Science (Washington, DC) 164: 262-270.
Odum, E. P. (1985) Trends expected in stressed ecosystems. BioScience 35: 419-422.
Odum, E. P. (1993) The ecosystem. In: Ecology and our endangered life-support systems. 2nd ed. Sunderland, MA:
Sinauer Associates, Inc.; pp. 38-67.
Oehme, M.; Biseth, A.; Schlabach, M.; Wiig, O. (1995) Concentrations of polychlorinated dibenzo-p-dioxins,
dibenzofurans and non-ortho substituted biphenyls in polar bear milk from Svalbard (Norway). Environ.
Pollut. 90: 401-407.
Olszyk, D. M.; Bytnerowicz, A.; Takemoto, B. K. (1989) Photochemical oxidant pollution and vegetation: effects of
gases, fog, and particles. Environ. Pollut. 61:11-29.
Omar, A. H.; Biegalski, S.; Larson, S. M.; Landsberger, S. (1999) Particulate contributions to light extinction and
local forcing at a rural Illinois site. Atmos. Environ. 33: 2637-2646.
Ormrod, D. P. (1984) Impact of trace element pollution on plants. In: Treshow, M., ed. Air pollution and plant life.
Chichester, United Kingdom: John Wiley & Sons; pp. 291-319.
Paerl, H. W.; Bales, J. D.; Ausley, L. W.; Buzzelli, C. P.; Crowder, L. B.; Eby, L. A.; Go, M.; Peierls, B. L.;
Richardson, T. L.; Ramus, J. S. (2000) Recent hurricanes result in continuing ecosystem impacts on USA's
largest lagoonal estuary: Pamlico Sound, North Carolina. EOS Trans. Am. Geophys. U.: In Press.
Pan, W.; Tatang, M. A.; McRae, G. J.; Prinn, R. G. (1997) Uncertainty analysis of direct radiative forcing by
anthropogenic sulfate aerosols. J. Geophys. Res. 102 (D18): 21,915-21,924.
Pan, W.; Tatang, M. A.; McRae, G. J:; Prinn, R. G. (1998) Uncertainty analysis of indirect radiative forcing by
anthropogenic sulfate aerosols. J. Geophys. Res. 103 (D4): 3815-3823.
Parish, S. B. (1910) The effect of cement dust on citrus trees. Plant World 13: 288-291.
Parker, A. (1955) The destructive effects of air pollution on materials. Colliery Guardian 191:447-450.
Parker, G. G. (1983) Throughfall and stemflow in the forest nutrient cycle. In: Macfadyen, A.; Ford, E. D., eds.
Advances in ecological research, v. 13; pp. 57-120.
Patz, J. A.; Engelberg, D.; Last, J. (2000a) The effects of changing weather on public health. Ann. Rev. Public
Health 21: 271-307.
March 2001
4-179
DRAFT-DO NOT QUOTE OR CITE
-------
1 Patz, J. A.; McGeehin, M. A.; Bernard, S. M.; Ebi, K. L.; Epstein, P. R.; Grambsch, A.; Gubler, D. J.; Reiter, P.;
2 Romieu, I.; Rose, J. B.; Samet, J. M.; Trtanj, J. (2000b) The potential health impacts of climate variability
3 and change for the United States: executive summary of the report of the health sector of the U.S. national
4 assessment. Environ. Health Perspect. 108: 367-376.
5 Pearcy, R. W.; Bjorkman, O.; Caldwell, M. M.; Keeley, J. E.; Monson, R. K.; Strain, B. R. (1987) Carbon gain by
6 plants in natural environments. BioScience 37: 21-29.
7 Percy, K. E. (1991) Effects of acid rain on forest vegetation: morphological and non-mensurational growth effects.
8 In: Rennie, P. J.; Robitaille, G., eds. Effects of Acid Rain on Forest Resources: Proceedings of the
9 Conference ; June, 1983; Sainte-Foy, Quebec. Ottawa, Canada: Canadian Forestry Service; pp. 97-111.
10 Percy, W. H.; Sutherland, J.; Christensen, J. (1991) Paradoxical relationship between substrates and agonist-induced
11 contractions of opossum esophageal body and sphincter invitro. Dig. Dis. Sci. 36: 1057-1065.
12 Perry, K. D.; Cahill, T. A.; Eldred, R. A.; Butcher, D. D.; Gill, T. E. (1997) Long-range transport of North African
13 dust to the eastern United States. J. Geophys. Res. [Atmos] 102: 11,225-11,238.
14 Peterjohn, W. T.; Adams, M. B.; Gilliam, F. S. (1996) Symptoms of nitrogen saturation in two central Appalachian
15 hardwood forest ecosystems. Biogeochemistry 35: 507-522.
16 Peterson, J. T.; Flowers, E. C.; Rudisill, J. H. (1978) Urban-rural solar radiation and atmospheric turbidity
17 < measurements in the Los Angeles basin. J. Appl. Meteorol. 17: 1595-1609.
18 Peterson, G.; Allen, C. R.; Rolling, C. S. (1998) Ecological resilience, biodiversity, and scale. Ecosystems 1: 6-18.
19 Petuskey, W. T.; Richardson, D. A.; Dolske, D. A. (1995) Aspects of the deterioration of sandstone masonry in
20 Anasazi dwelling ruins at Mesa Verde National Park, Colorado, USA. Sci. Total Environ. 167: 145-159.
21 Pimentel, D.; Wilson, C.; McCullum, C.; Huang, R.; Dwen, P.; Flack, J.; Tran, Q.; Saltman, T.; Cliff, B. (1997)
22 Economic and environmental benefits of biodiversity. BioScience 47: 747-757.
23 Pimm, S. L. (1984) The complexity and stability of ecosystems. Nature (London) 307: 321-326.
24 Pitchford, M. L.; Malm, W. C. (1994) Development and applications of a standard visual index. Atmos. Environ.
25 28: 1049-1054.
26 Pitchford, M. L.; McMurry, P. H. (1994) Relationship between measured water vapor growth and chemistry of
27 atmospheric aerosol for Grand Canyon, Arizona, in winter 1990. Atmos. Environ. 28: 827-839.
28 Prescott, C. E.; Parkinson, D. (1985) Effects of sulphur pollution on rates of litter decomposition in a pine forest.
29 Can. J. Bot. 63:1436-1443.
30 Pryor, S.; Steyn, D. (1994) Visibility and ambient aerosols in southwestern British Columbia during REVEAL
31 (REgional Visibility Experimental Assessment in the Lower Fraser Valley): application of analytical
32 techniques for source apportionment and related visibility impacts of the REVEAL fine particulate optical
33 and scene data. Victoria, BC, Canada: British Columbia Ministry of Environment, Lands and Parks.
34 Available from: NTIS, Springfield, VA; MIC-96-00314/INS.
35 Pueschel, R. F. (1993) Potential climatic effects of anthropogenic aerosols. In: Jennings, S. G., ed. Aerosol effects
36 on climate. Tucson, AZ: University of Arizona Press; pp. 110-132.
37 Rapport, D. J.; Whitford, W. G. (1999) How ecosystems respond to stress: common properties of arid and aquatic
38 systems. BioScience 49:193-203.
39 Raynal, D. J.; Joslin, J. D.; Thornton, F. C.; Schaedle, M.; Henderson, G. S. (1990) Sensitivity of tree seedlings to
40 aluminum: III. Red spruce and loblolly pine. J. Environ. Qual. 19: 180-187.
41 Realini, M.; Negrotti, R.; Appollonia, L.; Vaudan, D. (1995) Deposition of particulate matter on stone surfaces: an
42 experimental verification of its effects on Carrara marble. In: Saiz-Jimenez, C., ed. The deterioration of
43 monuments: proceedings of the 2nd international symposium on biodeterioration and biodegradation;
44 January 1994; Sevilla, Spain. Sci. Total Environ. 167: 67-72.
45 Rennie, P. J. (1955) The uptake of nutrients by mature forest growth. Plant Soil 7:49-95.
46 Reuss, J. O. (1983) Implications of the calcium-aluminum exchange system for the effect of acid precipitation on
47 soils. J. Environ. Qual. 12: 591-595.
48 Reuss, J. O.; Johnson, D. W. (1986) Acid deposition and the acidification of soils and waters. New York, NY:
49 Springer-Verlag. (Billings, W. D.; Golley, F.; Lange, O. L.; Olson, J. S.; Remmert, H., eds. Ecological
50 studies: analysis and synthesis: v. 59).
51 Richards, L. W. (1973) Size distribution of pigment particles: measurement and characterization by light-scattering
52 techniques. In: Patton, T. C., ed. Pigment handbook. V. III. Characterization and physical relationships.
53 New York, NY: John Wiley & Sons; pp. 89-100.
54 Richards, L. W. (1999) Use of the deciview haze index as an indicator for regional haze. J. Air Waste Manage.
55 Assoc. 49: 1230-1237.
March 2001
4-180
DRAFT-DO NOT QUOTE OR CITE
-------
1 Richards, L. W.; Blanchard, C. L.; Blumenthal, D. L., eds. (1991) Navajo generating station visibility study.
2 Final Report. Petaluma, CA: Sonoma Technology, Inc.; report no. STI-90200-1124-FR.
3 Riggan, P. J.; Lockwood, R. N.; Lopez, E. N. (1985) Deposition and processing of airborne nitrogen pollutants in
4 Mediterranean-type ecosystems of southern California. Environ. Sci. Technol. 19: 781-789.
5 Rodney, D. R. (1952) The entrance of nitrogen compounds through the epidermis of apple leaves. Proc. Am. Soc.
6 Horde. Sci. 59: 99-102.
7 Roelofs, J, G. M.; Kempers, A. J.; Houdijk, A. L. F. M.; Jansen, J. (1985) The effect of air-borne ammonium
8 sulphate on Pinus nigra var. maritima in the Netherlands. Plant Soil 84: 45-56.
9 Roelofs, J. G. M.; Boxman, A. W.; Van Dijk, H. F. G. (1987) Effects of airborne ammonium on natural vegetation
10 and forests. In: Asman, W. A. H.; Diederen, H. S. M. A., eds. Ammonia and acidification: proceedings [of a]
11 symposium of the European Association for the Science of Air Pollution (EURAS AP); April; Bilthoven,
12 The Netherlands. European Association for the Science of Air Pollution; pp. 266-276.
13 Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. (1993) Sources of fine organic
14 aerosol. 3. Road dust, tire debris, and organometallic brake lining dust: roads as sources and sinks.
15 Environ. Sci. Technol. 27: 1892-1904.
16 Rosen, H.; Hansen, A. D. A.; Gundel, L.; Novakov, T. (1978) Identification of the optically absorbing component in
17 urban aerosols. Appl. Opt. 17: 3859-3861.
18 Rosenberg, N. J.; Blad, B. L.; Verma, S. B. (1983) Microclimate: the biological environment. 2nd ed. New York
19 NY: John Wiley & Sons.
20 Ross, D. M.; Malm, W. C.; Iyer, H. K. (1997) Human visual sensitivity to plumes with a Gaussian luminance
21 distribution: experiments to develop an empirical probability of detection model. J. Air Waste Manage.
22 Assoc. 47: 370-382.
23 Rovira, A. D.; Davey, C. B. (1974) Biology of the rhizosphere. In: Carson, E. W., ed. The plant root and its
24 environment: proceedings of an institute; July, 1971; Blacksburg, VA. Charlottesville, VA: University Press
25 of Virginia; pp. 153-204.
26 Sabbioni, C. (1995) Contribution of atmospheric deposition to the formation of damage layers. In: Jimenez-Saiz, C.,
27 ed. The deterioration of monuments: proceedings of the 2nd international symposium on biodeterioration and
28 biodegradation; January 1994; Sevilla, Spain. Sci. Total Environ. 167: 49-56.
29 Sabbioni, C.; Zappia, G.; Gobbi, G. (1992) Carbonaceous particles on carbonate building stones in a simulated
30 system. In: Proceedings of the 1992 European aerosol conference; September; Oxford, United Kingdom.
31 J. Aerosol Sci. 23(suppl. 1): S921-S924.
32 Sabbioni, C.; Zappia, G.; Gobbi, G. (1996) Carbonaceous particles and stone damage in a laboratory exposure
33 system. J. Geophys. Res. [Atmos.] 101: 19,621-19,627.
34 Sabbioni, C.; Zappia, G.; Ghedini, N.; Gobbi, G.; Favoni, O. (1998) Black crusts on ancient mortars.
35 Atmos. Environ. 32:215-223.
36 Saiz-Jimenez, C. (1993) Deposition of airborne organic pollutants on historic buildings. Atmos. Environ Part B
37 27:77-85.
38 Samecka-Cymerman, A.; Kempers, A. J. (1999) Bioindication of heavy metals in the town Wro/caw (Poland) with
39 evergreen plants. Atmos. Environ. 33: 419-430.
40 Sanyal, B.; Singhania, G. K. (1956) Atmospheric corrosion of metals: part I. J. Sci. Ind. Res. Sect. B 15: 448-455.
41 Saunders, P. J. W.; Godzik, S. (1986) Terrestrial vegetation-air pollutant interactions: nongaseous air pollutants.
42 In: Legge, A. H.; Krupa, S. V., eds. Air pollutants and their effects on the terrestrial ecosystem. New York,
43 NY: John Wiley & Sons; pp. 389-394. (Advances in environmental science and technology: v. 18).
44 Saxena, P.; Hildemann, L. M.; McMurry, P. H.; Seinfeld, J. H. (1995) Organics alter hygroscopic behavior of
45 atmospheric particles. J. Geophys. Res. [Atmos.] 100: 18,755-18,770.
46 Schiavon, N.; Chiavari, G.; Schiavon, G.; Fabbri, D. (1995) Nature and decay effects of urban soiling on granitic
47 building stones. In: Jimenez-Saiz, C., ed. The deterioration of monuments: proceedings of the 2nd
48 international symposium on biodeterioration and biodegradation; January 1994; Sevilla, Spain. Sci. Total
49 Environ. 167:87-101.
50 Schier, G. A.; Jensen, K. (1992) Atmospheric deposition effects on foliar injury and foliar leaching in red spruce.
51 In: Eager, C.; Adams, M. B., eds. Ecology and decline of red spruce in the eastern United States. New York,
52 NY: Springer-Verlag; pp. 271-294. [Billings, W. D.; Golley, F.; Lange, O. L.; Olson, J. D.; Remmert, H.,
53 eds. Ecological Studies, Analysis and Synthesis Ecological Studies, v. 96].
March 2001
4-181
DRAFT-DO NOT QUOTE OR CITE
-------
1 Schnug, E. (1997) Significance of sulphur for the quality of domesticated plants. In: Cram, W. J.; De Kok, L. J.;
2 Stulen, I.; Brunold, C.; Rennenberg, H., eds. Sulphur metabolism in higher plants: molecular,
3 ecophysiological and nutritional aspects. Leiden, The Netherlands: Backhuys Publishers; pp. 109-130.
4 Schulze, E.-D. (1989) Air pollution and forest decline in a spruce (Picea abies) forest. Science (Washington, DC)
5 244:776-783.
6 Schuster, P. F.; Reddy, M. M.; Sherwood, S. I. (1994) Effects of acid rain and sulfur dioxide on marble dissolution.
7 Mater. Perform. 33: 76-80.
8 Schwar, M. J. R. (1998) Nuisance dust deposition and soiling rate measurements. Environ. Technol. 19: 223-229.
9 Seckmeyer, G.; McKenzie, R. L. (1992) Increased ultraviolet radiation in New Zealand (45 degrees S) relative to
10 Germany (48 degrees N). Nature 359:135ff.
11 Seigneur, C.; Wu, X. A.; Constantinou, E.; Gillespie, P.; Bergstrom, R. W.; Sykes, I.; Venkatram, A.;
12 Karamchandani, P. (1997) Formulation of a second-generation reactive plume and visibility model. J. Air
13 Waste Manage. Assoc. 47:176-184.
14 Seinfeld, J. H. (1989) Urban air pollution: state of the science. Science 243: 745-752.
15 Sereda, P. J. (1974) Weather factors affecting corrosion of metals. In: Corrosion in natural environments: three
16 symposia presented at the seventy-sixth annual meeting [of the] American Society for Testing and Materials;
17 June 1973; Philadelphia, PA. Philadelphia, PA: American Society for Testing and Materials; pp. 7-22.
18 (ASTM special technical publication 558).
19 Shannon, J. D.; Trexler, E. C., Jr.; Sonnenblick, R. (1997) Modeling visibility for assessment. Armos. Environ.
20 31:3719-3727.
21 Sievering, H.; Rusch, D.; Marquez, L. (1996) Nitric acid, particulate nitrate and ammonium in the continental free
22 troposphere: nitrogen deposition to an alpine tundra ecosystem. Armos. Environ. 30:2527-2537.
23 Shodor Education Foundation, Inc.; Industrial Extension Service of North Carolina State University; North Carolina
24 Supercomputing Center. (1996) Air quality meteorology. Session 8: physical meteorology and visibility.
25 Part 3: visibility. Research Triangle Park, NC: U.S. Environmental Protection Agency, National Exposure
26 Research Laboratory and U.S. National Oceanic and Atmospheric Administration. Available:
27 http://www.shodor.org/metweb/outline.html [1999, November 24].
28 Shortle, W. C.; Smith, K. T. (1988) Aluminum-induced calcium deficiency syndrome in declining red spruce.
29 Science (Washington, DC) 240: 1017-1018.
30 Shortle, W. C.; Bondietti, E. A. (1992) Timing, magnitude, and impact of acidic deposition on sensitive forest sites.
31 Water Air Soil Pollut. 61:253-267.
32 Showak, W.; Dunbar, S. R. (1982) Effect of 1 percent copper addition on atmospheric corrosion of rolled zinc after
33 20 years' exposure. In: Dean, S. W.; Rhea, E. C., eds. Atmospheric corrosion of metals: a symposium;
34 May 1980; Denver, CO. Philadelphia, PA: American Society for Testing and Materials; pp. 135-162.
3 5 (ASTM special technical publication 767).
36 Shriner, D. S.; Henderson, G. S. (1978) Sulfur distribution and cycling in a deciduous forest watershed. J. Environ.
37 Qual. 7:392-397.
38 Simonich; S. L.; Kites, R. A. (1995) Organic pollutant accumulation in vegetation. Environ. Sci. Technol.
39 29:2905-2914.
40 Simpson, J. W.; Horrobin, P. J., eds. (1970) The weathering and performance of building materials. New York, NY:
41 Wiley-Interscience.
42 Sisler, J. F.; Cahill, T. A. (1993) Spacial and temporal patterns and the chemical composition of the haze and its
43 impact on visibility in Alaska. In: Proceedings [of the] 86th annual meeting & exhibition of the Air & Waste
44 Management Association. Volume 8: general environmental topics; June; Denver, CO. Pittsburgh, PA: Air &
45 Waste Management Association; 93-MP-4.03.
46 Sisler, J. F.; Malm, W. C. (1994) The relative importance of soluble aerosols to spatial and seasonal trends of
47 impaired visibility in the United States. Atmos. Environ. 28: 851 -862.
48 Sisler, J. F.; Malm, W. C. (2000) Interpretation of trends of PM2 5 and reconstructed visibility from the IMPROVE
49 network. J. Air Waste Manage. Assoc.: submitted.
50 Sisler, J. F.; Huffman, D.; Lattimer, D. A.; Malm, W. C.; Pitchford, M. L. (1993) Spatial and temporal patterns and
51 the chemical composition of the haze in the United States: an anlysis of data from the IMPROVE network,
52 1988-1991. Fort Collins, CO: Colorado State University, Cooperative Institute for Research in the
53 Atmosphere (CIRA).
54 Skeffington, R. A.; Wilson, E. J. (1988) Excess nitrogen deposition: issues for consideration. Environ. Pollut.
55 54:159-184.
March 2001
4-182
DRAFT-DO NOT QUOTE OR CITE
-------
1 Skerry, B, S.; Alavi, A.; Lindgren, K. I. (1988) Environmental and electrochemical test methods for the evaluation
2 of protective organic coatings. J. Coat. Technol. 60: 97-106.
3 Sloane, C. S.; Watson, J.; Chow, J.; Pritchett, L.; Richards, L. W. (1991) Size-segregated fine particle
4 measurements by chemical species and their impact on visibility impairment in Denver. Atmos. Environ.
5 Part A 25: 1013-1024.
6 Smith, W. H. (1974) Air pollution - effects on the structure and function of the temperate forest ecosystem.
7 Environ. Pollut. 6: 111-129.
8 Smith, W. H. (1990a) Forests as sinks for air contaminants: soil compartment. In: Air pollution and forests:
9 interactions between air contaminants and forest ecosystems. 2nd ed. New York, NY: Springer-Verlag;
10 pp. 113-146. (Springer series on environmental management).
11 Smith, W. H. (1990b) Forest biotic agent stress: air pollutants and disease caused by microbial pathogens. In: Air
12 pollution and forests: interactions between air contaminants and forest ecosystems. 2nd ed. New York, NY:
13 Springer-Verlag; pp. 366-397. (Springer series on environmental management).
14 Smith, W. H. (1990c) Forests as sinks for air contaminants: vegetative compartment. In: Air pollution and forests:
15 interactions between air contaminants and forest ecosystems. 2nd ed. New York, NY: Springer-Verlag;
16 pp. 147-180. (Springer series on environmental management).
17 Smith, W. H. (1990d) Forests as sources of hydrocarbons, particulates, and other contaminants. In: Air pollution
18 and forests: interactions between air contaminants and forest ecosystems. 2nd ed. New York, NY:
19 Springer-Verlag; pp. 83-112. (Springer series on environmental management).
20 Smith, W. H. (1990e) Forest nutrient cycling: toxic ions. In: Air pollution and forests: interactions between air
21 contaminants and forest ecosystems. 2nd ed. New York, NY: Springer-Verlag; pp. 225-268. (Springer series
22 on environmental management).
23 Smith, W. H. (1991) Air pollution and forest damage. Chem. Eng. News 69(45): 30-43.
24 Smith, W. H.; Staskawicz, B. J. (1977) Removal of atmospheric particles by leaves and twigs of urban trees:
25 some preliminary observations and assessment of research needs. Environ. Manage. (N. Y.) 1: 317-330.
26 Soulen, P. E.; Frederick, J. E. (1999) Estimating biologically active UV irradiance from satellite radiance
27 measurements: a sensitivity study. J. Geophys. Res. 104: 4117-4126.
28 Spence, J. W.; Haynie, F. H. (1972) Paint technology and air pollution: a survey and economic assessment.
29 Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Programs; publication no.
30 AP-103. Available from: NTIS, Springfield, VA; PB-210736.
31 Spence, J. W.; Haynie, F.; Upham, J. B. (1975) Effects of gaseous pollutants on paints: a chamber study. J. Paint
32 Technol. 47: 57-63.
33 Spence, J. W.; Haynie, F. H.; Lipfert, F. W.; Cramer, S. D.; McDonald, L. G. (1992) Atmospheric corrosion model
34 for galvanized steel structures. Corrosion 48: 1009-1019.
35 Spiker, E. C.; Comer, V. J.; Hosker, R. P.; Sherwood, S. I. (1992) Dry deposition of SO2 on limestone and marble:
36 the role of humidity. In: Rodrigues, J. D.; Henriques, F.; Jeremias, F. T., eds. Proceedings of the 7th
37 international congress on deterioration and conservation of stone; June; Lisbon, Portugal. Lisbon, Portugal:
38 Laboratorio Nacional de Engenharia Civil; pp. 397-406.
39 Spinka, J. (1971) Vliv znecisteneho ovzdusi na ovocne stromy a zeleninu [Effects of polluted air on fruit trees and
40 legumes]. Ziva 19: 13-15.
41 Stevens, R. K.; Dzubay, T. G.; Lewis, C. W.; Shaw, R. W., Jr. (1984) Source apportionment methods applied to the
42 determination of the origin of ambient aerosols that affect visibility in forested areas. Atmos. Environ.
43 18:261-272.
44 Stoddard, J. L. (1994) Long-term changes in watershed retention of nitrogen: its causes and aquatic consequences.
45 In: Baker, L. A., ed. Environmental chemistry of lakes and reservoirs. Washington, DC: American Chemical
46 Society; pp. 223-284. (Advances in chemistry series no. 237).
47 Stoddard, J. L.; Murdoch, P. S. (1991) Catskill Mountains. In: Charles, D. F., ed. Acidic deposition and aquatic
48 ecosystems: regional case studies. New York, NY: Springer-Verlag; pp. 237-271.
49 Storm, G. L.; Fosmire, G. J.; Bellis, E. D. (1994) Persistence of metals in soil and selected vertebrates in the vicinity
50 of the Palmerton zinc smelters. J. Environ. Qual. 23: 508-514.
51 Strandberg, H.; Johansson, L.-G. (1997) The formation of black patina on copper in humid air containing traces of
52 SO2. J. Electrochem. Soc. 144: 81 -89.
53 Suzuki, D. T. (1997) Barriers to perception: from a world of interconnection to fragmentation. In: Raven, P. H., ed.
54 Nature and Human Society: the quest for a sustainable world. Washington, DC: National Academy Press;
55 pp. 11-21.
March 2001
4-183
DRAFT-DO NOT QUOTE OR CITE
-------
1 Sydberger, T.; Ericsson, R. (1977) Laboratory testing of the atmospheric corrosion of steel. Werkst. Korros.
2 28: 154-158.
3 Sykes, R. I.; Gabruk, R. S. (1997) A second-order closure model for the effect of averaging time on turbulent plume
4 dispersion. J. Appl. Meteorol. 36: 1038-1045.
5 Systems Applications International, Inc. (1998) User's guide to the regulatory modeling system for aerosols and
6 deposition (REMSAD). San Rafael, CA: Systems Applications International, Inc.; report no.
7 SYSAPP98-96\42r2.
8 Tang, I. N. (1997) Thermodynamic and optical properties of mixed-salt aerosols of atmospheric importance.
9 J. Geophys. Res. [Atmos.] 102: 1883-1893.
10 Taylor, G. E., Jr.; Johnson, D. W.; Andersen, C. P. (1994) Air pollution and forest ecosystems: a regional to global
11 perspective. Ecol. Appl. 4: 662-689.
12 Tegen, I.; Lacis, A. A. (1996) Modelling of particle size distribution and its influence on the radiative properties of
13 mineral dust aerosol. J. Geophys. Res. Atmos. 101: 19,237-19,244.
14 Tegen, I.; Lacis, A. A.; Fung, I. (1996) The influence on climate forcing of mineral aerosols from disturbed soils.
15 Nature (London) 380: 419-422.
16 Thornton, F. C.; Schaedle, M.; Raynal, D. J. (1987) Effects of aluminum on red spruce seedlings in solution culture.
17 Environ. Exp. Bot. 27: 489-498.
18 Tilman, D. (1996) Biodiversity: population versus ecosystem stability. Ecology 77: 350-363.
19 Tilman, D. (2000) Causes, consequences and ethics of biodiversity. Nature 405: 208-211.
20 Tilman, D.; Downing, J. A. (1994) Biodiversity and stability in grasslands. Nature (London) 367: 363-365.
21 Tong, S. T. Y. (1991) The retention of copper and lead particulate matter in plant foliage and forest soil.
22 Environ. Int. 17:31-37.
23 Trettin, C.C.; Johnson, D.W.; Todd, D.E., Jr. (1999) Forest nutrient and carbon pools at Walker Branch watershed:
24 changes during a 21-year-period. Soil Sci. Soc. Arner. J. 63: 1436-1448.
25 Trijonis, J.; Shapland, D. (1979) Existing visibility levels in the United States: isopleth maps of visibility in
26 suburban/nonurban areas during 1974-76. Research Triangle Park, NC: U.S. Environmental Protection
27 Agency, Office of Air Quality Planning and Standards; report no. EPA-450/5-79-010. Available from:
28 NTIS, Springfield, VA; PB80-156177.
29 Trijonis, J. C.; Pitchford, M.; McGown, M. (1987) Preliminary extinction budget results from the RESOLVE
30 program. In: Bhardwaja, P. S., ed. Visibility protection: research and policy aspects, an APCA international
31 specialty conference; September 1986; Grand Teton National Park, W Y. Pittsburgh, PA: Air Pollution
32 Control Association; pp. 872-883. (APCA transactions series no. TR-10).
33 Trijonis, J. C.; Malm, W. C.; Pitchford, M.; White, W. H. (1991) Visibility: existing and historical
34 conditions-causes and effects. In: Irving, P. M., ed. Acidic deposition: state of science and technology,
35 volume III: terrestrial, materials, health and visibility effects. Washington, DC: The U.S. National Acid
36 Precipitation Assessment Program. (State of science and technology report no. 24).
37 Turner, J.; Kelly, J. (1981) Relationships between soil nutrients and vegetation in a north coast forest, New South
38 Wales. Aust. For. Res. 11: 201-208
39 Turner, J.; Lambert, M. J. (1980) Sulfur nutrition of forests. In: Shriner, D. S.; Richmond, C. R.; Lingberg, S. E.,
40 eds. Atmospheric sulfur deposition: environmental impact and health effects, proceedings of the second life
41 sciences symposium; October 1979; Gatlinburg, TN. Ann Arbor, MI: Ann Arbor Science Publishers, Inc.;
42 pp. 321-333.
43 Turner, J.; Lambert, M. J.; Gessel, S. P. (1977) Use of foliage sulphate concentrations to predict response to urea
44 application by Douglas-fir. Can. J. For. Res. 7: 476-480.
45 Twomey, S. (1974) Pollution and the planetary albedo. Atmos. Environ. 8: 1251-1256.
46 Twomey, S. (1977) Atmospheric aerosols. Amsterdam, The Netherlands: Elsevier Scientific Publishing Company;
47 p. 237.
48 Tyler, G. (1972) Heavy metals pollute nature, may reduce productivity. Ambio 1: 52-59.
49 U.S. Code. (1991) Clean Air Act, §109, national ambient air quality standards. U. S. C. 42: §7409.
50 U.S. Environmental Protection Agency. (1979) Protecting visibility: an EPA report to Congress. Research Triangle
51 Park, NC: Office of Air Quality Planning and Standards; EPA report no. EPA-450/5-79-008. Available from:
52 NTIS, Springfield, VA; PB80-220320.
53 U.S. Environmental Protection Agency. (1982) Air quality criteria for particulate matter and sulfur oxides. Research
54 Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
55 Office; EPA report no. EPA-600/8-82-029aF-cF. 3v. Available from: NTIS, Springfield, VA; PB84-156777.
March 2001
4-184
DRAFT-DO NOT QUOTE OR CITE
-------
1 U.S. Environmental Protection Agency. (1986 ) Air quality criteria for lead. Research Triangle Park, NC: Office of
2 Health and Environmental Assessment, Environmental Criteria and Assessment Office; EPA report no.
3 EPA-600/8-83/028aF-dF. 4v. Available from: NTIS, Springfield, VA; PB87-142378.
4 U.S. Environmental Protection Agency. (1987) Assessing the risks of trace gases that can modify the stratosphere.
5 Washington, DC: U.S. Environmental Protection Agency, Office of Air Radiation; report no. EPA
6 400/1-87/001A-H.
7 U.S. Environmental Protection Agency. (1988) Workbook for plume visual impact screening and analysis. Research
8 Triangle Park, NC: Office of Air Quality Planning and Standards; report no. EPA-450/4-88-015.
9 Available from: NTIS, Springfield, VA; PB89-151286.
10 U.S. Environmental Protection Agency. (1992) Interagency workgroup on air quality modeling (IWAQM) work
11 plan rationale. Research Triangle Park, NC: Office of Air Quality Planning and Standards; report no.
12 EPA-454/R-92-001.
13 U.S. Environmental Protection Agency. (1993) Air quality criteria for oxides of nitrogen. Research Triangle Park,
14 NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment Office; report
15 nos. EPA/600/8-9!/049aF-cF. 3v. Available from: NTIS, Springfield, VA; PB95-124533, PB95-124525, and
16 PB95-124517.
17 U.S. Environmental Protection Agency. (1995a) Inter-agency workgroup on air quality modeling (IWAQM):
18 assessment of phase I recommendations regarding the use of MESOPUFF II. Research Triangle Park, NC:
19 U.S. Environmental Protection Agency, Emissions, Monitoring and Analysis Division; report no.
20 EPA-454/R-95-006. Available from: NTIS, Springfield, VA; PB95-241931. P. 106 (1995a).
21 U.S. Environmental Protection Agency. (1995b) Testing of meteorological and dispersion models for use in
22 regional air quality modeling. Research Triangle Park, NC: Office of Air Quality Planning and Standards;
23 report no. EPA-454/R-95-005. Available from: NTIS, Springfield, VA; PB95-215638.
24 U.S. Environmental Protection Agency. (1995c) Interim findings on the status of visibility research. Research
25 Triangle Park, NC: Office of Research and Development; report no. EPA/600/R-95/021.
26 U.S. Environmental Protection Agency. (1996a) Air quality criteria for particulate matter. Research Triangle Park,
27 NC: National Center for Environmental Assessment-RTP Office; report nos. EPA/600/P-95/001aF-cF. 3v.
28 U.S. Environmental Protection Agency. (1996b) Air quality criteria for ozone and related photochemical oxidants.
29 Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF.
30 3v. Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available
31 online at: www.epa.gov/ncea/ozone.htm.
32 U.S. Environmental Protection Agency. (1997a) Nitrogen oxides: impacts on public health and the environment.
33 Washington, DC: Office of Air and Radiation; August. Available:
34 www.epa.gov/rtncaaal/tl/reports/noxrept.pdf [1999, November 24].
35 U.S. Environmental Protection Agency. (1997b) Summary of the particulate monitoring program. Washington, DC:
36 Office of Air and Radiation.
37 U.S. Environmental Protection Agency. (1999) Protection of visibility. C. F. R. 40: "Section" 51.300-51.309.
38 U.S. Environmental Protection Agency. (2000a) Deposition of air pollutants to the great waters. Third report to
39 Congress. [Executive Summary]. Research Triangle Park, NC: U.S. Environmental Protection Agency,
40 Office of Air Quality Planning and Standards; report no. EPA-453/R-00-005.
41 U.S. Environmental Protection Agency. (2000b) Latest findings on national air quality: 1999 status and trends.
42 Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and
43 Standards; report no. EPA-454/F-00-002.
44 U. S. Global Change Research Program (USGCRP). (2000) Climate Change Impacts on the United States: the '
45 Potential Consequences of Climate Variability and Change (Overview), Report of National Assessment
46 Synthesis Team (NAST). NSTC Review Draft (September).
47 United Nations Environment Programme (UNEP). (1986) Report of the international conference on the assessment
48 of the role of carbon dioxide and of other greenhouse gases in climate variations and associated impacts;
49 October 1985; Villach, Austria. Paris, France: International Council of Scientific Unions. [World
50 Meteorological Organization series: WMO no. 661].
51 United Nations Environment Programme (UNEP). (1998) Environmental effects of ozone depletion: 1998
52 assessment. J. Photochem. Photobiol. B 46: 1-4.
53 Unsworth, M. H.; Wilshaw, J. C. (1989) Wet, occult and dry deposition of pollutants on forests. Agric. For.
54 Meteorol. 47: 221-238.
March 2001
4-185
DRAFT-DO NOT QUOTE OR CITE
-------
1 Van Breemen, N.; Van Dijk, H. F. G. (1988) Ecosystem effects of atmospheric deposition of nitrogen in The
2 Netherlands. In: Dempster, J. P.; Manning, W. J., eds. Excess nitrogen deposition. Environ. Pollut.
3 54:249-274.
4 Van Breemen, N.; Burrough, P. A.; Velthorst, E. J.; Van Dobben, H. F.; De Wit, T.; Ridder, T. B.; Reijnders, H. F.
5 R. (1982) Soil acidification from atmospheric ammonium sulphate in forest canopy throughfall. Nature
6 (London) 299: 548-550.
7 Van Breemen, N.; Mulder, J.; Driscoll, C. T. (1983) Acidification and alkalinization of soils. Plant Soil
8 75:283-308.
9 Van der Leun, J. C.; Tang, X.; Tevini, M. (1995) Environmental effects of ozone depletion: 1994 assessment.
10 Ambio24:138.
11 Van der Leun, J. C.; Tang, X.; Tevini, M., eds. (1998) Environmental effects of ozone depletion: 1998 assessment.
12 J. Photochem. Photobiol. B: Biology 46(1-3): 1-108.
13 Van Dijk, H. F. G.; Roelofs, J. G. M. (1988) Effects of excessive ammonium deposition on the nutritional status
14 and condition of pine needles. Physiol. Plant. 73: 494-501.
15 Van Dorland, R.; Dentener, F. J.; Lelieveld, J. (1997) Radiative forcing due to tropospheric ozone and sulfate
16 aerosols. J. Geophys. Res. [Atmos.] 102:28,079-28,100.
17 Van Miegroet, H.; Cole, D. W. (1984) The impact of nitrification on soil acidification and cation leaching in red
18 alder ecosystem. J. Environ. Qual. 13: 586-590.
19 Van Miegroet, H.; Johnson, D. W.; Todd, D. E. (1993) Foliar response of red spruce saplings to fertilization with
20 Ca and Mg in the Great Smoky Mountains National Park. Can. J. For. Res. 23: 89-95.
21 Vasconcelos, L. A. de P.; Kahl, J. D. W.; Liu, D.; Macias, E. S.; White, W. H. (1996) Patterns of dust transport to
22 the Grand Canyon. Geophys. Res. Lett. 23: 3187-3190.
23 Viles, H. A. (1990) The early stages of building stone decay in an urban environment. Atmos. Environ.
24 24A: 229-232.
25 Vitousek, P. M.; Mooney, H. A.; Lubchenco, J.; Melillo, J. M. (1997) Human domination of Earth's ecosystems.
26 Science (Washington, DC) 277:494-499.
27 Vogt, K. A.; Dahlgren, R.; Ugolini, F.; Zabowski, D.; Moore, E. E.; Zasoski, R. (1987a) Aluminum, Fe, Ca, Mg, K,
28 Mn, Cu, Zn and P in above- and belowground biomass. I. Abies amabilis and Tsuga mertensiana.
29 Biogeochemistry 4:277-294.
30 Vogt, K. A.; Dahlgren, R.; Ugolini, F.; Zabowski, D.; Moore, E. E.; Zasoski, R. (1987b) Aluminum, Fe, Ca, Mg, K,
31 Mn, Cu, Zn and P in above- and belowground biomass. II. Pools, and circulation in a subalpine Abies
32 amabilis stand. Biogeochemistry 4: 295-311.
33 Vose, J. M.; Swank, W. T. (1990) Preliminary estimates of foliar absorption of 15N-labeled nitric acid vapor
34 (HNO3) by mature eastern white pine (Pinus strobus). Can. J. For. Res. 20: 857-860.
35 Waggoner, A. P.; Weiss, R. E. (1980) Comparison of fine particle mass concentration and light scattering extinction
36 in ambient aerosol. Atmos. Environ. 14:623-626.
37 Waggoner, A. P.; Weiss, R. E.; Ahlquist, N. C.; Covert, D. S.; Will, S.; Charlson, R. J. (1981) Optical
38 characteristics of atmospheric aerosols. In: White, W. H.; Moore, D. J.; Lodge, J. P., Jr., eds. Plumes and
39 visibility: measurements and model components: proceedings of the symposium; November 1980;
40 Grand Canyon National Park, AZ. Atmos. Environ. 15: 1891 -1909.
41 Wagrowski, D. M.; Kites, R. A. (1997) Polycyclic aromatic hydrocarbon accumulation in urban, suburban, and rural
42 vegetation. Environ. Sci. Technol. 31: 279-282.
43 Wall, D. H. (1999) Biodiversity and ecosystem functioning. BioScience 49: 107-108.
44 Wall, D. H.; Moore, J. C. (1999) Interactions underground: soil biodiversity, mutualism, and ecosystem processes.
45 Bioscience49:109-117.
46 Walton, J. R.; Johnson, J. B.; Wood, G. C. (1982) Atmospheric corrosion initiation by sulphur dioxide and
47 particulate matter: II. characterisation and corrosivity of individual particulate atmospheric pollutants.
48 Br. Corros. J. 17: 65-70.
49 Wang, W. C.; Yung, Y. L.; Lacis, A. A.; Mo, T.; Hansen, J. E. (1976) Greenhouse effects due to man-made
50 perturbations of trace gases. Science (Washington, DC) 194: 685-690.
51 Waring, R. H. (1987) Nitrate pollution: a particular danger to boreal and subalpine coniferous forests. In:
52 Fujimori, T.; Kimura, M., eds. Human impacts and management of mountain forests: [proceedings of a
53 symposium]. Ibaraki, Japan: Forestry and Forest Products Research Institute; pp. 93-105.
54 Waring, R. H.; Schlesinger, W. H. (1985) The carbon balance of trees. In: Forest ecosystems: concepts and
55 management. Orlando, FL: Academic Press, Inc.; pp. 7-37.
March 2001
4-186
DRAFT-DO NOT QUOTE OR CITE
_
-------
1 Watson, J. G.; Chow, J. C. (1994) Clear sky visibility as a challenge for society. Annu. Rev. Energy Environ.
2 19:241-266.
3 Watson, J. G.; Chow, J. C.; Richards, L. W.; Haase, D. L.; McDade, C.; Dietrich, D. L.; Moon, D.; Sloane, C.
4 (1991) The 1989-90 Phoenix urban haze study, volume II: the apportionment of light extinction to sources
5 [final report]. Reno, NV: University and Community College System of Nevada, Desert Research Institute;
6 DRI document no. 8931.5F1.
7 Webb, A. H.; Bawden, R. J.; Busby, A. K.; Hopkins, J. N. (1992) Studies on the effects of air pollution on
8 limestone degradation in Great Britain. Atmos. Environ. Part B 26: 165-181.
9 Wedin, D. A.; Tilman, D. (1996) Influence of nitrogen loading and species composition on the carbon balance of
10 grasslands. Science 274: 1720-1723.
11 Weinbaum, S. A.; Neumann, P. M. (1977) Uptake and metabolism of 15N-labeled potassium nitrate by French
12 prune (Prunus domestica L.) leaves and the effects of two surfactants. J. Am. Soc. Hortic. Sci. 102: 601-604.
13 Wellburn, A. R. (1990) Tansley review no. 24: Why are atmospheric oxides of nitrogen usually phytotoxic and not
14 alternative fertilizers? New Phytol. 115: 395-429.
15 Wenny, B. N.; Schafer, J. S.; DeLuisi, J. J.; Saxena, V. K.; Barnard, W. F.; Petropavlovskikh, I. V.; Vergamini,
16 A. J. (1998) A study of regional aerosol radiative properties and effects on ultraviolet-B radiation.
17 J. Geophys. Res. [Atmos.] 103: 17,083-17,097.
18 Wesselink, L. G.; Meiwes, K.-J.; Matzner, E.; Stein, A. (1995) Long-term changes in water and soil chemistry in
19 spruce and beech forests, Soiling, Germany. Environ. Sci. Technol. 29: 51-58.
20 Westman, W. E. (1977) How much are nature's services worth? Measuring the social benefits of ecosystem
21 functioning is both controversial and illuminating. Science (Washington, DC) 197: 960-964.
22 White, W. H. (1976) Reduction of visibility by sulphates in photochemical smog. Nature (London) 264: 735-736.
23 White, W. H. (1997) Deteriorating air or improving measurements? On interpreting concatenate time series.
24 J. Geophys. Res. [Atmos.] 102: 6813-6821.
25 White, W. H.; Macias, E. S. (1990) Light scattering by haze and dust at Spirit Mountain, Nevada. In: Mathai, C. V.,
26 ed. Visibility and fine particles: an A&WMA/EPA international specialty conference; October 1989; Estes
27 Park, CO. Pittsburgh, PA; Air & Waste Management Association; pp. 914-922. (A&WMA transactions
28 series no. TR-17).
29 White, W. H.; Seigneur, C.; Heinold, D. W.; Eltgroth, M. W.; Richards, L. W.; Roberts, P. T.; Bhardwaja, P. S.;
30 Conner, W. D.; Wilson, W. E., Jr. (1985) Predicting the visibility of chimney plumes: an intercomparison of
31 four models with observations at a well-controlled power plant. Atmos. Environ. 19: 515-528.
32 White, W. H.; Macias, E. S.; Nininger, R. C.; Schorran, D. (1994) Size-resolved measurements of light scattering by
33 ambient particles in the southwestern U.S.A. Atmos. Environ. 28: 909-921.
34 White, W. H.; Macias, E. S.; Vasconcelos, L. A. de P.; Farber, R. J.; Mirabella, V. A.; Green, M. C.; Pitchford,
35 M. L. (1999) Tracking regional background in a haze attribution experiment. J. Air Waste Manage. Assoc.
36 49:599-602.
37 Williams, M. W.; Baron, J. S.; Caine, N.; Sommerfeld, R.; Sanford, R. (1996) Nitrogen saturation in the
38 Rocky Mountains. Environ. Sci. Technol. 30: 640-646.
39 Wilson, E. O. (1997) The creation of biodiversity. In: Raven, P. H., ed. Nature and human society: the quest for a
40 sustainable world: proceedings of the 1997 forum on biodiversity; October; sponsored by the National
41 Research Council. Washington, DC: National Academy Press.
42 Wilson, W. E.; Suh, H. H. (1997) Fine particles and coarse particles: concentration relationships relevant to
43 epidemiologic studies. J. Air Waste Manage. Assoc. 47: 1238-1249.
44 Winner, W. E.; Atkinson, C. J. (1986) Absorption of air pollution by plants, and consequences for growth.
45 Trends Ecol. Evol. 1: 15-18.
46 Winner, W. E.; Bewley, J. D. (1978a) Terrestrial mosses as bioindicators of SO2 pollution stress: synecological
47 analysis and the index of atmospheric purity. Oecologia 35: 221-230.
48 Winner, W. E.; Bewley, J. D. (1978b) Contrasts between bryophyte and vascular plant synecological responses in
49 an SO2-stressed white spruce association in central Alberta. Oecologia 33: 311-325.
50 Wittenburg, C.; Dannecker, W. (1992) Dry deposition and deposition velocity of airborne acidic species upon
51 different sandstones. In: Proceedings of the 1992 European aerosol conference; September; Oxford,
52 United Kingdom. J. Aerosol Sci. 23(suppl. 1): S869-S872.
53 Wolff, G. T.; Collins, D. C.; Rodgers, W. R.; Verma, M. H.; Wong, C. A. (1990) Spotting of automotive finishes
54 from the interactions between dry deposition of crustal material and wet deposition of sulfate. J. Air Waste
55 Manage. Assoc. 40: 1638-1648.
March 2001
4-187 DRAFT-DO NOT QUOTE OR CITE
-------
1 World Health Organization. (1997) Nitrogen oxides. 2nd ed. Geneva, Switzerland: World Health Organization.
2 (Environmental health criteria 188).
3 World Meteorological Organization. (1988) Developing policies for responding to climatic change: a summary of
4 the discussions and recommendations of workshops; September-October 1987; Villach, Austria; and
5 November 1987; Bellagio, Austria. Geneva, Switzerland: World Meteorological Organization; report no.
6 WMO/TD; no. 225. [World Climate Impact Programme series report no. WCIP-1].
7 World Meteorological Organization. (1999) Scientific assessment of ozone depletion: 1998. Geneva, Switzerland:
8 World Meteorological Organization, Global Ozone and Monitoring Project; report no. 44.
9 World Resources Institute. (2000) World resources 2000-2001: people and ecosystems: the fraying web of life.
10 Washington, DC: World Resources Institute.
11 Wu, P.-M.; Okada, K. (1994) Nature of coarse nitrate particles in the atmosphere-a single particle approach.
12 Atmos. Environ. 28:2053-2060.
13 Wu, X. A.; Seigneur, C.; Bergstrom, R. W. (1996) Evaluation of a sectional representation of size distributions for
14 calculating aerosol optical properties. J. Geophys. Res. [Atmos.] 101: 19,277-19,283.
15 Xu, J. R.; Balik, C. M. (1989) Infrared attenuated total reflectance study of latex paint films exposed to aqueous
16 sulfur dioxide. J. Appl. Polym. Sci. 38:173-183.
17 Yanai, R. D.; Siccama, T. G.; Arthur, M.. A.; Federer, C. A.; Friedland, A. J. (1999) Accumulation and depletion
18 of base cations in forest floors in the northeastern United States. Ecology 80: 2774-2787.
19 Yerrapragada, S. S.; Jaynes, J. H.; Chirra, S. R.; Gauri, K. L. (1994) Rate of weathering of marble due to dry
20 deposition of ambient sulfur and nitrogen dioxides. Anal. Chem. 66: 655-659.
21 Yocom, J. E.; Grappone, N. (1976) Effects of power plant emissions on materials. Palo Alto, CA: Electric Power
22 Research Institute; report no. EPRI/EC-139. Available from: NTIS, Springfield, VA; PB-257 539.
23 Yocom, J. E.; Upham, J. B. (1977) Effects of economic materials and structures. In: Stem, A. C., ed. Air pollution.
24 3rd ed. Volume II. New York, NY: Academic Press; pp. 93-94.
25 Young, P. (1996) Pollution-fueled "biodeterioration" threatens historic stone: researchers are probing the role
26 microorganisms play in the decay of buildings and artwork. Environ. Sci. Technol. 30: 206A-208A.
27 Young, J. R.; Ellis, E. C.; Hidy, G. M. (1988) Deposition of air-borne acidifiers in the western environment.
28 J. Environ. Qual. 17:1-26.
29 Youngs, W. D.; Rutzke, M.; Gutenmann, W. H.; Lisk, D. J. (1993) Nickel and vanadium in foliage in the vicinity of
30 an oil-fired power plant. Chemosphere 27: 1269-1272.
31 Zannetti, P.; Tombach, I.; Balson, W. (1990) Calculation of visual range improvements from SO2 emission
32 controls—I. semi-empirical methodology. Atmos. Environ. Part A 24: 2361-2368.
33 Zannetti, P.; Tombach, I.; Cvencek, S.; Balson, W. (1993) Calculation of visual range improvements from SO2
34 emission controls-II. an application to the eastern United States. Atmos. Environ. Part A 27:1479-1490.
35 Zappia, G.; Sabbioni, C.; Gobbi, G. (1993) Non-carbonate carbon content on black and white areas of damaged
3 6 stone monuments. Atmos. Environ. 27A: 1117-1121.
37 Zappia, G.; Sabbioni, C.; Pauri, M. G.; Gobbi, G. (1994) Mortar damage due to airborne sulfur compounds in a
38 simulation chamber. Mater. Struct. 27: 469-473.
39 Zappia, G.; Sabbioni, C.; Riontino, C.; Gobbi, G.; Favoni, O. (1998) Exposure tests of building materials in urban
40 atmosphere. Sci. Total Environ. 224: 235-244.
41 Zepp, R. G.; Callaghan, T. V.; Erickson, D. J. (1998) Effects of enhanced solar ultraviolet radiation on
42 biogeochemical cycles. J. Photochem. Photobiol. B 46: 69-82.
43 Zhang, X. Q.; McMurry, P. H.; Hering, S. V.; Casuccio, G. S. (1993) Mixing characteristics and water content of
44 submicron aerosols measured in Los Angeles and at the Grand Canyon. Atmos. Environ. Part A
45 27:1593-1607.
46 Zhang, X.; Turpin, B. J.; McMurry, P. H.; Hering, S. V.; Stolzenburg, M. R. (1994) Mie theory evaluation of
47 species contributions to 1990 wintertime visibility reduction in the Grand Canyon. Air Waste 44: 153-162.
48
March 2001
4-188
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
APPENDIX 4A
Excerpted Key Points from the Executive Summary of the
World Meteorological Organization 1998
Assessment of Stratospheric Ozone Depletion
(World Meteorological Organization, 1999)
Among the provisions of the 1987 Montreal Protocol on Substances that Deplete the Ozone
Layer was the requirement that the Parties to the Protocol base their future decisions on available
scientific, environmental, technical, and economic information, as assessed by worldwide expert
communities. Advances in the understanding of ozone science over this decade were assessed in
1988, 1989, 1991, and 1994. This information was input to the subsequent Amendments and
Adjustments of the 1987 Protocol. The 1998 assessment summarized below is the fifth in that
series.
Recent Major Scientific Findings and Observations
Since the Scientific Assessment of Ozone Depletion: 1994, significant advances have
continued to be made in understanding of the impact of human activities on the ozone layer, the
influence of changes in chemical composition on the radiative balance of the Earth's climate, and,
indeed, the coupling of the ozone layer and the climate system. Numerous laboratory
investigations, atmospheric observations, and theoretical and modeling studies have produced
several key ozone- and climate-related findings that are discussed below.
• The total combined abundance of ozone-depleting compounds in the lower atmosphere
peaked in about 1994 and now is slowly declining. Total chlorine is declining, but total
bromine is still increasing. As forecast in the 1994 Assessment, the long period of increasing
total chlorine abundances—primarily from the chlorofluorocarbons (CFCs), carbon
tetrachloride (CC14), and methyl chloroform (CH3CCI3)—has ended. The peak total
tropospheric chlorine abundance was 3.7 ± 0.1 parts per billion (ppb) between mid-1992 and
mid-1994. The declining abundance of total chlorine results principally from reduced
emissions of methyl chloroform. Chlorine from the major CFCs is still increasing slightly. The
March 2001
4A-1
DRAFT-DO NOT QUOTE OR CITE
-------
1 abundances of most of the halons continue to increase (for example, Halon-1211, almost 6%
2 per year in 1996), but the rate has slowed in recent years. These halon increases likely are
3 caused by emissions in the 1990s from the halon "bank," largely in developed countries, and
4 new production of halons in developing countries. The observed abundances of CFCs and
5 chlorocarbons hi the lower atmosphere are consistent with reported emissions.
6 • The observed abundances of the substitutes for the CFCs are increasing. Abundances of
7 the hydrochlorofluorocarbons (HCFCs) and hydrofluorocarbons (HFCs) are increasing as a
8 result of continuation of earlier uses and of their use as substitutes for the CFCs. hi 1996, the
9 HCFCs contributed about 5% to the tropospheric chlorine from the long-lived gases. This
10 addition from the substitutes offsets some of the decline in tropospheric chlorine associated
11 with methyl chloroform, but is still about 10 times less than that from the total tropospheric
12 chlorine growth rate throughout the 1980s. The atmospheric abundances of HCFC-141b and
13 HCFC-142b calculated from reported emissions data are factors of 1.3 and 2, respectively,
14 smaller than observations. Observed and calculated abundances agree for HCFC-22 and
15 HFC-134a.
16 • The combined abundance of stratospheric chlorine and bromine is expected to peak
17 before the year 2000. The delay in this peak in the stratosphere compared with the lower
18 atmosphere reflects the average time required for surface emissions to reach the lower
19 stratosphere. The observations of key chlorine compounds in the stratosphere up through the
20 present show the expected slower rate of increase and show that the peak had not occurred at
21 the time of the most recent observations that were analyzed for this assessment.
22 • The role of methyl bromide as an ozone-depleting compound is now considered to be less
23 than was estimated in the 1994 Assessment, although significant uncertainties remain.
24 The current best estimate of the Ozone Depletion Potential (ODP) for methyl bromide (CH3Br)
25 is 0.4, compared with an ODP of 0.6 estimated previously. The change is caused primarily by
26 both an increase in estimates of ocean removal processes and identification of an uptake by
27 soils, with a smaller contribution from change in our estimate of the atmospheric removal rate.
28 Recent research has shown that the science of atmospheric methyl bromide is complex and still
29 not well understood. Current understanding of the sources and sinks of atmospheric methyl
30 bromide is incomplete.
31
March 2001
4A-2
DRAFT-DO NOT QUOTE OR CITE
-------
1 • The rate of decline in stratospheric ozone at mid-latitudes has slowed; hence, the
2 projections of ozone loss made in the 1994 assessment are larger than what has actually
3 occurred. Total column ozone decreased significantly at mid-latitudes (25 to 60°) between
4 1979 and 1991, with estimated linear downward trends of 4.0, 1.8, and 3.8% per decade,
5 respectively, for northern mid-latitudes in winter/spring, northern mid-latitudes in summer/fall,
6 and southern mid-latitudes year round. However, since 1991, the linear trend observed during
7 the 1980s has not continued, rather, total column ozone has been almost constant at mid-
8 latitudes in both hemispheres since the recovery from the 1991 Mt. Pinatubo eruption. The
9 observed total column ozone losses from 1979 to the period 1994 to 1997 are about 5.4, 2.8,
10 and 5.0%, respectively, for northern mid-latitudes in winter/spring, northern mid-latitudes in
11 summer/fall, and southern mid-latitudes year round, rather than the values projected in the 1994
12 assessment assuming a linear trend: 7.6, 3.4, and 7.2%, respectively. Understanding of how
13 changes in stratospheric chlorine/bromine and aerosol loading affect ozone suggests some of
14 the reasons for the unsuitability of using a linear extrapolation of the pre-1991 ozone trend to
15 the present.
16 • The springtime Antarctic ozone hole continues unabated. The extent of ozone depletion has
17 remained essentially unchanged since the early 1990s. This behavior is expected given the
18 near-complete destruction of ozone within the Antarctic lower stratosphere during springtime.
19 The factors contributing to the continuing depletion are well understood.
20 • The link between the long-term buildup of chlorine and the decline of ozone in the upper
21 stratosphere has been firmly established. Model predictions based on the observed buildup
22 of stratospheric chlorine in the upper stratosphere indicate a depletion of ozone that is in good
23 quantitative agreement with the altitude and latitude dependence of the measured ozone decline
24 during the past several decades, which peaks at about 7% per decade near 40 km at mid-
25 latitudes in both hemispheres.
26 • The late-winter/spring ozone values in the Arctic were unusually low in six out of the last
27 nine years, the six being years that are characterized by unusually cold and protracted
28 stratospheric winters. The possibility of such depletions was predicted in the 1989
29 assessment. Minimum Arctic vortex temperatures are near the threshold for large chlorine
30 activation. Therefore, the year-to-year variability in temperature, which is driven by
31 meteorology, leads to particularly large variability in ozone for current chlorine loading. As a
March 2001
4A-3
DRAFT-DO NOT QUOTE OR CITE
-------
1 result, it is not possible to forecast the behavior of Arctic ozone for a particular year. Elevated
2 stratospheric halogen abundances over the next decade or so imply that the Arctic will continue
3 to be vulnerable to large ozone losses.
4 • The understanding of the relation between increasing surface UV-B radiation and
5 decreasing column ozone has been further strengthened by ground-based observations,
6 and newly developed satellite methods show promise for establishing global trends in UV
7 radiation. The inverse dependence of surface UV radiation and the overhead amount of ozone,
8 which was demonstrated in earlier assessments, has been further demonstrated and quantified
9 by ground-based measurements under a wide range of atmospheric conditions. In addition, the
10 influences of other variables, such as clouds, particles, and surface reflectivity, are better
11 understood. These data have assisted the development of a satellite-based method to estimate
12 global UV changes, taking into account the role of cloud cover. The satellite estimates for 1979
13 through 1992 indicate that the largest UV increases occur during spring at high latitudes in both
14 hemispheres.
15 • Stratospheric ozone losses have caused a cooling of the global lower stratosphere and
16 global-average negative radiative forcing of the climate system. The decadal temperature
17 trends in the stratosphere have now been better quantified. Model simulations indicate that
18 much of the observed downward trend in lower stratospheric temperatures (about 0.6 °C per
19 decade from 1979 to 1994) is attributed to the ozone loss in the lower stratosphere. A lower
20 stratosphere that is cooler results in less infrared radiation reaching the surface/troposphere
21 system. Radiative calculations, using extrapolations based on the ozone trends reported in the
22 1994 assessment for reference, indicate that stratospheric ozone losses since 1980 may have
23 offset about 30% of the positive forcing because of increases in the well-mixed greenhouse
24 gases (i.e., carbon dioxide, methane, nitrous oxide, halocarbons) over the same time period. The
25 climatic impact of the slowing of mid-latitude ozone trends and the enhanced ozone loss in the
26 Arctic has not yet been assessed.
27 • Based on past emissions of ozone-depleting substances and a projection of the maximum
28 allowances under the Montreal Protocol into the future, the maximum ozone depletion is
29 estimated to lie within the current decade or the next two decades, but its identification
30 and the evidence for the recovery of the ozone layer lie still further ahead. The falloff of
31 total chlorine and bromine abundances in the stratosphere in the next century will be much
March 2001
4A-4
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14 .
15
16
17
slower than the rate of increase observed in past decades, because of the slow rate at which
natural processes remove these compounds from the stratosphere. The most vulnerable period
for ozone depletion will be extended into the coming decades. However, extreme
perturbations, such as natural events like volcanic eruptions, could enhance the loss from
ozone-depleting chemicals. Detection of the beginning of the recovery of the ozone layer could
be achievable early in the next century if decreasing chlorine and bromine abundances were the
only factor. However, potential future increases or decreases in other gases important in ozone
chemistry (such as nitrous oxide, methane, and water vapor) and climate change will influence
the recovery of the ozone layer. When combined with the natural variability of the ozone layer,
these factors imply that unambiguous detection of the beginning of the recovery of the ozone
layer is expected to be well after the maximum stratospheric loading of ozone-depleting gases.
REFERENCE
World Meteorological Organization (WMO). (1999) Scientific assessment of ozone depletion: 1998. Geneva,
Switzerland: World Meteorological Organization, Global Ozone and Monitoring Project; report no. 44.
March 2001
4A-5
DRAFT-DO NOT QUOTE OR CITE
-------
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
APPENDIX 4B
Excerpted Key Points from the Executive Summary of the United Nations
Environmental Programme 1998 Assessment of Environmental Effects of
Ozone Depletion (United Nations Environmental Programme, 1998)
Decreased quantities of total-column ozone now are observed over large parts of the globe,
permitting increased penetration of solar UV-B radiation (280 to 315 nm) to the Earth's surface.
The present assessment deals with the possible consequences. The Atmospheric Science Panel
predicts that the ozone layer will be in its most vulnerable state during the coming two decades.
Some of the effects are expected to occur during most of the next century. Recent studies show
that the effects of ozone depletion would have been dramatically worse without protective
measures taken under the 1987 Montreal Protocol. The assessment is given in seven papers,
summarized below:
(1) Changes in Ultraviolet Radiation
• Stratospheric ozone levels are near their lowest points since measurements began, so
current UV-B radiation levels are thought to be close to their maximum. Total
stratospheric content of ozone-depleting substances is expected to reach a maximum before the
year 2000. All other things being equal, the current ozone losses and related UV-B increases
should be close to their maximum. Increases in surface erythemal (sunburning) UV radiation
relative to the values in the 1970s are estimated to be
• about 7% at Northern Hemisphere mid-latitudes in winter/spring;
• about 4% at Northern Hemisphere mid-latitudes in summer/fall;
• about 6% at Southern Hemisphere mid-latitudes on a year-round basis;
• about 130% in the Antarctic in the spring; and
• about 22% in the Arctic in the spring.
• The correlation between increases in surface UV-B radiation and decreases in overhead
ozone has been demonstrated further and quantified by ground-based instruments under
a wide range of conditions. Improved measurements of UV-B radiation are now providing
March 2001
4B-1
DRAFT-DO NOT QUOTE OR CITE
-------
1 better geographical and temporal coverage. Surface UV-B radiation levels are highly variable
2 because of sun angle, cloud cover, and, also, because of local effects, including pollutants and
3 surface reflections. With a few exceptions, the direct detection of UV-B trends at low- and
4 mid-latitudes remains problematic because of this high natural variability, the relatively small
5 ozone changes, and the practical difficulties of maintaining long-term stability in networks of
6 UV-measuring instruments. Few reliable UV-B radiation measurements are available from
7 pre-ozone-depletion days.
8 • Satellite-based observations of atmospheric ozone and clouds are being used, together
9 with models of atmospheric transmission, to provide global coverage and long-term
10 estimates of surface UV-B radiation. Estimates of long-term (1979 to 1992) trends in zonally
11 averaged UV irradiances that include cloud effects are nearly identical to those for clear-sky
12 estimates, providing evidence that clouds have not influenced the UV-B trends. However, the
13 limitations of satellite-derived UV estimates should be recognized. To assess uncertainties
14 inherent in this approach, additional validations involving comparisons with ground-based
15 observations are required.
16 • Direct comparisons of ground-based UV-B radiation measurements between a few
17 mid-latitude sites in the Northern and Southern Hemispheres have shown larger
18 differences than those estimated using satellite data. Ground-based measurements show that
19 summertime erythemal UV irradiances in the Southern Hemisphere exceed those at comparable
20 latitudes of the Northern Hemisphere by up to 40%, whereas corresponding satellite-based
21 estimates yield only 10 to 15% differences. Atmospheric pollution may be a factor in this
22 discrepancy between ground-based measurements and satellite-derived estimates. UV-B
23 measurements at more sites are required to determine whether the larger observed differences
24 are globally representative.
25 • High levels of UV-B radiation continue to be observed in Antarctica during the recurrent
26 spring-time ozone hole. For example, during ozone-hole episodes, measured biologically
27 damaging radiation at Palmer Station, Antarctica (64 °S) has been found to approach and
28 occasionally even exceed maximum summer values at San Diego, CA (32 °N).
29 • Long-term predictions of future UV-B levels are difficult and uncertain. Nevertheless,
30 current best estimates suggest that a slow recovery to pre-ozone-depletion levels may be
31 expected during the next half-century. Although the maximum ozone depletion, and hence
March 2001
4B-2
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
maximum UV-B increase, is likely to occur in the current decade, the ozone layer will continue
to be in its most vulnerable state into the next century. The peak depletion and the recovery
phase could be delayed by decades because of interactions with other long-term atmospheric
changes (e.g., increasing concentrations of greenhouse gases). Other factors that could influence
the recovery include nonratification or noncompliance with the Montreal Protocol and its
Amendments and Adjustments and future volcanic eruptions. The recovery phase for surface
UV-B irradiances probably will not be detectable until many years after the ozone minimum.
(2) Effects on Human and Animal Health
• Recent estimates suggest that the increase in the risk of cataract and skin cancer because
of ozone depletion would not have been controlled adequately by implementation of the
Montreal Protocol (1987) alone, but can be achieved through implementation of its later
provisions. Risk assessments for the United States and northwestern Europe indicate large
increases in cataracts and skin cancers under either the "no Protocol" or the early Montreal
Protocol scenarios. Under scenarios based on later amendments (Copenhagen, 1992) and
Montreal (1997), increases in cataracts and skin cancer attributable to ozone depletion return
almost to zero by the end of the next century.
• The increases in UV-B radiation associated with ozone depletion are likely to lead to
increases in the incidence or severity of a variety of short- and long-term health effects, if
current exposure practices are not modified by changes in behavior.
• Adverse effects on the eye will affect all populations irrespective of skin color. Adverse
impacts could include more cases of acute reactions such as "snowblindness", increases in
cataract incidence or severity (and thus the incidence of cataract-associated blindness), and
increases in the incidence (and mortality) from ocular melanoma and squamous cell carcinoma
of the eye.
• Effects on the immune system also will affect all populations but may be both adverse and
beneficial. Adverse effects include depressed resistance to certain tumors and infectious
diseases, potential impairment of vaccination responses, and possibly increased severity of
some autoimmune and allergic responses. Beneficial effects could include decreases in the
severity of certain immunologic disease conditions, such as psoriasis and nickel allergy.
March 2001
4B-3
DRAFT-DO NOT QUOTE OR CITE
-------
1 • Effects on the skin could include increases in photoaging and skin cancer with risk
2 increasing with fairness of skin. Increases in UV-B are likely to accelerate the rate of
3 photoaging, as well as increase the incidence (and associated mortality) of melanoma and
4 nonmelanoma skin cancer, basal cell carcinoma, and squamous cell carcinoma.
5 • Research is generating much new information being used to help reduce uncertainties
6 associated with current risk estimates. Evaluation of the impact of susceptibility genes is
7 helping to identify highly susceptible populations so that their special risk can be assessed.
8 Examination of the impacts of behavior changes, such as consuming diets that are high in
9 antioxidants, avoiding sun exposure during the 4 h around solar noon, and wearing of covering
10 apparel (e.g., hats, sunglasses), is beginning to identify important exposure patterns, as well as
11 possible mitigation strategies.
12 • Quantitative risk assessments for a variety of other effects, such as UV-B-induced
13 immunosuppression of infectious diseases, are not yet possible. New information continues
14 to confirm the reasonableness of these concerns, but data that is adequate for quantitative risk
15 assessment are not yet available.
16
17 (3) Effects on Terrestrial Ecosystems
18 • Increased UV-B can be damaging for terrestrial organisms including plants and
19 microbes, but all these organisms also have protective and repair processes. The balance
20 between damage and protection varies among species and even varieties of crop species; many
21 species and varieties can accommodate increased UV-B. Tolerance of elevated UV-B by some
22 species and crop varieties provides opportunities for genetic engineering and breeding to deal
23 with potential crop-yield reductions because of elevated UV-B in agricultural systems.
24 • Research in the past few years indicates that increased UV-B exerts effects more often
25 through altered patterns of gene activity rather than damage. These UV-B effects on
26 regulation manifest themselves in many ways including changes in life-cycle timing, changes hi
27 plant form, and production of plant chemicals not directly involved in primary metabolism.
28 These plant chemicals play a role in protecting plants from pathogens and insect attack and
29 affect food quality for humans and grauine animals.
30 • Terrestrial ecosystem responses to increased UV-B are evident primarily in interactions
31 among species, rather than in the performance of individual species. Much of the recent
March 2001
4B-4
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
experimentation indicates that increased UV-B affects the balance of competition among higher
plants, the degree to which higher plants are consumed by insects, and susceptibility of plants to
pathogens. These effects can be mediated, in large part, by changes in plant form and
chemistry, but effects of UV-B on insects and microbes are also possible. The direction of
these UV-B mediated interactions among species is often difficult to predict based only on
single-organism responses to increased UV-B.
• Effects of increased UV-B radiation can accumulate from year to year in long-lived
perennial plants and from generation to generation in annual plants. This effect has been
shown in a few recent studies, but the generality of this accumulation among species is not
presently known. If this phenomenon is widespread, this would amplify otherwise subtle
responses to UV-B seen in a single growing season, for example, in forest trees.
• Effects of increased UV-B must be taken into account together with other environmental
factors including those associated with global change. Responses of plants and other
organisms to increased UV-B are modified by other environmental factors (e.g., CO, water
stress, mineral nutrient availability, heavy metals, temperature). Many of these factors also are
changing as the global climate is altered.
(4) Effects on Aquatic Ecosystems
• Recent studies continue to demonstrate that solar UV-B and UV-A have adverse effects on
the growth, photosynthesis, protein and pigment content, and reproduction of
phytoplankton, thus affecting the food web. These studies have determined biological
weighting functions and exposure-response curves for phytoplankton and have developed new
models for the estimation of UV-related photoinhibition. In spite of this increased
understanding and enhanced ability to model aquatic impacts, considerable uncertainty remains
with respect to quantifying effects of ozone-related UV-B increases at the ecosystem level.
• Macroalgae and sea grasses show a pronounced sensitivity to solar UV-B. They are
important biomass producers in aquatic ecosystems. Most of these organisms are attached and
so cannot avoid being exposed to solar radiation at their growth site. Effects have been found
throughout the top 10 to 15 m of the water column.
• Zooplankton communities, as well as other aquatic organisms including sea urchins,
corals, and amphibians, are sensitive to UV-B. There is evidence that, for some of these
March 2001
4B-5
DRAFT-DO NOT QUOTE OR CITE
-------
1 populations, even current levels of solar UV-B radiation, acting in conjunction with other
2 environmental stresses, may be a limiting factor, but quantitative evaluation of possible effects
3 remains uncertain.
4 • UV-B radiation is absorbed by and breaks down dissolved organic carbon (DOC) and
5 participate organic carbon (POC) and makes the products available for bacterial
6 degradation and remineralization. The degradation products are of importance in the cycling
7 of carbon in aquatic ecosystems. Because UV-B breaks down DOC as it is absorbed, increase
8 in UV-B can increase the penetration of both UV-B and UV-A radiation into the water column.
9 As a consequence, the quantity of UV-B penetrating to a given depth both influences and is
10 influenced by DOC. Warming and acidification result in faster degradation of these substances
11 and, thus, enhance the penetration of UV radiation into the water column.
12 • Polar marine ecosystems, where ozone-related UV-B increases are the greatest, are
13 expected to be the oceanic ecosystems most influenced by ozone depletion. Oceanic
14 ecosystems are characterized by large spatial and temporal variabilities that make it difficult to
15 select out UV-B-specific effects on single species or whole phytoplankton communities.
16 Although estimates of reduction in both Arctic and Antarctic productivity are based on
17 measurable short-term effects, there remain considerable uncertainties in estimating long-term
18 consequences, including possible shifts in community structure. Reduced productivity of fish
19 and other marine crops could have an economic impact, as well as affect natural predators;
20 however, quantitative estimation of the possible effects of reduced production remain
21 controversial.
22 • Potential consequences of enhanced levels of exposure of aquatic ecosystems to UV-B
23 radiation include reduced uptake capacity for atmospheric carbon dioxide (CO2),
24 resulting in the potential augmentation of global warming. The oceans play a key role with
25 respect to the budget of greenhouse gases. Marine phytoplankton are a major sink for
26 atmospheric CO2 and they have a decisive role in the development of future trends of CO2
27 concentrations in the atmosphere. The relative importance of the net uptake of CO2 by the
28 biological pump and the possible role of increased UV-B in the ocean are still controversial.
29
30
March 2001
4B-6
DRAFT-DO NOT QUOTE OR CITE
-------
1 (5) Effects on Biogeochemical Cycles
2 • Effects of increased UV-B on emissions of CO2 and carbon monoxide (CO) and on
3 mineral nutrient cycling in the terrestrial biosphere have been confirmed by recent
4 studies of a range of species and ecosystems. The effects, both in magnitude and direction, of
5 UV-B on trace gas emissions and mineral nutrient cycling are species specific and operate on a
6 number of processes. These processes include changes in the chemical composition in living
7 plant tissue; photodegradation (breakdown by light) of dead plant matter, including litter;
8 release of CO from vegetation previously charred by fire; changes in the communities of
9 microbial decomposers; and effects on nitrogen-fixing micro-organisms and plants. Long-term
10 experiments are in place to examine UV-B effects on carbon capture and storage in biomass
11 within natural terrestrial ecosystems.
12 • Studies in natural aquatic ecosystems have indicated that organic matter is the primary
13 regulator of UV-B penetration. Enhanced UV-B can affect the balance between the
14 biological processes that produce the organic matter and the chemical and microbial processes
15 that degrade it. Changes in the balance have broad impacts on the effects of enhanced UV-B on
16 biogeochemical cycles. These changes, which are reinforced by changes in climate and
17 acidification, result from clarification of the water and changes in light quality.
18 • Increased UV-B has positive and negative impacts on microbial activity in aquatic
19 ecosystems that can affect carbon and mineral nutrient cycling, as well as the uptake and
20 release of greenhouse and chemically reactive gases. Photoinhibition of surface aquatic
21 micro-organisms by UV- B can be offset partially by photodegradation of dissolved organic
22 matter to produce substrates, such as organic acids and ammonium, that stimulate microbial
23 activity.
24 • Modeling and experimental approaches are being developed to predict and measure the
25 interactions and feedbacks between climate change in UV-B-induced changes in marine
26 and terrestrial biogeochemical cycles. These interactions include alterations in the oxidative
27 environment in the upper ocean and in the marine boundary layer and oceanic production and
28 release of CO, volatile organic compounds (VOC), and reactive oxygen species (ROS, such as
29 hydrogen peroxide and hydroxyl radicals). Climate-related changes in temperature and water
30 supply in terrestrial ecosystems interact with UV-B radiation through biogeochemical processes
31 operating on a wide range of time scales.
March 2001
4B-7
DRAFT-DO NOT QUOTE OR CITE
-------
1 (6) Effects on Air Quality
2 • Increased UV-B will increase the chemical activity in the lower atmosphere (the
3 troposphere). Troposphere ozone levels are sensitive to local concentrations of nitrogen
4 oxides (NOx) and hydrocarbons. Model studies suggest that additional UV-B radiation reduces
5 tropospheric ozone in clean environments (low NOX) and increases tropospheric ozone in
6 polluted areas (high NOJ.
7 • Assuming other factors remain constant, additional UV-B will increase the rate at which
8 primary pollutants are removed from the troposphere. Increased UV-B is expected to
9 increase the concentration of hydroxyl radicals (OH) and result in faster removal of pollutants.
10 Increased concentrations of oxidants such as hydrogen peroxide and organic peroxides also are
11 expected. The effects of UV-B increases on tropospheric ozone, OH, methane, CO, and
12 possibly other tropospheric constituents, although not negligible, will be difficult to detect
13 because the concentrations of these species also are influenced by many other variable factors
14 (e.g., emissions).
15 -No significant effects on humans or the environment have been identified from
16 trifluoroacetic acid (TFA) produced by atmospheric degradation of HCFCs and MFCs.
17 Numerous studies have shown that TFA has, at most, moderate short-term toxicity. Insufficient
18 information is available to assess potential chronic, developmental, or reproductive effects. The
19 atmospheric degradation mechanisms of most substitutes for ozone-depleting substances are
20 well established. HCFCs and HFCs are two important classes of substitutes. Atmospheric
21 degradation of HCFC-123 (CF3CHC12), HCFC-124 (CF3CHFCI), and H FC-I34a (CF3CH2F)
22 produces TFA. Reported measurements of TFA in rain, rivers, lakes, and oceans show it to be
23 an ubiquitous component of the hydrosphere, present at levels much higher than can be
24 explained by currently reported sources. The levels of TFA currently produced by the
25 atmospheric degradation of HFCs and HCFCs are estimated to be orders of magnitude below
26 those of concern and make only a minor contribution to the current environmental burden of
27 TFA.
28
29 (7) Effects on Materials
30 • Physical and mechanical properties of polymers are affected negatively by increased
31 UV-B in sunlight. Increased UV-B reduces the useful lifetimes of synthetic polymer products
March 2001
4B-8
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
used outdoors and of biopolymer materials such as wood, paper, wool, and cotton. The
reduction in service life of materials depends on the synergistic effect of increased UV-B and
other factors, especially the temperature of the material during exposure to sunlight. Even
under harsh UV exposure conditions, the higher temperatures largely determine the extent of
increased UV-induced damage to photostabilized polyethylenes. However, accurate assessment
of such damage to various materials is presently difficult to make because of limited availability
of technical data, especially on the relationship between the dose of UV-B radiation and the
resulting damage of the polymer or other material.
• Conventional photostabilizers are likely to be able to mitigate the effects of increased UV
levels in sunlight. More effective photostabilizers for plastics have been commercialized in
recent years. The use of these compounds allows plastic polymer products to be used in a wide
range of different UV environments found worldwide. It is reasonable to expect existing
photostabilizer technologies to be able to mitigate these effects of an increased UV-B on
polymer materials. This, however, would increase the cost of the relevant polymer products,
surface coatings, and treated biopolymer materials. However, the efficiencies of even the
conventional photostabilizers under the unique exposure environments resulting from an
increase in solar UV-B have not been well studied.
REFERENCE
United Nations Environment Programme (UNEP). (1998) Environmental effects of ozone depletion: 1998
assessment. J. Photochem. Photobiol. B 46: 1-4.
March 2001
4B-9
DRAFT-DO NOT QUOTE OR CITE
-------
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
APPENDIX 4C
Excerpted Key Points from the Executive Summary of the Special Report
of the International Panel on Climate Change Working Group II on the
Regional Impacts of Climate Change: An Assessment of Vulnerability
Excerpts from Executive Summary materials from a Special Report of the IPCC Working
Group II, The Regional Impacts of Climate Change: An Assessment of Vulnerability (IPCC,
1998) are incorporated below in this appendix to provide an overview of key points regarding the
vulnerability of aquatic and terrestrial ecosystems, water resources, agriculture, and human
habitability in North America to climate change.
Scope of the Assessment
The report was prepared at the request of the Conference of the Parties to the United
Nations Framework Convention on Climate Change (UNFCCC) and its subsidiary bodies
(specifically, the Subsidiary Body for Scientific and Technological Advice-SBSTA). The special
report provides, on a regional basis, a review of state-of-the-art information on the vulnerability
to potential changes in climate of ecological systems, socioeconomic sectors (agriculture,
fisheries, water resources, and human settlements), and human health. The report reviews the
sensitivity of these systems as well as options for adaptation. Though the report draws heavily on
the sectoral impact assessments of the Second Assessment Report (SAR), it also draws on more
recent peer-reviewed literature (inter alia, country studies programs).
Nature of the Issue
Human activities (primarily burning of fossil fuels and changes in land use and land cover)
are increasing atmospheric concentrations of greenhouse gases, which alter radiative balances
and tend to warm the atmosphere, and, in some regions, aerosols, which have an opposite effect
on radiative balances and tend to cool the atmosphere. At present, in some locations primarily in
the Northern Hemisphere, the cooling effects of aerosols can be large enough to more than offset
the warming caused by greenhouse gases. Because aerosols do not remain in the atmosphere for
March 2001 4C-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 long periods and global emissions of their precursors are not projected to increase substantially,
2 aerosols will not offset the global long-term effects of greenhouse gases, which are long-lived.
3 Aerosols can have important consequences for continental-scale patterns of climate change.
4 These changes in greenhouse gases and aerosols, taken together, are projected to lead to
5 regional and global changes in temperature, precipitation, and other climate variables, resulting
6 in global changes in soil moisture; an increase in global mean sea level; and prospects for more
7 severe extreme high-temperature events, floods, and droughts in some places. Based on the
8 range of sensitivities of climate to changes in the atmospheric concentrations of greenhouse gases
9 (IPCC, 1996; WG I) and plausible changes in emissions of greenhouse gases and aerosols
10 (lS92a-f, scenarios that assume no climate policies), climate models project that the mean annual
11 global surface temperature will increase by 1 to 3.5 °C by 2100, that global mean sea level will
12 rise by 15 to 95 cm, and that changes in the spatial and temporal patterns of precipitation will
13 occur. The average rate of warming probably would be greater than any seen in the past 10,000
14 years, although the actual annual to decadal rate would include considerable natural variability,
15 and regional changes could differ substantially from the global mean value. These long-term,
16 large-scale, human-induced changes will interact with natural variability on time scales of days to
17 decades (e.g., the El Nino-Southern Oscillation [ENSO] phenomenon) and, thus, influence social
18 and economic well-being. Possible local climate effects caused by unexpected events such as a
19 climate-change-induced change of flow pattern of marine water streams (e.g., the Gulf Stream)
20 have not been considered, because such changes cannot be predicted with confidence at present.
21 Scientific studies show that human health, ecological systems, and socioeconomic sectors
22 (e.g., hydrology and water resources, food and fiber production, coastal systems, human
23 settlements), all of which are vital to sustainable development, are sensitive to changes in
24 climate, including both the magnitude and rate of climate change, as well as to changes in
25 climate variability. Whereas many regions are likely to experience adverse effects of climate
26 change, some of which are potentially irreversible, some effects of climate change are likely to be
27 beneficial. Climate change represents an important additional stress on those systems already
28 affected by increasing resource demands, unsustainable management practices, and pollution,
29 which in many cases may be equal to or greater than those of climate change. These stresses will
30 interact in different ways across regions but can be expected to reduce the ability of some
31 environmental systems to provide, on a sustained basis, key goods and services needed for
March 2001
4C-2
DRAFT-DO NOT QUOTE OR CITE
-------
1 successful economic and social development, including adequate food, clean air, and water;
2 energy; safe shelter; low levels of disease; and employment opportunities. Climate change also
3 will take place in the context of economic development, which may make some groups or
4 countries less vulnerable to climate change, for example, by increasing the resources available for
5 adaptation. Those that experience low rates of growth, rapid increases in population, and
6 ecological degradation may become increasingly vulnerable to potential changes.
7
8 Approach of the Assessment
9 The report assesses the vulnerability of natural and social systems of major regions of the
10 world to climate change. Vulnerability is defined as the extent to which a natural or social
11 system is susceptible to sustaining damage from climate change. Vulnerability is a function of
12 the sensitivity of a system to changes in climate (the degree to which a system will respond to a
13 given change in climate, including both beneficial and harmful effects) and the ability to adapt
14 the system to changes in climate (the degree to which adjustments in practices, processes, or
15 structures can moderate or offset the potential for damage or take advantage of opportunities
16 created because of a given change in climate). Under this framework, a highly vulnerable system
17 would be one that is highly sensitive to modest changes hi climate, where the sensitivity includes
18 the potential for substantial harmful effects, and one for which the ability to adapt is severely
19 constrained.
20 Because the available studies have not employed a common set of climate scenarios and
21 methods, and because of uncertainties regarding the sensitivities and adaptability of natural and
22 social systems, the assessment of regional vulnerabilities is necessarily qualitative. However, the
23 report provides substantial and indispensable information on what currently is known about
24 vulnerability to climate change.
25 In a number of instances, quantitative estimates of impacts of climate change are cited in
26 the report. Such estimates are dependent on the specific assumptions employed regarding future
27 changes in climate, as well as on the particular methods and models applied hi the analyses.
28 In interpreting these estimates, it is important to bear hi mind that uncertainties regarding the
29 character, magnitude, and rates of future climate change remain. These uncertainties impose
30 limitations on the ability of scientists to project impacts of climate change, particularly at
31 regional and smaller scales.
March 2001
4C-3
DRAFT-DO NOT QUOTE OR CITE
-------
1 It is in part because of the uncertainties regarding how climate will change that the report
2 takes the approach of assessing vulnerabilities rather than assessing quantitatively the expected
3 impacts of climate change. The estimates are interpreted best as illustrative of the potential
4 character and approximate magnitudes of impacts that may result from specific scenarios of
5 climate change. They serve as indicators of sensitivities and possible vulnerabilities. Most
6 commonly, the estimates are based on changes in equilibrium climate that have been simulated to
7 result from an equivalent doubling of carbon dioxide (CO2) in the atmosphere. Usually, the
8 simulations have excluded the effects of aerosols. Increases in global mean temperatures
9 corresponding to these scenarios mostly fall in the range of 2 to 5 °C. To provide a temporal
10 context for these scenarios, the range of projected global mean warming by 2100 is 1 to 3.5 °C,
11 accompanied by a mean sea-level rise of 15 to 95 cm, according to the IPCC Second Assessment
12 Report. General circulation model (GCM) results are used in this analysis to justify the order of
13 magnitude of the changes used in the sensitivity analyses. They are not predictions that climate
14 will change by specific magnitudes in particular countries or regions. The amount of literature
15 available for assessment varies in quantity and quality among the regions.
16
17 Overview of Regional Vulnerabilities to Global Climate Change
18 The report's assessment of regional vulnerability to climate change focuses on ecosystems,
19 hydrology and water resources, food and fiber production, coastal systems, human settlements,
20 human health, and other sectors or systems (including the climate system) important to
21 10 regions that encompass the Earth's land surface. Wide variation in the vulnerability of similar
22 sectors or systems is to be expected across regions, as a consequence of regional differences in
23 local environmental conditions; preexisting stresses to ecosystems; current resource-use patterns;
24 and the framework of factors affecting decision making, including government policies, prices,
25 preferences, and values. Nonetheless, some general observations, based on information
26 contained in the IPCC Second Assessment Report (SAR) (IPCC, 1995) and synthesized from the
27 regional analyses in the 1998 assessment, provide a global context for assessment of each
28 region's vulnerability. The general types of vulnerabilities are discussed first below, followed by
29 more specific discussion of projected likely regional vulnerabilities most directly applicable to
30 the United States.
31
March 2001
4C-4
DRAFT-DO NOT QUOTE OR CITE
-------
1 Ecosystems
2 Ecosystems are of fundamental importance to environmental function and to sustainability,
3 and they provide many goods and services critical to individuals and societies. These goods and
4 services include (1) providing food, fiber, fodder, shelter, medicines, and energy; (2) processing
5 and storing carbon and nutrients; (3) assimilating wastes; (4) purifying water, regulating water
6 runoff, and moderating floods; (5) building soils and reducing soil degradation; (6) providing
7 opportunities for recreation and tourism; and (7) housing the Earth's entire reservoir of genetic
8 and species diversity. In addition, natural ecosystems have cultural, religious, aesthetic, and
9 intrinsic existence values. Changes in climate have the potential to affect the geographic location
10 of ecological systems, the mix of species that they contain, and their ability to provide the wide
11 range of benefits on which societies rely for their continued existence. Ecological systems are
12 intrinsically dynamic and are constantly influenced by climate variability. The primary influence
13 of anthropogenic climate change on ecosystems is expected to be through the rate and magnitude
14 of change in climate means and extremes; climate change is expected to occur at a rapid rate
15 relative to the speed at which ecosystems can adapt and reestablish themselves; and through the
16 direct effects of increased atmospheric CO2 concentrations, which may increase the productivity
17 and efficiency of water use in some plant species. Secondary effects of climate change involve
18 changes in soil characteristics and disturbance regimes (e.g., fires, pests, diseases), which would
19 favor some species over others and thus change the species composition of ecosystems.
20 Based on model simulations of vegetation distribution, which use GCM-based climate
21 scenarios, large shifts of vegetation boundaries into higher latitudes and elevations can be
22 expected. The mix of species within a given vegetation class likely will change. Under
23 equilibrium GCM climate scenarios, large regions show drought-induced declines in vegetation,
24 even when the direct effects of CO2 fertilization are included. By comparison, under transient
25 climate scenarios, in which trace gases increase slowly over a period of years, the full effects of
26 changes in temperature and precipitation lag the effects of a change in atmospheric composition
27 by a number of decades; hence, the positive effects of CO2, precede the full effects of changes in
28 climate.
29 Climate change is projected to occur at a rapid rate relative to the speed at which forest
30 species grow, reproduce, and reestablish themselves (past tree species' migration rates are
31 believed to be on the order of 4 to 200 km per century). For mid-latitude regions, an average
March 2001
4C-5
DRAFT-DO NOT QUOTE OR CITE
-------
1 warming of 1 to 3.5 °C over the next 100 years would be equivalent to a poleward shift of the
2 present geographic bands of similar temperatures (or "isotherms") by approximately 150 to
3 550 km, or an altitude shift of about a 150 to 550 m. Therefore, the species composition of
4 forests is likely to change; in some regions, entire forest types may disappear, and new
5 assemblages of species and, hence, new ecosystems may be established. As a consequence of
6 possible changes in temperature and water availability under doubled equivalent-CO2 equilibrium
7 conditions, a substantial fraction (a global average of one-third, varying by region from
8 one-seventh to two-thirds) of the existing forested area of the world likely would undergo major
9 changes hi broad vegetation types, with the greatest changes occurring in high latitudes and the
10 least in the tropics. In tropical rangelands, major alterations in productivity and species
11 composition would occur because of altered rainfall amount and seasonally and increased
12 evapotranspiration, although a mean temperature increase alone would not lead to such changes.
13 Inland aquatic ecosystems will be influenced by climate change through altered water
14 temperatures, flow regimes, water levels, and thawing of permafrost at high latitudes. In lakes
15 and streams, warming would have the greatest biological effects at high latitudes, where
16 biological productivity would increase and lead to expansion of cool-water species' ranges, and
17 at the low-latitude boundaries of cold- and cool-water species ranges, where extinctions would be
18 greatest. Increases in flow variability, particularly the frequency and duration of large floods and
19 droughts, would tend to reduce water quality, biological productivity, and habitat in streams. The
20 geographical distribution of wetlands is likely to shift with changes in temperature and
21 precipitation, with uncertain implications for net greenhouse gas emissions from nontidal
22 wetlands. Some coastal ecosystems (saltwater marshes, mangrove ecosystems, coastal wetlands,
23 coral reefs, coral atolls, and river deltas) are particularly at risk from climate change and other
24 stresses. Changes in these ecosystems would have major negative effects on freshwater supplies,
25 fisheries, biodiversity, and tourism.
26 Adaptation options for ecosystems are limited, and their effectiveness is uncertain. Options
27 include establishment of corridors to assist the "migration" of ecosystems, land-use management,
28 plantings, and restoration of degraded areas. Because of the projected rapid rate of change
29 relative to the rate at which species can reestablish themselves, the isolation and fragmentation
30 of many ecosystems, the existence of multiple stresses (e.g., land-use change, pollution), and
March 2001
4C-6
DRAFT-DO NOT QUOTE OR CITE
-------
1 limited adaptation options, ecosystems (especially forested systems, montane systems, and coral
2 reefs) are vulnerable to climate change.
3
4 Hydrology and Water Resources
5 Water availability is an essential component of welfare and productivity. Currently,
6 1.3 billion people do not have access to adequate supplies of safe water, and 2 billion people do
7 not have access to adequate sanitation. Although these people are dispersed throughout the
8 globe, reflecting subnational variations in water availability and quality, some 19 countries
9 (primarily in the Middle East and northern and southern Africa) face such severe shortfalls that
10 they are classified as either water-scarce or water-stressed; this number is expected to roughly
11 double by 2025, in large part because of increases in demand resulting from economic and
12 population growth. For example, most policy makers now recognize drought as a recurrent
13 feature of Africa's climate. However, climate change will further exacerbate the frequency and
14 magnitude of droughts in some places.
15 Changes in climate could exacerbate periodic and chronic shortfalls of water, particularly in
16 arid and semi-arid areas of the world. Developing countries are highly vulnerable to climate
17 change because many are located in arid and semi-arid regions, and most derive their water
18 resources from single-point systems such as bore holes or isolated reservoirs. These systems, by
19 their nature, are vulnerable because there is no redundancy in the system to provide resources,
20 should the primary supply fail. Also, given the limited technical, financial, and management
21 resources possessed by developing countries, adjusting to shortages or implementing adaptation
22 measures will impose a heavy burden on their national economies. There is evidence that
23 flooding is likely to become a larger problem in many temperate and humid regions, requiring
24 adaptations not only to droughts and chronic water shortages but also to floods and associated
25 damages, raising concerns about dam and levee failures.
26 The impacts of climate change will depend on the baseline condition of the water supply
27 system and the ability of water resources managers to respond not only to climate change but also
28 to population growth and changes in demands; technology; and economic, social, and legislative
29 conditions.
30 Various approaches are available to reduce the potential vulnerability of water systems to
31 climate change. Options include pricing systems, water efficiency initiatives, engineering and
March 2001
4C-7
DRAFT-DO NOT QUOTE OR CITE
-------
1 structural improvements to water supply infrastructure, agriculture policies, and urban planning
2 and management. At the national/regional level, priorities include placing greater emphasis on
3 integrated, cross-sectoral water resources management, using river basins as resource
4 management units, and encouraging sound pricing and management practices. Given increasing
5 demands, the prevalence and sensitivity of many simple water management systems to
6 fluctuations in precipitation and runoff, and the considerable time and expense required to
7 implement many adaptation measures, the water resources sector in many regions and countries
8 is vulnerable to potential changes in climate.
9
10 Food and Fiber Production
11 Currently, 800 million people are malnourished; as the world's population increases and
12 incomes in some countries rise, food consumption is expected to double over the next three to
13 four decades. The most recent doubling in food production occurred over a 25-year period and
14 was based on irrigation, chemical inputs, and high-yielding crop varieties. Whether the
15 remarkable gains of the past 25 years will be repeated is uncertain. Problems associated with
16 intensifying production on land already in use (e.g., chemical and biological runoff, waterlogging
17 and salinization of soils, soil erosion and compaction) are becoming increasingly evident.
18 Expanding the amount of land under cultivation (including reducing land deliberately taken out
19 of production to reduce agricultural output) also is an option for increasing total crop production,
20 but it could lead to increases in competition for land and pressure on natural ecosystems,
21 increased agricultural emissions of greenhouse gases, a reduction in natural sinks of carbon, and
22 expansion of agriculture to marginal lands, all of which could undermine the ability to
23 sustainably support increased agricultural production.
24 Changes in climate will interact with stresses that result from actions to increase
25 agricultural production, affecting crop yields and productivity in different ways, depending on the
26 types of agricultural practices and systems in place. The main direct effects will be through
27 changes in factors such as temperature, precipitation, length of growing season, and timing of
28 extreme or critical threshold events relative to crop development, as well as through changes in
29 atmospheric CO2 concentration (which may have a beneficial effect on the growth of many crop
30 types): Indirect effects will include potentially detrimental changes in diseases, pests, and weeds,
31 the effects of which have not yet been quantified in most available studies. Evidence continues
March 2001
4C-8
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
to support the findings of the IPCC SAR that "global agricultural production could be maintained
relative to baseline production" for a growing population under 2xCO2 equilibrium climate
conditions. In addition, the regional findings of this special report lend support to concerns over
the "potential serious consequences" of increased risk of hunger in some regions, particularly the
tropics and subtropics. Generally, middle to high latitudes may experience increases in
productivity, depending on crop type, growing season, changes in temperature regimes, and the
seasonality of precipitation. In the tropics and subtropics, where some crops are near their
maximum temperature tolerance and where dry-land, nonirrigated agriculture predominates,
yields are likely to decrease. The livelihoods of subsistence farmers and pastoral peoples, who
make up a large portion of rural populations in some regions, also could be affected negatively.
In regions where there is a likelihood of decreased rainfall, agriculture could be significantly
affected.
Fisheries and fish production are sensitive to changes in climate and currently are at risk
from overfishing, diminishing nursery areas, and extensive inshore and coastal pollution.
Globally, marine fisheries production is expected to remain about the same in response to
changes in climate; high-latitude freshwater and aquaculture production is likely to increase,
assuming that natural climate variability and the structure and strength of ocean currents remain
about the same. The principal impacts will be felt at the national and local levels, as centers of
production shift. The positive effects of climate change, such as longer growing seasons, lower
natural winter mortality, and faster growth rates in higher latitudes, may be offset by negative
factors such as changes in established reproductive patterns, migration routes, and ecosystem
relationships.
Given the many forces bringing profound change to the agricultural sector, adaptation
options that enhance resilience to current natural climate variability and potential changes in
means and extremes and address other concerns (e.g., soil erosion, salinization) offer no- or
low-regret options. For example, linking agricultural management to seasonal climate
predictions can assist in incremental adaptation, particularly in regions where climate is strongly
affected by ENSO conditions. The suitability of these options for different regions varies, in part
because of differences in the financial and institutional ability of the private sector and
governments in different regions to implement them. Adaptation options include changes in
crops and crop varieties; development of new crop varieties; changes in planting schedules and
March 2001
4C-9
DRAFT-DO NOT QUOTE OR CITE
-------
1 tillage practices; introduction of new biotechnologies; and improved water-management and
2 irrigation systems, which have high capital costs and are limited by availability of water
3 resources. Other options, such as minimum- and reduced-tillage technologies, do not require
4 such extensive capitalization but do require high levels of agricultural training and support.
5 In regions where agriculture is well adapted to current climate variability or where market
6 and institutional factors are in place to redistribute agricultural surpluses to make up for
7 shortfalls, vulnerability to changes in climate means and extremes generally is low. However, in
8 regions where agriculture is unable to cope with existing extremes, where markets and
9 institutions to facilitate redistribution of deficits and surpluses are not in place, or where
10 adaptation resources are limited, the vulnerability of the agricultural sector to climate change
11 should be considered high. Other factors also will influence the vulnerability of agricultural
12 production in a particular country or region to climate change, including the extent to which
13 current temperatures or precipitation patterns are close to or exceed tolerance limits for important
14 crops, per capita income, the percentage of economic activity based on agricultural production,
15 and the preexisting condition of the agricultural land base.
16
17 Coastal Systems
1 g Coastal zones are characterized by a rich diversity of ecosystems and a great number of
19 socioeconomic activities. Coastal human populations in many countries have been growing at
20 double the national rate of population growth. Currently, it is estimated that about half of the
21 global population lives in coastal zones, although there is large variation among countries.
22 Changes in climate will affect coastal systems through sea-level rise and an increase in
23 storm-surge hazards and possible changes in the frequency or intensity of extreme events.
24 Coasts in many countries currently face severe sea-level rise problems as a consequence of
25 tectonically and anthropogenically induced subsidence. An estimated 46 million people per year
26 currently are at risk of flooding from storm surges. Climate change will exacerbate these
27 problems, leading to potential impacts on ecosystems and human coastal infrastructure. Large
28 numbers of people also potentially are affected by sea-level rise, for example, tens of millions of
29 people in Bangladesh would be displaced by a 1 -m increase (the top of the range of IPCC
30 Working Group I estimates for 2100) in the absence of adaptation measures. A growing number
31 of extremely large cities are located in coastal areas, which means that large amounts of
March 2001
4C-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 infrastructure may be affected. Although annual protection costs for many nations are relatively
2 modest, about 0.1% of gross domestic product (GDP), the average annual costs to many small
3 island states total several percent of GDP. For some island nations, the high cost of providing
4 storm-surge protection would make it essentially infeasible, especially given the limited
5 availability of capital for investment.
6 Beaches, dunes, estuaries, and coastal wetlands adapt naturally and dynamically to changes
7 in prevailing winds and seas, as well as sea-level changes; in areas where infrastructure
8 development is not extensive, planned retreat and accommodation to changes may be possible.
9 It also may be possible to rebuild or relocate capital assets at the end of their design life. In other
10 areas, however, accommodation and planned retreat are not viable options, and protection using
11 hard structures (e.g., dikes, levees, floodwalls, barriers) and soft structures (e.g., beach
12 nourishment, dune restoration, wetland creation) will be necessary. Factors that limit the
13 implementation of these options include inadequate financial resources, limited institutional and
14 technological capability, and shortages of trained personnel. In most regions, current coastal
15 management and planning frameworks do not take account of the vulnerability of key systems to
16 changes in climate and sea level or long lead times for implementation of many adaptation
17 measures. Inappropriate policies encourage development in impact-prone areas. Given
18 increasing population density in coastal zones; long lead times for implementation of many
19 adaptation measures; and institutional, financial, and technological limitations (particularly in
20 many developing countries), coastal systems should be considered vulnerable to changes in
21 climate.
22
23 Human Health
24 In much of the world, life expectancy is increasing; in addition, infant and child mortality
25 in most developing countries is droping. Against this positive backdrop, however, there appears
26 to be a widespread increase in new and resurgent vectorborne and infectious diseases, such as
27 dengue, malaria, hantavirus, and cholera. In addition, the percentage of the developing world's
s:
28 population living in cities is expected to increase from 25% (in 1960) to more than 50% by 2020,
29 with percentages in some regions far exceeding these averages. These changes will bring
30 benefits only if accompanied by increased access to services such as sanitation and potable water
31 supplies; they also can lead to serious urban environmental problems, including air pollution
March 2001
4C-11
DRAFT-DO NOT QUOTE OR CITE
-------
1 (e-g-j participates, surface ozone, lead), poor sanitation, and associated problems in water quality
2 and potability, if access to services is not improved.
3 Climate change could affect human health through increases in heat-stress mortality,
4 tropical vector-borne diseases, urban air pollution problems, and decreases in cold-related
5 illnesses. Compared with the total burden of ill health, these problems are not likely to be large.
6 In the aggregate, however, the direct and indirect impacts of climate change on human health do
7 constitute a hazard to human population health, especially in developing countries in the tropics
8 and subtropics; these impacts have considerable potential to cause significant loss of life, affect
9 communities, and increase health-care costs and lost work days. Model projections (which entail
10 necessary simplifying assumptions) indicate that the geographical zone of potential malaria
11 transmission would expand in response to global mean temperature increases at the upper part of
12 the IPCC-projected range (3 to 5 °C by 2100), increasing the affected proportion of the world's
13 population from approximately 45% to approximately 60% by the latter half of the next century.
14 Areas where malaria is currently endemic could experience intensified transmission (on the order
15 of 50 to 80 million additional annual cases, relative to an estimated global background total of
16 500 million cases). Some increases in non-vector-borne infectious diseases, such as
17 salmonellosis, cholera, and giardiasis, also could occur as a result of elevated temperatures and
18 increased flooding. However, quantifying the projected health impacts is difficult because the
19 extent of climate induced health disorders depends on other factors, such as migration, provision
20 of clean urban environments, improved nutrition, increased availability of potable water,
21 improvements in sanitation, the extent of disease vector-control measures, changes in resistance
22- of vector organisms to insecticides, and more widespread availability of health care. Human
23 health is vulnerable to changes in climate, particularly in urban areas, where access to space
24 conditioning may be limited, as well as in areas where exposure to vector-borne and
25 communicable diseases may increase and health-care delivery and basic services, such as
26 sanitation, are poor.
27
28 Regional Vulnerability to Global Climate Change
29 Discussions about two geographic regions (North American and Polar regions) assessed in
30 the report are included here because of their relevance to the continental United States and
31 Alaska, respectively.
March 2001
4C-12
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
North American Region
This region consists of Canada and the United States south of the Arctic Circle. Within the
region, vulnerability to and the impacts of climate change vary significantly from sector to sector
and from subregion to subregion. This "texture" is important in understanding the potential
effects of climate change on North America, as well as in formulating and implementing viable
response strategies.
Ecosystems. Most ecosystems are moderately to highly sensitive to changes in climate.
Effects are likely to include both beneficial and harmful changes. Potential impacts include
northward shifts of forest and other vegetation types, which would affect biodiversity by altering
habitats and would reduce the market and nonmarket goods and services they provide; declines in
forest density and forested area in some subregions, but gains in others; more frequent and larger
forest fires; expansion of arid land species into the great basin region; drying of prairie pothole
wetlands that currently support over 50% of all waterfowl in North America; and changes in
distribution of habitat for cold-, cool-, and warm-water fish. The ability to apply management
practices to limit potential damages is likely to be low for ecosystems that are not already
intensively managed.
Hydrology and Water Resources. Water quantity and quality are particularly sensitive to
climate change. Potential impacts include increased runoff in winter and spring and decreased
soil moisture and runoff in summer. The Great Plains and prairie regions are particularly
vulnerable. Projected increases in the frequency of heavy rainfall events and severe flooding also
could be accompanied by an increase in the length of dry periods between rainfall events and in
the frequency or severity of droughts in parts of North America. Water quality could suffer and
would decline where minimum river flows decline. Opportunities to adapt are extensive, but
their costs and possible obstacles may be limiting.
Food and Fiber Production. The productivity of food and fiber resources of North
America is moderately to highly sensitive to climate change. Most studies, however, have not
fully considered the effects of potential changes in climate variability; water availability; stresses
from pests, diseases, and fire; or interactions with other, existing stresses. Warmer climate
scenarios (4 to 5 °C increases in North America) have yielded estimates of negative impacts in
eastern, southeastern, and corn belt regions and positive effects in northern plains and western
regions. More moderate warming produced estimates of predominately positive effects in some
March 2001
4C-13
DRAFT-DO NOT QUOTE OR CITE
-------
1 warm-season crops. Vulnerability of commercial forest production is uncertain, but is likely to
2 be lower than less intensively managed systems because of changing technology and
3 management options. The vulnerability of food and fiber production in North America is thought
4 to be low at the continental scale, though subregional variation in losses or gains is likely. The
5 ability to adapt may be limited by information gaps; institutional obstacles; high economic,
6 social, and environmental costs; and the rate of climate change.
7 Coastal Systems. Sea level has been rising relative to the land along most of the coast of
8 North America, and falling in a few areas, for thousands of years. During the next century, a
9 50-cm rise in sea level from climate change alone could inundate 8500 to 19,000 square
10 kilometers of dry land, expand the 100-year flood plain by more than 23,000 square kilometers,
11 and eliminate as much as 50% of North America's coastal wetlands. The projected changes in
12 sea level because of climate change alone would underestimate the total change in sea level from
13 all causes along the eastern seaboard and Gulf Coast of North America. In many areas, wetlands
14 and estuarine beaches may be squeezed between advancing seas and dikes or seawalls built to
15 protect human settlements. Several local governments are implementing land-use regulations to
16 enable coastal ecosystems to migrate landward as sea level rises. Saltwater intrusion may
17 threaten water supplies in several areas.
18 Human Settlements. Projected changes in climate could have positive and negative impacts
19 on the operation and maintenance costs of North American land and water transportation. Such
20 changes also could increase the risks to property and human health and life as a result of possible
21 increased exposure to natural hazards (e.g., wildfires, landslides, extreme weather events) and
22 result in increased demand for cooling and decreased demand for heating energy, with the overall
23 net effect varying across geographic regions.
24 Human Health. Climate can have wide-ranging and potentially adverse effects on human
25 health via direct pathways (e.g., thermal stress, extreme weather and climate events) and indirect
26 pathways (e.g., disease vectors and infectious agents, environmental and occupational exposures
27 to toxic substances, food production). In high-latitude regions, some human health impacts are
28 expected because of dietary changes resulting from shifts in migratory patterns and abundance of
29 native food sources.
30 Conclusions. Taken individually, any one of the impacts of climate change may be within
31 the response capabilities of a subregion or sector. The fact that they are projected to occur
March 2001
4C-14
DRAFT-DO NOT QUOTE OR CITE
-------
1 simultaneously and in concert with changes in population, technology, economics, and other
2 environmental and social changes, however, adds to the complexity of the impact assessment and
3 the choice of appropriate responses. The characteristics of subregions and sectors of North
4 America suggest that neither the impacts of climate change nor the response options will be
5 uniform.
6 Many systems of North America are moderately to highly sensitive to climate change, and
7 the range of estimated effects often includes the potential for substantial damages. The
8 technological capability to adapt management of systems to lessen or avoid damaging effects
9 exists in many instances. The ability to adapt may be diminished, however, by the attendant
10 costs, lack of private incentives to protect publicly owned natural systems, imperfect information
11 regarding future changes in climate and the available options for adaptation, and institutional
12 barriers. The most vulnerable sectors and regions include long-lived natural forest ecosystems in
13 the east and ulterior west, water resources in the southern plains, agriculture in the southeast and
14 southern plains, human health in areas currently experiencing diminished urban air quality,
15 northern ecosystems and habitats, estuarine beaches in developed areas, and low-latitude cool
16 and cold-water fisheries. Other sectors and subregions may benefit from opportunities associated
17 with warmer temperatures or, potentially, from CO2 fertilization, including west coast coniferous
18 forests; some western rangelands; reduced energy costs for heating in the northern latitudes;
19 reduced salting and snow-clearance costs; longer open-water seasons in northern channels and
20 ports; and agriculture in the northern latitudes, the interior west, and the west coast.
21
22 Polar (Arctic and Antarctic) Regions
23 The polar regions include some very diverse landscapes, and the Arctic and the Antarctic
24 are very different in character. The Arctic is defined here as the area within the Arctic Circle; the
25 Antarctic here includes the area within the Antarctic Convergence, including the Antarctic
26 continent, the Southern Ocean, and the sub-Antarctic islands. The Arctic can be described as a
27 frozen ocean surrounded by land, and the Antarctic as a frozen continent surrounded by ocean.
28 The projected warming in the polar regions is greater than for many other regions of the world.
29 Where temperatures are close to freezing on average, global warming will reduce land ice and
30 sea ice, the former contributing to sea-level rise. However, in the interiors of ice caps, increased
March 2001
4C-15
DRAFT-DO NOT QUOTE OR CITE
-------
1 temperature may not be sufficient to lead to melting of ice and snow, and will tend to have the
2 effect of increasing snow accumulation.
3 Ecosystems. Major physical and ecological changes are expected in the Arctic. Frozen
4 areas close to the freezing point will thaw and undergo substantial changes with warming.
5 Substantial loss of sea ice is expected in the Arctic Ocean. As warming occurs, there will be
6 considerable thawing of permafrost, leading to changes in drainage, increased slumping, and
7 altered landscapes over large areas. Polar warming probably will increase biological production
8 but may lead to different species composition on land and in the sea. On land, there will be a
9 tendency for polar shifts in major biomes such as tundra and boreal forest and associated
10 animals, with significant impacts on species such as bear and caribou. However, the Arctic
11 Ocean geographically limits northward movement. Much smaller changes are likely for the
12 Antarctic, but there may be species shifts. In the sea, marine ecosystems will move poleward.
13 Animals dependent on ice may be disadvantaged in both polar areas.
14 Hydrology and Water Resources. Increasing temperature will thaw permafrost and melt
15 more snow and ice. There will be more running and standing water. Drainage systems in the
16 Arctic are likely to change at the local scale. River and lake ice will break up earlier and freeze
17 later.
18 Food and Fiber Production. Agriculture is severely limited by the harsh climate. Many
19 limitations will remain in the future, although some small northern extension of farming into the
20 Arctic may be possible. In general, marine ecological productivity should rise. Warming should
21 increase growth and development rates of nonmammals; ultraviolet-B (UV-B) radiation is still
22 increasing, however, which may adversely affect primary productivity as well as fish
23 productivity.
24 Coastal Systems. As warming occurs, the Arctic could experience a thinner and reduced
25 ice cover. Coastal and river navigation will increase, with new opportunities for water transport,
26 tourism, and trade. The Arctic Ocean could become a major global trade route. Reductions in
27 ice will benefit offshore oil production. Increased erosion of Arctic shorelines is expected from a
28 combination of rising sea level, permafrost thaw, and increased wave action as a result of
29 increased open water. Further breakup of ice shelves in the Antarctic peninsula is likely.
30 Elsewhere in Antarctica, little change is expected in coastlines and probably in its large ice
31 shelves.
March 2001
4C-16
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Human Settlements. Human communities in the Arctic will be substantially affected by the
projected physical and ecological changes. The effects will be particularly important for
indigenous peoples leading traditional lifestyles. There will be new opportunities for shipping,
the oil industry, fishing, mining, tourism, and migration of people. Sea ice changes projected for
the Arctic have major strategic implications for trade, especially between Asia and Europe.
Conclusions. The Antarctic peninsula and the Arctic are very vulnerable to projected
climate change and its impacts. Although the number of people directly affected is relatively
small, many native communities will face profound changes that impact on traditional lifestyles.
Direct effects could include ecosystem shifts, sea and river-ice loss, and permafrost thaw.
Indirect effects could include feedbacks to the climate system such as further releases of
greenhouse gases, changes in ocean circulation drivers, and increased temperature and higher
precipitation with loss of ice, which could affect climate and sea level globally. The interior of
Antarctica is less vulnerable to climate change, because the temperature changes envisaged over
the next century are likely to have little impact and very few people are involved. However,
there are considerable uncertainties about the mass balance of the Antarctic ice sheets and the
future behavior of the West Antarctic ice sheet (low probability of disintegration over the next
century). Changes in either could affect sea level and Southern Hemisphere climates.
REFERENCES
Intergovernmental Panel on Climate Change (IPCC). (1996) Climate change 1996: contribution of working group I
to the second assessment of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom:
Cambridge University Press; p. 572. [IPCC second assessment report].
Intergovernmental Panel on Climate Change (IPCC). (1998) The regional impacts of climate change: an assessment
of vulnerability. Cambridge, United Kingdom: Cambridge University Press.
March 2001
4C-17
DRAFT-DO NOT QUOTE OR CITE
-------
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
APPENDIX 4D
Excerpted Materials from the U.S. Global Change Research
Program Assessment Overview Report on Climate Change Impacts
on the United States (U.S. Global Change Research Program, 2000) and
Subsidiary Regional Assessment Reports
The subject Assessment Overview on Climate Change Impacts on the United States
(USGCRP, 2000) was prepared by the USGCRP National Assessment Synthesis Team (NAST)
and represents an important landmark for the U.S. national process of research analyses and
dialog about coming changes in climate, their impacts, and possible adaptation measures that can
be taken. NAST consists of a committee of experts drawn from governments, academia industry,
and nongovernmental organizations (NGO's). The subject overview is based on a much more
extensive, detailed "foundation" report, written by NAST in coordination with independent
regional and sector assessment teams. The subject assessment, required by a 1990 U.S. law, was
conducted under the USGCRP in response to a request from the President's Science Advisor.
The materials presented below are excerpted from the September 2000 NSTC review draft of the
assessment overview and, if necessary, later will be appropriately corrected to reflect the final
versions of the report due out in fall 2000, after completion of all peer-review and clearance
precesses. Selected material derived from one or another of the specific regional assessments
also are presented in this appendix. The materials selected for presentation here are meant to
provide an informative introduction to the latest available expert assessment of potential sector
and regional-scale impacts of climate change in the United States and to illustrate the difficulties
in projecting likely varying location-specific mixes of potential deleterious and beneficial effects
of climate change.
The past record of 1000 years of global temperature and CO2 emissions change, as
depicted by the assessment overview, is shown in Figure 4D-1. As noted in the Figure 4D-1,
there appears to be a relatively close correlation between marked parallel increases in
anthropogenic carbon emissions starting roughly in the latter part of the 18th century, increasing
atmospheric CO2 concentrations, and notable increasing global average temperature trends.
March 2001
4D-1
DRAFT-DO NOT QUOTE OR CITE
-------
Temperature Change
380-
Q-, 360
•§ 340
2"
§ 320
W 300
e
ra
0- 280
O
260
CO2 Concentrations
& s s
Figure 4D-1. Records of Northern Hemisphere surface temperatures, CO2 concentrations,
and carbon emissions show a close correlation. Temperature Change:
reconstruction of annual-average Northern Hemisphere surface air
temperatures derived from historical records, tree rings, and corals (blue),
and air temperatures directly measured (purple). CO2 Concentrations:
record of global CO2 concentration for the last 1000 years, derived from
measurements of CO2 concentration in air bubbles in the layered ice cores
drilled in Antarctica (blue line) and from atmospheric measurements since
1957. Carbon Emissions: reconstruction of past emissions of CO2 as a result
of land clearing and fossil fuel combustion since about 1750 (in billions of
metric tons of carbon per year.
March 2001
4D-2
DRAFT-DO NOT QUOTE OR CITE
-------
1 The two primary models used to project future changes in global climate in the USGCRP
2 assessment were developed at the Canadian Climate Centre and the Hadley Centre in the United
3 Kingdom and have been peer-reviewed extensively by other scientists. Both incorporate similar
4 assumptions about future emissions of carbon dioxide and other major greenhouse gases (both
5 approximate the IPCC "business as usual" scenario with a 1% per year increase in greenhouse
6 gases and growing sulfur emissions). These models were the best fit to a list of criteria
7 developed for the U.S. National Assessment. Climate models developed at the National Center
8 for Atmospheric Research (NCAR), NOAA's Geophysical Fluid Dynamics Laboratory (GFDL),
9 NASA's Goddard Institute for Space Studies (GISS), and Max Planck Institute (MPI) in
10 Germany also were used in various aspects of the assessment. Although the physical principles
11 driving the models are similar, they differ in how they represent the effects of some important
12 processes, with the two primary models yielding different views of 21 st century climate. On
13 average over the United States, the Hadley model projects a much wetter climate than does the
14 Canadian model, although the Canadian model projects a greater increase in temperature than
15 does the Hadley. Both projections are plausible, given current understanding. See Figure 4D-2
16 for plots of U.S. average temperature increases projected by the different models. In all climate
17 models, increases in temperature for the United States are significantly higher than global
18 average temperature increases (see Table 4D-1), because of the fact that all models project
19 warming to be greatest at middle to high latitudes (partly because melting snow and ice make the
20 surface less reflective of sunlight, allowing it to absorb more heat). Warming also will be greater
21 over land than over the oceans because it takes longer for the oceans to warm.
22 Uncertainties about future climate stem from a wide variety of factors (e.g., questions about
23 how to represent clouds and precipitation in climate models and uncertainties about how
24 emissions of greenhouse gases will change). These uncertainties result in differences in climate
25 model projections. Examining these differences aids in understanding the range of risk or
26 opportunity associated with a plausible range of future climate changes. These differences in
27 model projections also raise questions about how to interpret model results, especially at the
28 regional level, where projections can differ significantly.
29 One of the most important world-wide consequences of the overall global warming
30 increases projected for the 21 st century is sea level rise, and it can be expected to impact Alaska,
31 coastal areas of the continental United States, and U.S. Hawaiian and Carribean islands regions,
March 2001
4D-3
DRAFT-DO NOT QUOTE OR CITE
-------
LL.
o
Q)
O>
O
Q)
•*->
£
CD
D.
I
9-
8
7
6
5
3-
2-
1-
0-
-1-
-3
-4-
-5
Changes in Temperature over the U.S.
Simulated by Climate Models
Hadley Centre - Version 2
Canadian Centre
Max Planck Institute
Geophysical Fluid Dynamics Laboratory
Hadley Centre - Version 3
NCAR - Parallel Version
NCAR - Climate System Model
1850
1900
I
1950
2000
2050
2100
Figure 4D-2. Simulation of decadal average changes in temperature from leading climate
models on historic and projected changes in CO2 and sulfate atmospheric
concentrations. For the 21st century, the projected global temperature
increase is 4.9 °F for the Hadley model and 7.4 °F for the Canadian model.
The model with the smallest projected increase in global temperature is the
NCAR Climate System Model at 3.6 °F. By comparison, the projected
increase in temperature for the 21st century over the contiguous United
States is Canadian, 9.4 °F, Hadley, 5.5 °F, and NCAR Climate System
Model, 4.0 °F.
March 2001
4D-4
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4D-1. RANGE OF PROJECTED WARMING IN THE 21ST CENTURY
Global United States
Hadley Model +4.9 °F +5.5 °F
Canadian Model +7.4 °F +9.4 °F
NCAR Climate System Model +3.6 °F +4.0 °F
1 as well. Figure 4D-3 illustrates sea level rise predicted by the Canadian and Hadley models used
2 in the USGCRP assessment. The sea level rise projected by either model can be expected to pose
3 threats, not only in terms of potential inundation of low lying portions of the Hawaiian and
4 Carribean islands, but also in terms of shoreline erosion in portions of Alaska and the continental
5 United States. Those continental United States areas most vulnerable to future sea level rise are
6 those low lying areas already experiencing rapid erosion rates, as depicted in Figure 4D-4.
7 Substantial impacts can be expected, including losses of coastal wetlands important for migratory
8 birds and degradation of estuarine sound complexes providing shallow water fishery nurseries
9 (most immediately because of salt water incursions resulting from sea level rise and other
10 impacts resulting from more frequent and extensive algal-toxic blooms impacting coastal
11 commercial fish and shell fish harvests secondary to increased nutrient out flows caused by
12 extreme rain fall events [e.g., during hurricanes]).
13 The main climate models used all predict notable increases in the minimum and maximum
14 annual average temperatures in the United States during the next 100 years. Projected changes in
15 temperature minimum and maxima are likely more important than average temperatures, in that
16 they influence such things as human comfort, heat and cold stress in plants and animals,
17 maintenance of snow pack, and pest populations (low temperatures kill many pests and higher
18 minimum temperatures may allow increased overwinter survival of pests). The largest increases
19 in temperature are projected over much of the southern United States in summer, dramatically
20 raising the heat index (a measure of discomfort based on temperature and humidity). Also,
21 following an average 5 to 10% increase in average U.S. precipitation over the last century, the
22 climate models project notable changes in precipitation during the 21 st century. The Canadian
March 2001
4D-5
DRAFT-DO NOT QUOTE OR CITE
-------
Canadian Model (Thermal Expansion)
Hadley Model (Themial Expansion)
Hadley Model (T.E. + glacial melt)
1850
1950
2000
2050
2100
Figure 4D-3. Historic and projected changes in sea level (in inches) based on the Canadian
and Hadley model simulations. The Canadian model projection includes
only the effects of thermal expansion of warming ocean waters. The Hadley
projection includes both thermal expansion and the additional sea-level rise
projected because of melting of land-based glaciers. Neither model includes
consideration of possible sea-level changes because of polar ice melting or
accumulation of snow on Greenland and Antarctica.
Severely eroding
Figure 4D-4. This map is a preliminary classification of annual shoreline erosion
throughout the United States, in coarse detail and resolution. The areas most
vulnerable to future sea-level change are those with low relief that are already
experiencing rapid erosion rates, such as the Southeast and Gulf Coast.
March 2001
4D-6
DRAFT-DO NOT QUOTE OR CITE
-------
1 model predicts the largest percentage increases in precipitation in California and the Southwest;
2 but east of the Rocky Mountains, the southern half of the United States is projected to have
3 decreased precipitation (with especially large decreases in eastern Colorado, western Kansas, and
4 in an arc stretching from Louisiana to Virginia). The Hadley model predicts largest percentage
5 increases in southern California and the Southwest, along with lesser increases for the rest of the
6 nation, except for small areas of the Northwest and the Gulf Coast. Both models also predict
7 increases in frequency of heavy precipitation effects, largely because of shifts in storm activity
8 and tracks. Soil moisture critical for both agriculture and natural ecosystems may, despite
9 increased precipitation, actually undergo marked decreases in some areas, because of offsetting
10 evaporation rates increased by higher temperatures during projected scenarios of likely increased
11 periods of drought for some U.S. regions.
12 The predicted changes in temperature and precipitation are expected to result in varying
13 impacts of climate change on ecosystems various U.S. regions. Such impacts will likely include
14 the following.
15 • Changes in productivity and carbon storage capacity of ecosystems (decreases in some places
16 and increases in others are very likely).
17 • Shifts in the distribution of major plant and animal species are likely.
18 • Some ecosystems, such as alpine meadows, are likely to disappear in some places because the
19 new local climate will not support them or there are barriers to their movement.
20 • In many places, it is very likely that ecosystem services, such as air and water purification,
21 landscape stabilization against erosion, and carbon storage capacity will be reduced. These
22 losses likely will occur in the wake of episodic, large-scale disturbances that trigger species
23 migrations or local extinctions.
24 • In some places, it is very likely that ecosystems services will be enhanced where climate-
25 related stresses are reduced.
26 The USGCRP assessment provides extensive detailed evaluations of the above and other
27 types of impacts projected to occur as consequences of changing weather patterns (and
28 consequent shifts in temperature, precipitation, etc.). Such evaluations are summarized in the
29 overview assessment in relation to several overall sectors (water resources, agriculture, forests,
30 coastal areas and marine resources, and human health) and in relation to different regions of the
31 United States.
March 2001
4D-7
DRAFT-DO NOT QUOTE OR CITE
-------
1 The following concise statements highlight some of the more salient points emerging from
2 the overall sector evaluations.
3 • Water. Rising temperatures and greater precipitation are likely to lead to more evaporation
4 and greater swings between wet and dry conditions. Changes in the amount and timing of rain,
5 snow, runoff, and soil moisture are very likely. Water management, including pricing and
6 allocation, will very likely be important in determining many impacts.
7 • Agriculture. Overall productivity of American agriculture likely will remain high and is
8 projected to increase throughout the 21 st century, with northern regions faring better than
9 southern ones. Though agriculture is highly dependent on climate, it is also highly adaptive
10 Weather extremes, pests, and weeds likely will present challenges in a changing climate.
11 Falling commodity prices and competitive pressures are likely to stress farmers and rural
12 communities.
13 • Forests. Rising CO2 concentrations and modest warming are likely to increase forest
14 productively in many regions. With larger increases in temperature, increased drought is likely
15 to reduce forest productivity in some regions, notably in the Southeast and Northwest. Climate
16 change is likely to cause shifts in species ranges, as well as large changes in disturbances such
17 as fire and pests.
18 • Coastal Areas and Marine Resources. Coastal wetlands and shorelines are vulnerable to
19 sea-level rise and storm surges, especially when climate impacts are combined with the
20 growing stressed of increasing human population and development. It is likely that coastal
21 communities will be affected increasingly by extreme events. The negative impacts on natural
22 ecosystems are very likely to increase.
23 • Human Health. Heat-related illnesses and deaths, air pollution, injuries and deaths from
24 extreme weather events, and diseases carried by water, food, insects, ticks, and rodents, have
25 all been raised as concerns for the United States in a warmer world. Modern public health
26 efforts will be important in identifying and adapting to these potential impacts.
27 The USGCRP Assessment also evaluated sector impacts in relation to various U.S. regions,
28 broken out as depicted in Figure 4D-5 derived from the overview assessment (USGCRP, 2000).
29 That assessment highlighted the following important points in relation to expected major impacts
30 in each of the regions evaluated.
31
March 2001
4D-8
DRAFT-DO NOT QUOTE OR CITE
-------
C
O
en
es
U
O
•o
-------
1 • Alaska. Sharp winter and springtime temperature increases are very likely to cause continued
2 thawing of permafrost, further disrupting forest ecosystems, roads, and buildings.
3 • Northwest. Increasing stream temperatures are very likely to further stress migrating fish,
4 complicating restoration efforts.
5 • Mountain West. Higher winter temperatures are very likely to reduce snowpack and peak
6 runoff and shift the peak to earlier in the spring, reducing summer runoff and complicating
7 water management for flood control, fish runs, cities, and irrigation.
8 • Southwest. With an increase in precipitation, the desert ecosystems native to this region are
9 likely to decline, whereas grasslands and shrublands likely are to expand.
10 • Midwest/Great Plains. Higher CO2 concentrations are likely to offset the effects of rising
11 temperatures on forests and agriculture for several decades, increasing productivity.
12 • Southern Great Plains. Prairie potholes, which provide important habitat for ducks and other
13 migratory waterfowl, are likely to dry up in a warmer climate.
14 • Great Lakes. Lake levels are likely to decline, leading to reduced water supply and more
15 costly transportation. Shoreline damage caused by high water levels is likely to decrease.
16 • Northern and Mountain Regions. It is very probable that warm weather recreational
17 opportunities, such as hiking, will expand, whereas cold weather activities, such as skiing, will
18 contract.
19 • Northeast, Southeast, and Midwest. Rising temperatures are very likely to increase the heat
20 index dramatically in summer, with impacts to health and comfort. Warmer winters are likely
21 to reduce cold-related stresses.
22 • Appalachians. Warmer and moister air very likely will lead to more intense rainfall events,
23 increasing the potential for flash floods.
24 • Southeast. Under warmer wetter scenarios, the range of southern tree species is likely to
25 expand. Under hotter and drier scenarios, it is likely that far southeastern forests will be
26 displaced by grasslands and savannas.
27 • Southeast Atlantic Coast. It is very probable that rising sea levels and storm surge will
28 threaten natural ecosystems and human coastal development and reduce buffering capacity
29 against storm impacts.
30 • Southeast Gulf Coast. Inundation of coastal wetlands will very likely increase, threatening
31 fertile areas for marine life, migrating birds, and waterfowl.
March 2001 4D-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 • Islands. More intense El Nino and La Nina events are possible and are likely to create extreme
2 fluctuations in water resources for island citizens and the tourists who sustain local economies.
3 Other materials from the overview assessment summarize regional concerns with regard to
4 different types of sector impacts. Tables 4D-2 and 4D-3 present two examples drawn from the
5 overview assessment, denoting concerns or impacts regarding water resources and types of
6 ecosystems, respectively, likely to be impacted in different U.S. regions.
TABLE 4D-2. TYPES OF WATER CONCERNS PROJECTED TO BE IMPORTANT
FOR U.S. REGIONS CONSEQUENT TO FUTURE CLIMATE CHANGE"
Region
Northeast
Southeast
Midwest
Great Plains
West
Northwest
Alaska
Islands
Floods
X
X
X
X
X
X
X
Droughts
X
X
X
X
X
X
X
X
Snowpack
X
X
X
X
X
X
Groundwater
X
X
X
X
X
X
Lake, River, and
Reservoir Levels
X
X
X
X
Quality
X
X
X
X
X
X
This table identifies some of the key regional concerns about water. Many of these issues were raised and
discussed by stakeholders during regional workshops and other Assessment meetings held between 1997 and
2000.
1 As seen in Table 4D-2, different types of water impacts are projected to be of important
2 widespread concern across many different U.S. regions. It should be noted that some limited
3 beneficial effects may occur in some regions (e.g., longer periods of open-water transportation on
4 navigable rivers and sounds in and around Alaska).
5 The overview assessment notes that the information presented in Table 4D-3 represents
6 only a partial list of potential impacts for major ecosystem types and that, although the impacts
7 often are stated in terms of plant-community impacts, it is important to recognize that such plant-
8 community changes also will have animal habitat effects and consequent impacts on both
9 terrestrial and aquatic animal species. Both the plant and animal impacts can have further
10 consequent impacts on human health and welfare, which also can be expected to vary
11 considerably from region to region.
March 2001
4D-11
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4D-3. PROJECTED FUTURE CLIMATE-CHANGE-INDUCED IMPACTS
ON TYPES OF ECOSYSTEMS OF CONCERN TO DIFFERENT U.S. REGIONS
Ecosystem
Type
Impacts
U.S. Regions
Forests Changes in tree species composition and
alteration of animal habitat
Displacement of forests by woodlands and
grasslands under a warmer climate in
which soils are drier
Grasslands Displacement of grasslands by woodlands
and forests under a wetter climate
Increase in success of nonnative invasive
plant species
Tundra Loss of alpine meadows as their species
are displaced by lower elevation species
Loss of northern tundra as trees migrate
poleward
Changes in plant community composition
and alteration of animal habitat
Semi-arid Increase in woody species and loss of
and Arid desert species under wetter climate
Freshwater Loss of prairie pot holes with more
frequent drought conditions
Habitat changes in rivers and lakes as
amount and timing of runoff changes and
water temperatures rise
Coastal and Loss of coastal wetlands as sea level rises
Marine and coastal development prevents
landward migration
Loss of barrier islands as sea-level rise
prevents landward migration
Changes in quantity and quality of
freshwater delivery to estuaries and bays
alter plant and animal habitats
Loss of coral reefs as water temperature
increases
Changes in ice location and duration alter
marine mammal habitat
NE SE MW GP WE PNW AK IS
X
X
X
X
X
X
X
March 2001
4D-12
DRAFT-DO NOT QUOTE OR CITE
-------
1 The USGCRP (2000) assessment included extensive detailed evaluation of projected
2 climate change impacts on the different U.S. regions depicted in Figure 4D-5. Overview reports
3 on those detailed evaluations by various regional assessment teams are in various stages of
4 preparation, with information pertaining to each being available via the internet at the following
5 address: http://www.nacc.usgcrp.gov/regions/.
6 The wide variation in the types of projected impacts of climate change, both deleterious and
7 some possible beneficial effects, can be readily illustrated by one example illustrated in
8 Figure 4D-6. The figure depicts projected types of changes that may occur (with varying degrees
9 of certainty indicated) as the consequence of climate change impacts on the Mid-Atlantic Region
10 (MAR) of the United States, including both potentially negative and positive impacts.
11
Summary of MAR impacts
Positive Impact
Most Certain
• Agricultural production
• Coastal zones
• Temperature related health status
erosion,
saltwater intrusion
soybeans,
possibly corn
and treefruits
Moderately Certain
• Forestery production
• Temperature related health status
less cold stress
Uncertain
• Biodiversity
• Fresh water quantity
• Fresh water quality
• Ecological functioning
• Vector and water-borne disease health status
• Environmental effects from agriculture
migration barriers,
invasive species
forest composition,
cold water fisheries
nutrient leaching,
runoff
Figure 4D-6. Projected climate change impacts in the Mid-Atlantic Region (MAR) of the
United States.
Source: Mid-Atlantic Regional Assessment Team (2000).
March 2001
4D-13
DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Mid-Atlantic Regional Assessment (MARA) Team. (2000) Preparing for a changing climate: the potential
3 consequences of climate variability and change (Mid-Atlantic overview). Washington, DC: U.S. Global
4 Change Research Program (USGCRP).
5 „ U. S. Global Change Research Program (USGCRP, 2000) Climate Change Impacts on the United States: the
6 Potential Consequences of Climate Variability and Change (Overview), Report of National Assessment
7 Synthesis Team (NAST). NSTC Review Draft (September 2000).
8
March 2001
4D-14
DRAFT-DO NOT QUOTE OR CITE
-------
APPENDIX 4E
Recent Model Projections of Excess Mortality Expected in U.S. Cities
During Summer and Winter Seasons Because of Future Climate Change,
Based on Kalkstein and Greene (1997)
March 2001
4E-1
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4E-1. MODELED PROJECTIONS OF DIRECT HUMAN HEALTH
IMPACTS OF CLIMATE CHANGE: ESTIMATED TOTAL EXCESS MORTALITY
IN U.S. URBAN AREAS FOR AN AVERAGE SUMMER SEASON,
ASSUMING FULL ACCLIMATIZATION3
Year 2020 Climate
SMSA
Anaheim
Atlanta
Baltimore
Birmingham
Boston
Buffalo
Chicago
Cincinnati
Cleveland
Columbus
Dallas
Denver
Detroit
Ft. Lauderdale
Greensboro
Hartford
Houston
Indianapolis
Jacksonville
Kansas City
Los Angeles
Louisville
Memphis
Miami
Minneapolis
Nassau
New Orleans
New York
March 2001
Present
climate
0
25
84
42
96
33
191
14
29
33
36
42
110
0
22
38
7
36
0
49
68
17
25
0
59
29
20
307
GFDL 89
0
,43
57
26
113
15
243
16
21
24
45
29
84
0
28
21
7
23
0
79
74
0
42
0
55
59
0
363
Year 2050 Climate
UKMO Max Planck GFDL 89
0
62
148
47
165
52
538
90
55
83
62
41
240
0
43
42
16
93
0
173
123
2
27
0
185
84
0
753
0
22
63
14
134
36
421
49
44
51
45
30
164
0
27
32
7
51
0
93
83
0
57
0
148
84
0
498
4E-2
0
60
124
40
155
34
359
54
46
51
107
35
130
0
37
38
15
55
0
121
110
0
40
0
123
110
0
460
UKMO Max Planck
0
138
164
47
194
73
583
81'
58
90
64
39
271
0
45
50
17
86
0
156
128
1
29
0
215
116
0
999
DRAFT-DO NOT QUOTE
0
33 .
131
21
160
59
550
67
53
78
44
32
256
0
29
41
6
69
0
105
116
1
49
0
186
116
0
727
OR CITE
-------
TABLE 4E-1 (cont'd). MODELED PROJECTIONS OF DIRECT HUMAN HEALTH
IMPACTS OF CLIMATE CHANGE: ESTIMATED TOTAL EXCESS MORTALITY
IN U.S. URBAN AREAS FOR AN AVERAGE SUMMER SEASON,
ASSUMING FULL ACCLIMATIZATION8
Year 2020 Climate
SMSA
Newark
Philadelphia
Phoenix
Pittsburgh
Portland
Providence
Riverside
Salt Lake City
San Antonio
San Diego
San Francisco
San Jose
Seattle
St. Louis
Tampa
Washington, DC
Total
Present
climate
26
129
0
39
9
47
4
0
4
0
28
0
5
79
28
0
1,840
GFDL 89
83
99
0
32
13
39
6
0
0
0
24
0
1
149
68
0
1,981
UKMO
173
362
0
66
.22
80
10
0
0
0
23
0
0
173
95
0
4,128
Max Planck
111
191
0
64
11
52
6
0
0
0
23
0
2
158
28
0
2,799
Year 2050 Climate
GFLD 89
150
246
0
61
23
73
8
0
0
0
18
0
0
212
95
0
3,790
UKMO
127
477
0
83
31
96
11
0
0
0
24
0
0
155
100
0
4,748
Max Planck
161
323
0
95
14
74
7
0
0
0
23
0
1
189
47
0
3,863
"Abbreviations: SMSA, standard metropolitan statistical area; GFDL, Geophysical Fluid Dynamics Laboratory
Model; UKMO, United Kingdom Meteorological Office Model; Max Planck, Max Planck Institute Model.
Values given are estimated excess deaths.
Source: Kalkstein and Green (1997).
March 2001
4E-3
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4E-2. MODELED PROJECTIONS OF DIRECT HUMAN HEALTH
IMPACTS OF GLOBAL CLIMATE CHANGE: ESTIMATED TOTAL EXCESS
MORTALITY IN U.S. URBAN AREAS FOR AN AVERAGE WINTER SEASON,
ASSUMING FULL ACCLIMATIZATION"
Year 2020 climate
SMSA
Anaheim
Atlanta
Baltimore
Birmingham
Boston
Buffalo
Chicago
Cincinnati
Cleveland
Columbus
Dallas
Denver
Detroit
Ft. Lauderdale
Greensboro
Hartford
Houston
Indianapolis
Jacksonville
Kansas City
Los Angeles
Louisville
Memphis
Miami
Minneapolis
Nassau
Present
Climate
2
37
0
25
0
7
2
0
2
12
32
9
34
36
0
0
24
16
0
12
100
16
23
46
0
24
GFDL 89
0
53
0
12
0
18
4
0
9
1
41
10
15
4
0
0
33
32
0
51
102
12
20
35
0
21
UKMO
0
48
0
8
0
8
3
0
10
2
33
11
20
4
0
0
29
28
0
36
78
17
17
35
0
4
Max Planck
1
52
0
11
0
17
4
0
10
1
43
10
15
5
0
0
35
33
0
46
100
12
19
37
0
20
Year 2050 climate
GFDL 89
0
50
0
11
0
5
4
0
15
3
36
11
18
3
0
0
29
34
0
42
77
19
19
32
0
5
UKMO
0
47
0
7
0
5
2
0
10
2
31
11
25
3
0
0
27
28 •
0
35
88
15
15
32
0
3
Max Planck
0
52
0
12
0
' 18
5
0
9
1
41
11
14
5
0
0
33
32
0
46
81
12
19
36
0
21
March 2001
4E-4
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4E-2 (cont'd). MODELED PROJECTIONS OF DIRECT HUMAN HEALTH
IMPACTS OF GLOBAL CLIMATE CHANGE: ESTIMATED TOTAL EXCESS
MORTALITY IN U.S. URBAN AREAS FOR AN AVERAGE WINTER SEASON,
ASSUMING FULL ACCLIMATIZATION3
Year 2020 climate
SMSA
New Orleans
New York
Newark
Philadelphia
Phoenix
Pittsburgh
Portland
Providence
Riverside
Salt Lake City
San Antonio
San Diego
San Francisco
San Jose
Seattle
St. Louis
Tampa
Washington, DC
Total
Present
Climate
52
102
48
85
26
19
17
27
10
5
9
17
85
3
13
50
21
19
1,067
GFDL 89
56
123
23
80
25
20
15
21
26
7
10
26
39
2
40
61
26
31
1,104
UKMO
51
150
8
14
26
29
12
34
29
9
6
16
30
4
45
68
24
38
984
Max Planck
54
120
20
73
25
21
15
33
26
8
11
24
42
2
37
60
26
30
1,098
Year 2050 climate
GFDL 89
51
152
10
36
26
24
12
35
27
8
5
16
30
3
46
53
22
20
989
UKMO
47
93
6
9
27
31
10
36
27
10
4
18
21
5
47
61
20
35
894
Max Planck
54
121
23
82
26
21
13
21
26
9
9
16
26
4
43
61
25
31
1,059
"Abbreviations: SMSA, standard metropolitan statistical area; GFDL, Geophysical Fluid Dynamics Laboratory
Model; UKMO, United Kingdom Meteorological Office Model; Max Planck, Max Planck Institute Model.
Values given are estimated excess deaths.
REFERENCE
Kalkstein, L. S.; Greene, J. S. (1997) An evaluation of climate/mortality relationships in large U.S. cities and the
possible impacts of a climate change. Environ. Health Perspect. 105: 84-93.
March 2001
4E-5
DRAFT-DO NOT QUOTE OR CITE
-------
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
5. HUMAN EXPOSURE TO PARTICULATE MATTER
AND ITS CONSTITUENTS
5.1 INTRODUCTION
5.1.1 Purpose
Exposure is defined as the contact by an individual with a pollutant for a specific duration
of time at a visible external boundary (modified from Duan 1982,1991). For airborne particulate
matter (PM), the breathing zone is considered the point of contact and the lung is the external
boundary of concern. An individual's exposure is measured as the PM air concentration in
his/her breathing zone over time. Understanding exposure is important, because it is individuals
who experience adverse health effects associated with elevated PM concentrations in ambient air.
The U.S. Environmental Protection Agency's (EPA's) regulatory authority for PM applies
primarily to ambient air and those sources that contribute to ambient PM air concentrations.
»
Thus, a major emphasis must be to develop an understanding of exposure to PM from ambient
sources. However, personal exposure to total PM may result from exposure to PM from both
ambient and nonambient sources. Therefore, it will be necessary to account for both in order to
fully understand the relationship between PM and health effects. Personal exposure to PM from
nonambient sources may be a confounder in community-based epidemiological studies in which
ambient PM measures are correlated with community health parameters. In addition, an
individual's personal exposure to ambient, nonambient, and total PM would provide useful
information for studies where health outcomes are tracked individually.
The overall purpose of this chapter is to provide current exposure information that will aid
in the understanding and interpretation of PM dosimetry, toxicology, and epidemiology studies
assessed in later chapters. The specific objectives of this chapter, which are described below, are
fourfold.
(1) To provide an overall conceptual framework of exposure science as applied to PM, including
the identification and evaluation of factors that determine personal exposure to total PM and
to PM from ambient and nonambient PM sources
March 2001
5-1
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15,
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
(2) To provide a concise summary and review of recent data (since 1996) and findings from
pertinent studies of personal exposure to total PM and to PM from ambient and nonambient
sources
(3) To characterize quantitative relationships between ambient air quality measurements (mass,
chemical components, number, etc.), as determined by a community monitoring site, and
total personal PM exposure as well as its ambient and nonambient components
(4) To evaluate the implications of using ambient PM concentrations as a surrogate for exposure
in epidemiologicaL studies of PM health effects
5.1.2 Particulate Matter Mass and Constituents
Current EPA PM regulations are based on mass as a function of aerodynamic size.
However, EPA also measures the chemical composition of PM in both monitoring and research
studies. The composition of PM is variable and adverse health effects may be related to PM
characteristics other than mass. PM from ambient and nonambient sources also may have
differing physical and chemical characteristics and differing health effects. Ultimately, to
understand and control health impacts caused by PM, it is important to quantify and understand
exposure to those chemical constituents responsible for the adverse health effects. The National
Research Council (NRC) recognized the distinction between measuring exposure to PM mass
and to chemical constituents when setting Research Priorities for Airborne Particulate Matter I:
Immediate Priorities and a Long-range Research Portfolio (NRC, 1998). Specifically, NRC
Research Topic 1 recommends evaluating the relationship between outdoor measures versus
actual human exposure for PM mass. The NRC Research Topic 2 recommends evaluating
exposures to biologically important constituents and specific characteristics of PM that cause
responses in potentially susceptible subpopulations and the general population. It also was
recognized by the NRC that, "a more targeted set of studies under this research topic (#2) should
await a better understanding of the physical, chemical, and biological properties of airborne
particles associated with the reported mortality and morbidity outcomes" (NRC, 1999). The
NRC also stated that the studies "should be designed to determine the extent to which members
of the population contact these biologically important constituents and size fraction of concern in
outdoor air, outdoor air that has penetrated indoors, and air pollutants generated indoors" (NRC,
1999). Thus, when biologically important constituents are identified, exposure studies should
March 2001
5-2
DRAFT-DO NOT QUOTE OR CITE
-------
1 include contributions from all sources. The emphasis in this chapter on PM mass reflects the
2 current state of the science. Where available, data also have been provided on chemical
3 constituents, although in most cases, the data are limited. As recognized by the NRC, a better
4 understanding of exposures to chemical constituents will be required to more fully identify,
5 understand, and control those sources of PM with adverse health effects and to accurately define
6 the relationship between PM exposure and health outcomes.
7
8 5.1.3 Relationship to Past Documents
9 Early versions of PM criteria documents did not emphasize total human exposure but rather
10 focused almost exclusively on outdoor air concentrations. For instance, the 1969 Air Quality
11 Criteria for Particulate Matter (PM AQCD) (National Air Pollution Control Administration,
12 1969) did not discuss either exposure or indoor concentrations. The 1982 PM AQCD (U.S.
13 Environmental Protection Agency, 1982) provided some discussion of indoor PM concentrations,'
14 reflecting an increase in microenvironmental and personal exposure studies. The new data
15 indicated that personal activities, along with PM generated by personal and indoor sources (e.g.,
16 cigarette smoking), could lead to high indoor levels and high personal exposures to total PM.
17 Some studies reported indoor concentrations that exceeded PM concentrations found in the air
18 outside the monitored microenvironments or at nearby monitoring sites.
19 Between 1982 and 1996, many more studies of personal and indoor PM exposure
20 demonstrated that, in most inhabited domestic environments, indoor PM concentrations and
21 personal PM exposures of the residents were greater than ambient PM concentrations measured
22 simultaneously (e.g., Sexton et al., 1984; Spengler et al., 1985; Clayton et al., 1993). As a result,
23 the NRC (1991) recognized the potential importance of indoor sources of contaminants
24 (including PM) in causing adverse health outcomes.
25 The 1996 AQCD (U.S. Environmental Protection Agency, 1996) reviewed the human PM
26 exposure literature through early 1996. Many of the studies cited showed poor correlations
27 between personal exposure or indoor measurements of PM and outdoor or ambient site
28 measurements. Conversely, Janssen et al. (1995) and Tamura et al. (1996a) showed that in the
29 absence of major nonambient sources, total PM exposures to individuals tracked through time
30 were highly correlated with ambient PM concentrations. Analyses of these latter two studies led
31 to consideration of ambient and nonambient exposures as separate components of total personal
March 2001 5-3 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
exposure. As a result, the 1996 PM AQCD (U.S. Environmental Protection Agency, 1996), for
the first time, distinguished between ambient and nonambient PM exposure. This chapter builds
on the work of the 1996 PM AQCD by further evaluating the ambient and nonambient
components of PM, as well as reporting research that evaluates the relationship between ambient
concentrations and total, ambient, and nonambient personal exposure.
5.2 STRUCTURE FOR THE CHAPTER
The chapter is organized to provide information on the principles of exposure, review the
existing literature, and summarize key findings and limitations in the information; the specific
sections are described below.
• Section 5.3 discusses the basic concepts of exposure, including definitions, methods for
estimating exposure, and methods for estimating ambient components of exposure.
• Section 5.4 presents PM mass data, including a description of the key available studies,
correlations of PM exposures with ambient concentrations, and factors that effect the
correlations.
• Section 5.5 presents data on PM constituents, including a description of the key available
studies, correlations with ambient concentrations, and factors that effect the correlations.
• Section 5.6 discusses the implications of using ambient PM concentrations in epidemiological
studies of PM health effects.
• Section 5.7 summarizes key findings and limitations of the information.
5.3 BASIC CONCEPTS OF EXPOSURE
5.3.1 Components of Exposure
The total exposure of an individual over a discrete period of time includes exposures to
many different particles from various sources while in different microenvironments 0/e's). Duan
(1982) defined a microenvironment as "a [portion] of air space with homogeneous pollutant
concentration." It also has been defined as a volume in space, for a specific time interval, during
which the variance of concentration within the volume is significantly less than the variance
March 2001
5-4
DRAFT-DO NOT QUOTE OR CITE
-------
1 between that //e and surrounding ^ue's (Mage, 1985). In general, people pass through a series of
2 /^e's, including outdoor, in-vehicle, and indoor //e's, as they go through time and space. Thus,
3 total daily exposure for a single individual to PM can be expressed as the sum of various
4 microenvironmental exposures that the person occupies in the day (modified from National
5 Research Council, 1991).
6 In a given //e, particles may originate from a wide variety of sources. For example, in an
7 indoor /we, PM may be generated by (1) indoor activities, (2) outdoor PM entering the indoor /^e,
8 (3) the chemical interaction of outdoor air pollutants and indoor air or indoor sources,
9 (4) transport from another indoor //e, or (5) personal activities. All of these disparate sources
10 have to be accounted for in a total human PM exposure assessment.
11 An analysis of personal exposure to PM mass (or constituent compounds) requires
12 definition and discussion of several classes of particles and exposure. In this chapter, PM
13 metrics may be described in terms of exposure or as an air concentration. PM also may be
14 described according to both its source (i.e., ambient, nonambient) and the microenvironment
15 where exposure occurs. Table 5-1 provides a summary of the terms used in this chapter, the
16 notation used for these terms, and their definition. These terms will be used throughout this
17 section and will provide the terminology for evaluating personal exposure to total PM and PM
18 from ambient and nonambient sources.
19
20 5.3.2 Methods To Estimate Personal Exposure
21 Personal exposure may be estimated using either direct or indirect approaches. Direct
22 approaches measure the contact of the person with the chemical concentration in the exposure
23 media over an identified period of time. Direct measurement methods include personal exposure
24 monitors (PEMs) for PM that are worn continuously by individuals as they encounter various
25 microenvironments and perform their daily activities. Indirect approaches use available
26 information on concentrations of chemicals in microenvironments, along with information about
27 the time individuals spend in those microenvironments and personal PM generating activities.
28 The indirect approach then uses models and data on microenvironmental air concentrations and
29 time spent in microenvironments to estimate personal exposure. This section describes the
30 methods to directly measure personal exposures and microenvironmental concentrations, as well
March 2001
5-5
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-1. CLASSES OF PARTICULATE MATTER EXPOSURE AND
CONCENTRATION DEFINITIONS
Term
Notation
Definition
Concentration
Personal Exposure
Microenvironment
Ambient PM
Ambient-Outdoor PM
Indoor PM
Ambient-Indoor PM
Indoor-Generated PM
Personal Exposure to
Indoor-Generated PM
Personal Exposure to
Ambient-Generated PM
Personal Exposure to
Personal-Activity PM
Personal Exposure to
Nonambient PM
C
E
//e
Ca
Cao
C,
Cai
Epact
General Definitions
Air concentration of PM in a given microenvironment, expressed in
/zg/m3
Contact at visible external boundaries of an individual with a pollutant
for a specific duration of time; quantified by the amount of PM available
in concentration units 0"g/m3) at the oral/nasal contact boundary for a
specified time period (At). General term for any exposure variable.
Volume in space, for a specific time interval, during which the variance
of concentration within the volume is significantly less than the variance
between that /ze and surrounding /zes
Concentration Variables
PM in the atmosphere measured at a community ambient monitoring site
either emitted into the atmosphere directly (primary PM) or formed in it
(secondary PM). Major sources of PM species are industry, motor
vehicles, commerce, domestic emissions such as wood smoke, and
natural wind-blown dust or soil.
Ambient PM in an outdoor microenvironment
All PM found indoors
Ambient PM that has infiltrated indoors (i.e., has penetrated indoors and
remains suspended)
PM generated or formed indoors
Exposure Variables
Sum of personal exposure resulting from indoor-generated PM
Sum of personal exposure caused by ambient-outdoor and ambient
indoor PM (does not include resuspended ambient PM previously
deposited indoors)
Small-scale PM-generating activities that primarily influence exposure of
the person performing the activity itself
Sum of personal exposure to indoor-generated and personal activity PM
= Ei + Eact
Personal Exposure to
Total PM
Sum of all personal exposures to ambient and nonambient PM
1 as the models used to estimate exposure. Several approaches to estimate personal exposure to
2 ambient PM also are described.
March 2001
5-6
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
5.3.2.1 Direct Measurement Methods
5.3.2.1.1 Personal Exposure Monitoring Methods
In theory, personal exposure to total PM is measured by sampling the concentration of PM
in inhaled air entering the nose or mouth. Practically, it is defined as that PM collected by a
PEM worn by a person and sampling from a point near the breathing zone (but not impacted by
exhaled breath). The inlet to a PEM normally is placed at the outer limit of the breathing zone to
avoid a negative sampling bias resulting from dilution of the collected air by exhaled breath
depleted of PM. However, such placement does not allow for the sampling of directly inhaled
cigarette smoke or inhaled air that passes through a dust mask. PEMs for PM use measurement
techniques similar to those used for ambient PM. The PEM is a filter-based mass measurement
of a particle size fraction (PM10 or PM2 5), usually integrated over either a 24- or 12-h period at
flow rates of 2 to 4 L/min using battery-operated pumps. PEMs must be worn by study
participants and, therefore, they must be quiet, compact, and battery-operated. These
requirements limit the type of pumps and the total sample volume that can be collected.
Generally, small sample volumes limit personal exposure measurements to PM mass and a few
elements detected by XRF. In most studies, PM2 5 and PMIO have not been collected
concurrently.
Other methods used for ambient PM also have been adapted for use as a personal exposure
monitor. For example, a personal nephelometer that measures particle number within a specific
particle size range using light scattering has been used in personal exposure studies to obtain
real-time measurements of PM.
5.3.2.1.2 Microenvironmental Monitoring Methods
Direct measurements of microenvironmental PM concentrations, which are used with
models to estimate personal exposure to PM, also use methods similar to those for ambient PM.
These methods differ from PEMs in that they are stationary with respect to the microenvironment
(such as a stationary PEM). Microenvironmental monitoring methods include filter-based mass
measurements of particle size fractions (PM,0, PM2 5), usually integrated over either a 24- or 12-h
period. Flow rates vary between various devices from 4 to 20 L/min. Larger sample volumes
allow more extensive chemical characterization to be conducted on microenvironmental samples.
Because more than one pumping system can be used in a microenvironment, PM2 5 and PM10 can
March 2001
5-7
DRAFT-DO NOT QUOTE OR CITE
-------
1 be collected simultaneously. Other continuous ambient PM measurement methods that have
2 been utilized for microenvironmental monitoring are the Tapered Element Oscillating
3 Microbalance (TEOM) and nephelometers. Various continuous techniques for counting particles
4 by size also have also been used (Climet, LASX, SMPS, APS). Measurement techniques are
5 discussed in Chapter 2.
6
7 5 3.2.2 Indirect Methods (Modeling Methods)
8 5.3.2.2.1 Personal Exposure Models
9 Exposure modeling for PM2.5 mass and chemical constituents is a relatively new field
10 facing significant methodological challenges and input data limitations. Exposure models
11 typically use one of two general approaches: (1) a time-series approach that estimates
12 microenvironmental exposures sequentially as individuals go through time or (2) a time-averaged
13 approach that estimates microenvironmental exposures using average microenvironmental
14 concentrations and the total time spent hi each microenvironment. Although the time-series
15 approach to modeling personal exposures provides the appropriate structure for accurately
16 estimating personal exposures (Esmen and Hall, 2000; Mihlan et al., 2000), a time-averaged
17 approach typically is used when the input data needed to support a time-series model are not
18 available. In addition, the time-varying dose profile of an exposed individual can be modeled
19 only by using the time-series approach (McCurdy, 1997, 2000). We define the personal
20 exposure of an individual to a chemical in air to be (NRC, 1991)
21
E=
22 where
23 E is the personal exposure during the time period from t, to t,, and
24 C(t) is the concentration near the nose and mouth not impacted by
25 exhaled air, at time t.
26 In general, personal exposure models combine microenvironmental concentration data with
27 human activity pattern data to estimate personal exposures. Time-averaged models also can be
28 used to estimate personal exposure for an individual or for a defined population. Total personal
March 2001
5-8
DRAFT-DO NOT QUOTE OR CITE
-------
1 exposure models estimate exposures for all of the different microenvironments in which a person
2 spends time, and total average personal exposure is calculated from the sum of these
3 microenvironmental exposures:
(5-1)
4
5 where Ey is the personal exposure in each microenvironment,/ (Duan, 1982). Example
6 microenvironments include outdoors, indoors at home, indoors at work, and in transit. Each
7 microenvironmental exposure, E;, is calculated from the average concentration in
8 microenvironment/, Cj , weighted by the time spent in microenvironment/, t;. T is the sum of t,
9 over all/. It is important to note that, although measurement data may be an average
10 concentration over some time period (i.e., 24 h), significant variations in PM concentrations can
11 occur during that time period. Thus, an error may be introduced if real-time concentrations are
12 highly variable, and an average concentration for a microenvironment is used to estimate
13 exposure when the individual is in that microenvironment for only a fraction of the total time.
14 This exposure formulation has been applied to concentration data in a number of studies (Ott,
15 1984; Ott et al., 1988, 1992; Miller et al., 1998; Klepeis et al., 1994; Lachenmyer and Hidy,
16 2000).
17 Microenvironmental concentrations used in the exposure models can be measured directly
18 or estimated from one or more microenvironmental models. Microenvironmental models vary in
19 complexity, from a simple indoor/outdoor ratio to a multi-compartmental mass-balance model.
20 A discussion of microenvironmental models is presented below in Section 5.3.2.2.2.
21 On the individual level, the time spent in the various microenvironments is obtained from
22 time/activity diaries that are completed by the individual. For population-based estimates, the
23 time spent in various microenvironments is obtained from human activity databases. Many of
24 the largest human activity databases have been consolidated by EPA's National Exposure
25 Research Laboratory (NERL) into one comprehensive database called the Consolidated Human
26 Activity Database (CHAD). CHAD contains over 22,000 person-days of 24-h activity data from
27 11 different human activity pattern studies. Population cohorts with diverse characteristics can
March 2001
5-9
DRAFT-DO NOT QUOTE OR CITE
-------
1 be constructed from the activity data in CHAD and used for exposure analysis and modeling
2 (McCurdy, 2000). Table 5-2 is a summary listing of the human activity studies in CHAD.
3 Methodologically, personal exposure, models can be divided into three general types:
4 (1) statistical models based on empirical data obtained from one or more personal monitoring
5 study, (2) simulation models based upon known or assumed physical relationships, and
6 (3) physical-stochastic models that include Monte Carlo or other techniques to explicitly address
7 variability and uncertainty in model structure and input data (Ryan, 1991; Macintosh et al.,
8 1995). The attributes, strengths, and weaknesses of these model types are discussed by Ryan
9 (1991), National Research Council (1991), Frey and Rhodes (1996), and Ramachandran and
10 Vincent (1999). GIS-based approaches to estimate health risks of environmental concentrations
11 also have been developed (e.g., Beyea and Hatch, 1999; Jensen, 1999). A recent summary
12 review of the logic of exposure modeling is found in Klepeis (1999).
13 Personal exposure models that have been developed for PM are summarized in Table 5-3.
14 The regression-based models (Johnson et al., 2000; Janssen et al., 1997; Janssen et al., 1998a)
15 were developed for a specific purpose (i.e., to account for the observed difference between
16 personal exposure and microenvironmental measurements) and are based on data from a single
17 study, which limits their utility for broader purposes. Other types of models in Table 5-3 were
18 limited by a lack of data for the various model inputs. For example, ambient PM monitoring data
19 is not generally of adequate spatial and temporal resolution for these models. Lurmann and Korc
20 (1994) used site-specific coefficient of haze (COH) information to stochastically develop a time
21 series of 1-h PM10 data from every sixth day 24-h PM10 measurements. A mass-balance model
22 typically was used for indoor microenvironments when sufficient data was available, such as for
23 a residence. For most other microenvironments, indoor/outdoor ratios were used because of the
24 lack of data for a mass-balance model. In addition, only the deterministic model PMEX included
25 estimation of inhaled dose from activity-specific breathing rate information. Data from recent
26 PM personal exposure and microenvironmental measurement studies will help facilitate the
27 development of improved personal exposure models for PM.
28 An integrated human exposure source-to-dose modeling system that will include exposure
29 models to predict population exposures to environmental pollutants such as PM currently is
30 being developed by NERL. A first-generation population exposure model for PM, called the
31 Stochastic Human Exposure and Dose Simulation (SHEDS-PM) model, recently has been
March 2001
5-10
DRAFT-DO NOT QUOTE OR CITE
-------
ati
ce
Is,
Calen
Time
II
i.
D
a.
il
•o „
u~
tl
|£-
,2 o
•3
•s
-
ON
ON
rt of COP
ys; 55 P-D
P-D
OO ^—. ,-x
ON J3 «
^ON ON
_; ON ON
rt ^- ^^
| "3 73
§ S o
1^ S>
5i '^
z1
— «2
¥ 5¥
Akl
Joh
4h St
h Sta
73 S 73 E
§ S
Standard
5
— \o
_ oo
vo —•
r
85
e.^ -Z
rH _L *r? Q.
22 22 <:
2 2*
If II I
s-s
IS
s
!
o.
ati
days
a
c-
*0
•&
-------
1
z
1
1
4.
X
tu
I
GO
X
u
S
oi
UJ
*C3
§
I
-day
§
M
*
;S.8 §
S
00
11111:1
i
Distributioi
population
!l
™J3' § " S
^ S o
.fill
S
I
00
o
March 2001
5-12
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
developed. The SHEDS-PM model uses a 2-stage Monte Carlo sampling technique previously
applied by Macintosh et al. (1995) for benzene exposures. This technique allows for separate
characterization of variability and uncertainty in the model predictions (to predict the distribution
of total exposure to PM for the population of an urban/metropolitan area and to estimate the
contribution of ambient PM to total PM exposure). This model is yet to be evaluated and is
discussed for information purposes only because results from the case study have been only
recently reported in a journal article submitted for peer review (Burke et al., 2001).
5.3.2.2.2 Microenvironmental Models
The mass balance model has been used extensively in exposure analysis to estimate PM
concentrations in indoor microenvironments (Calder, 1957; Sexton and Ryan, 1988; Duan, 1982,
1991; McCurdy, 1995; Johnson, 1995; Klepeis et al., 1995; Dockery and Spengler, 1981; Ott,
1984; Ott et al., 1988, 1992, 2000; Miller et al., 1998; Mage et al., 1999; Wilson et al., 2000).
The mass balance model describes the infiltration of particles from outdoors into the indoor
microenvironment and the generation of particles from indoor sources:
V
= vPCa-vQ-kVCi+Qi,
(5-2)
where
V
Ca
k
volume of the well-mixed indoor air (cubic meters),
concentration of indoor PM;
volumetric air exchange rate between indoors and outdoors (cubic
meters per hour);
penetration ratio, the fraction of ambient (outdoor) PM that is not
removed from ambient air during its entry into the indoor volume;
concentration of PM in the ambient air (micrograms per cubic meter);
removal rate (per hour); and
indoor sources of particles (micrograms per hour).
Qi contains a variety of indoor, particle-generating sources, including combustion or
mechanical processes, condensation of vapors formed by combustion or chemical reaction,
March 2001
5-13
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
suspension from bulk material, and resuspension of previously deposited PM. The removal rate,
k, includes dry deposition to interior surfaces by diffusion, impaction, electrostatic forces, and
gravitational fallout. It may include other removal processes such as filtration by forced air
heating, ventilation, or air-conditioning (HVAC) or by independent air cleaners. All parameters
except V are functions of time. P and k also are functions of particle aerodynamic diameter
andv.
In addition to the mass balance model, a number of single-source or single-
microenvironment models exist. However, most are used to estimate personal exposures to
environmental tobacco smoke (ETS). These models include both empirically based statistical
models and physical models based on first principles; some are time-averaged, whereas others
are time-series. These models evaluate the contribution of ETS to total PM exposure in an
enclosed microenvironment and can be applied as activity-specific components of total personal
exposure models. Examples of ETS-oriented personal exposure models are Klepeis (1999),
Klepeis et al. (1996,2000), Mage and Ott (1996), Ott (1999), Ott et al. (1992, 1995), and
Robinson etal. (1994).
5.3.2.3 Methods of Estimating Personal Exposure to Ambient Particulate Matter
In keeping with the various components of PM exposure described above in Section 5.3.1,
personal exposure to PM can be expressed as the sum of exposure to particles from different
sources summed over all microenvironments in which exposure occurs. Total personal exposure
may be expressed as
Et — Eag + Eig + Epact
Et = Eag + Ei
(5-3)
.nonag,
where Et is the total personal exposure to ambient and nonambient PM, Eag is personal exposure
to ambient PM (the sum of ambient PM while outdoors and ambient PM that has infiltrated
indoors, while indoors), Eig is personal exposure to indoor-generated PM, Epact is personal
exposure to PM from personal activity, and Enonag is personal exposure to nonambient PM.
Although personal exposure to ambient and nonambient PM cannot be measured directly, they
March 2001
5-14
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
can be calculated or estimated from other measurement data. Approaches for estimating these
components of PM exposure are described in the following section.
5.3.2.3.1 Mass Balance Approach
Ambient-Indoor Concentrations of Participate Matter
The mass balance model described above (Equation 5-2) has been used to estimate PM
concentrations in indoor microenvironments. This model also may be used to estimate ambient-
indoor (Cai) and indoor-generated (Cig) PM concentrations. The mass balance model can be
solved for Cai and Cig assuming equilibrium conditions, and assuming that all variables remain
constant (Ott et al., 2000; Dockery and Spengler, 1981; Koutrakis et al., 1992). By substituting
dCai + dCig for dQ in equation 5-2 and assuming dCai and dCig = 0, ambient-indoor PM (Cai) and
indoor-generated PM (Cig), at equilibrium, are given by
(5-4)
15
16
17
18
19
20
21
22
23
24
25
(5-5)
where a = v/V, the number of air exchanges per hour. Equations 5-4 and 5-5 assume equilibrium
conditions and, therefore, are valid only when the parameters k, a, Cao, and Qi are not changing
rapidly and when the Cs are averaged over several hours. Under certain conditions (e.g.,
air-conditioned homes, homes with HVAC or air cleaners that cycle on and off, ambient
pollutants with rapidly varying concentrations), nonequilibrium versions of the mass balance
model (Ott et al., 2000; Freijer and Bloeman, 2000; Isukapalli and Georgopoulos, 2000) are
likely to provide a more accurate estimate of Cai and Cig. However, the equilibrium model
provides a useful, if simplified, example of the basic relationships (Ott et al., 2000).
Equation 5-4 may be rearranged further to give Cai/Cao, the equilibrium fraction of ambient
PM that is found indoors, defined as the infiltration factor (F^) (Dockery and Spengler, 1981).
March 2001
5-15
DRAFT-DO NOT QUOTE OR CITE
-------
Fai
INF = „
Pa
(5-6)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
The penetration ratio (P) and the decay rate (&) can be estimated using a variety techniques.
A discussion of these variables and estimation techniques is given in Section 5.4.3.2.2. Because
both P and k are a function of particle aerodynamic diameter, F^p also will be a function of
particle aerodynamic diameter.
Personal Exposure to Ambient-Generated Participate Matter
Personal exposure to ambient-generated PM (Eag) may be estimated using ambient-indoor
PM concentration (Cai) from the mass balance model, ambient outdoor PM concentrations (Cao)
and information on the time an individual spent hi the various microenvironments.
Mathematically, this may be expressed as
ag
(5-7)
is the fraction of time that an individual spent outdoors, and (1 -y) is the fraction of time
spent indoors.
It is convenient to express personal exposure to ambient generated PM (Eag) as the product
of the ambient PM concentration (Cao or CJ and a personal exposure or attenuation factor.
Following the usage in several recent papers (Zeger et al., 2000; Dominici et al., 2000; Ott et al.,
2000), the symbol a will be used for this attenuation factor. Equation 5-7 can be rearranged to
obtain an expression for a:
' Pa 1
a+k}
(5-8)
March 2001
5-16
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
Substituting equation 5-6 in equation 5-8 gives a relationship for a in terms of the infiltration
factor FJNF and the fraction of time spent in the various microenvironments:
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
-y)
(5-9)
Thus, personal exposures to ambient PM (Eag) may be calculated from measurable quantities:
— OC Cao.
(5-10)
The factor a can be measured directly or calculated from measured or estimated values of the
parameters a, k, and P and the time spent in various microenvironments from activity pattern
diaries (Wilson et al., 2000).
The use of a mass balance model to separate personal exposure into two components
because of exposure to ambient and nonambient concentrations is not novel. This approach,
based on Equation 5-3 as given in Duan (1982) and called superposition of component
concentrations, has been applied using multiple microenvironments to carbon monoxide (Ott,
1984; Ott et al., 1988, 1992), volatile organic compounds (Miller et al., 1998), and particles
(Koutrakis et al., 1992; Klepeis et al., 1994). However, in these studies, and in most of the
exposure literature, the ambient and nonambient components are added to yield a personal
exposure from all sources of the pollutant. The use of the mass balance model, ambient
concentrations, and exposure parameters to estimate exposure to ambient-generated PM and
exposure to indoor-generated PM separately as different classes of exposure has been discussed
in Wilson and Suh (1997) and in Wilson et al. (2000).
5.3.2.3.2 Tracer Species as Surrogates of Ambient-Generated Particulate Matter
The ratio of personal exposure to ambient concentration for a PM component that has no
indoor sources may be used as a measure of the ratio of personal exposure to ambient PM to the
ambient concentration of PM for PM of similar aerodynamic diameter (Wilson et al., 2000).
Sulfate, in particular, often is used as a marker of outdoor air in indoor microenvironments
March 2001
5-17
DRAFT-DO NOT QUOTE OR CITE
-------
1 (Jones et al., 2000). It is found primarily in the PM2 5 fraction of the aerosol (Cohen et al, 2000).
2 Ozkaynak et al. (1996a, b) and Janssen et al. (1999a) report, in the PTEAM and Netherlands
3 studies respectively, that XRF analyses of indoor PM and the immediate outdoor PM show that
4 sulfur is the only element reported with virtually identical mass concentrations in both indoor and
5 outdoor air. Therefore, where there are no indoor sources of fine-mode sulfates, one may deduce
6 that the ambient-to-personal relationship found for sulfates probably would be the same as that
7 for unspeciated particulate matter of the same aerodynamic size range. This assumption has not
8 been validated, however, and ambient PM with different physical or chemical characteristics may
9 not behave similarly to sulfate.
10 Particulate sulfate is formed in the ambient air via photochemical oxidation of gaseous
11 sulfur dioxide arising from the primary emissions from the combustion of fossil fuels containing
12 sulfur. They also arise from the direct emissions of sulfur-containing particles from
13 nonanthropogenic sources (e.g., volcanic activity, wind-blown soil). In the indoor environment,
14 the only common sources of sulfate may be resuspension by human activity of deposited PM
15 containing ammonium sulfates or soil sulfates that were tracked into the home. In some homes
16 an unvented kerosene heater using a high-sulfur fuel may be a major contributor during winter
17 (Leaderer et al., 1999). Use of matches to light cigarettes or gas stoves are also a source of
18 sulfates. Studies that have used sulfate as a surrogate for ambient PM are discussed in
19 Section 5.4.3.1 (i.e., Oglesby et al., 2000a; Sarnat et al., 2000; Ebelt, 2000).
20
21 5.3.2.5.3 Source-Apportionment Techniques
22 Source apportionment techniques provide a method for determining personal exposure to
23 PM from specific sources. If a sufficient number of samples are analyzed with sufficient
24 compositional detail, it is possible to use statistical techniques to derive source category
25 signatures, identify indoor and outdoor source categories, and estimate their contribution to
26 indoor and personal PM. Daily contributions from sources that have no indoor component can
27 be used as tracers to generate exposure to ambient PM of similar aerodynamic size or directly as
28 exposure surrogates in epidemiologic analyses. Studies that have used source-apportionment are
29 discussed in Section 5.4.3.3 (i.e., Ozkaynak and Thurston, 1987; Yakovleva et al., 1999; Mar
30 et al. 2000; Laden et al., 2000).
31
March 2001
5-18
DRAFT-DO NOT QUOTE OR CITE
-------
1 5.4 SUMMARY OF PARTICULATE MATTER MASS DATA
2 5.4.1 Types of Particulate Matter Measurement Studies
3 A variety of field measurement studies have been conducted to quantify personal exposure
4 to PM mass, measure microenvironmental concentrations of PM, and evaluate the relationship
5 between personal exposure to PM and PM air concentrations measured at ambient sites.
6 In general, exposure measurement studies are of two types depending on how the participants are
7 selected for the study. In a probability study, participants are selected using a probability
8 sampling design where every member of the defined population has a known, positive probability
9 of being included into the sample. Probability study results can be used to make statistical
10 inferences about the target population. In & purposeful or nonprobability design, any convenient
11 method may be used to enlist participants and the probability of any individual in the population
12 being included in the sample is unknown. Participants in purposeful samples (also referred to as
13 a "convenience" samples) may not have same the characteristics that would lead to exposure as
14 the general population. Thus, results of purposeful studies apply only to the subjects sampled on
15 the days that they were sampled. In a purposeful study, statistically valid inferences cannot be
16 made to any other population or period of time. Although such studies may report significant
17 differences, confidence intervals, andp values, they have no inferential validity (Lessler and
18 Kalsbeek, 1992). However, most purposeful studies of PM personal exposure can provide data to
19 develop relationships on important exposure factors and useful information for developing and
20 evaluating either statistical or physical/chemical human exposure models.
21 Regardless of the sampling design (probability or purposeful) there are three general
22 categories of study design that can be used to measure personal exposure to PM and evaluate the
23 relationship between personal PM exposure levels and ambient PM concentrations measured
24 simultaneously: (1) longitudinal, (2) daily-average, and (3) pooled. These are discussed in
25 Section 5.4.3.1.1.
26
27 5.4.2 Available Data
28 5.4.2.1 Personal Exposure Data
29 Table 5-4 gives an overview of the personal exposure studies that have been conducted and
30 are reviewed in this section. This includes studies that have been reported since the 1996 AQCD.
March 2001
5-19
DRAFT-DO NOT QUOTE OR CITE
-------
_
H
02
P
O
s>
w
o
C/3
05
6-J
PH
1
S
§
g
o
1
en
"T
W
S
H
8
0
^
^
1 *fl
^ *M
i1
'i
s
S
55
S
cu
•i
l|
S
DU
It
f
£
5
t
t/5
o |
o S1
•0
Study Location an
Population
&
Study Des
en
a ^s S
^5 -a •« g g -s g
OO O •? O cu Z en
2 ^
= a 3"= 2
S S SS 2
cu cu cu cu Cu
0" cu 0" 0
•J* H-T ^T ^-T
cu cC cu oT
Q>
"5*3 *o *o
So o S
cu a. cu a,
CA
-3
S
1
1
(S
• _
js. §
£" ^, g "~T o
^ ^ T3 . Ct 2 J T3 g
"S3 "a> w4_.'rt2'crt
g | i s « 1 | s
II 1 1 1 I 1 1
A
O
z
os o
8S 8
g2 1 igs g g i
5 ^ 555^^55
S S S)SSS*S2S
ft. ftt O-, O-r G« D^ O^ O-t D-
« o o" o"cf o o o
< «r _r _r ,-r i-r »-r *-? <3
cu" &T cu" aT cu" cu" cu* cu* cu*
II g SS •* 2 • « -
0-1 Tf m J2 OO ^O
il S s " 33
g- - | il i
2°* oo o\'^1^ 2T!~ §•
1 If 1 11 Ji II 1 !l
co <^ 3C ^^ v^ ^o o oo o ^o
5 §
•3 t Q £ 1 -
' ill Hti ll II js It if ll
c c G c c c c cj
I a alalal
'ob *ob "So 'ob oo CD 5o DO
33 333333
March 2001
5-20
DRAFT-DO NOT QUOTE OR CITE
-------
G
8
_c
co
1
H
t/5
fjj
s
£3
00
O
td
SONAL]
K
g
PH
H
U
O
fe
53
1^
1
**.
^
Is
0
u
TABLE 5-4
8
P
£
2 «a
— is
o
1
TO
U
2
8
OT
s
a.
"S
|S
Q w
0>
OX)
T3
O
1
>»
1
t/3
°f
6 ^o*
•a
Study Location an
Population
§>
%
Q
t
OT
5
is
ieful Studies (
§.
1
r-
o\
ON
O
"3
u
1
0
s"
Cu
P, A, School
oo
2
o
to
2
f
1
^
Amsterdam and
Wageningen, Neth.,
school children
I
•a
i?
oo ^-^
OS O
c. §,
*t5 »-:
t3 g
4J •*-*
1 1
C?
OT
li
0
Jg"
(X
o "S
s s"
0. 0.
oT cu
S 2
oo
— -H
oo"
*" Os
•* i c"
OS S O.
en — '
Amsterdam, adults
Baltimore, elderly
subjects
c c
I I
^
"3 '
S
1
o1
_
Cu
OT
O
— "
oT
rn
s
o\
oo
Tokyo, Japan,
elderly housewives
c
•3
1
s
ON OO
O\ Os
• "3 —
w «
S "
1 1
Q-
aO
o
OH
0
S
°i 2
OH CM
01
.£
o 1 g" 1
E!E 1
Cw, cu O O
f, ~
It
o! fc
o 3 |
O\ "^"S
S C *o
ca o. c
u- &o 5
*«o o
(N O
8
c 1
1 f.
13
-S
3
I 1
1
ja
TO
HjJ
vimetric measurements.
: indoors, O = outdoors,
Si
8 I
S 8.
March 2001
5-21
DRAFT-DO NOT QUOTE OR CITE
-------
1 Major studies that were reported before that time also have been included to provide a
2 comprehensive evaluation of data in this area. Table 5-4 gives information on the sampling and
3 study designs, the study population, the season, number of participants, PM exposure metric, and
4 the PM size fraction measured.
5 Although there are a number of studies listed in the table, the data available to answer the
6 important questions related to exposure are limited. Few are based on probability sampling
7 designs that allow study results to be inferred to the general population. Unfortunately, none of
8 these probability studies uses a longitudinal study design. This limits our ability to provide
9 population estimates on the relationship between personal PM exposures and ambient site
10 measurements. In addition, most of the probability studies of PM exposure were conducted
11 during a single season, thus variations in ambient concentrations, air exchange rates, and
12 personal activities are not accounted for across seasons. In these cases, study results are only
13 applicable to a specific time period. Longitudinal studies, on the other hand, generally have
14 small sample sizes and use a purposeful sampling design. Many of these studies did not include
15 ambient site measurements to allow comparisons with the exposure data, and approximately half
16 of these studies monitored PM25.
17 Four large-scale probability studies that quantify personal exposure to PM under normal
18 ambient source conditions have been reported in the literature. These include the EPA's Particle
19 Total Exposure Assessment Methodology (PTEAM) study (Clayton et al., 1993; Ozkaynak et al.,
20 1996a,b); the Toronto, Ontario, study (Clayton et al., 1999a and Pellizzari et al., 1999); the Air
21 Pollution Exposure Distribution within Adult Urban Populations in Europe (EXPOLIS) exposure
22 study (Jantunen et al., 1998, 2000; Oglesby, et al., 2000); and a study of a small, highly polluted,
23 area in Mexico City (Santos-Burgoa et al., 1998). Only preliminary results have been reported
24 for the EXPOLIS study. A fifth study conducted in Kuwait during the last days of the oil-well
25 fires (Al-Raheem et al., 2000) is not reported here because the ambient PM levels were not
26 representative of normal ambient source conditions.
27 Recent longitudinal exposure studies have focused on potentially susceptible
28 subpopulations such as the elderly with preexisting respiratory and heart diseases (hypertension,
29 chronic obstructive pulmonary disease, and congestive heart disease). This is in keeping with air
30 pollution analyses that indicate mortality associated with high levels of ambient PM2 5 is greatest
31 for elderly people with cardiopulmonary disease (U.S. Environmental Protection Agency, 1996).
March 2001
5-22
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Longitudinal studies were conducted in the Netherlands by Janssen (1998) and Jassen et al.
(1997, 1998a,b, 1999a,b) on purposefully selected samples of adults (50 to 70 years old) and
children (10 to 12 years old). Several additional studies have focused on nonsmoking elderly
populations in Amsterdam and Helsinki (Janssen et al., 2000), Tokyo (Tamura et al., 1996a),
Baltimore (Liao et al., 1999; Williams et-al., 2000a,b,c), and Fresno, CA (Evans et al. 2000).
These cohorts were selected because of the low incidence of indoor sources of PM (such as
combustion or cooking). This should allow an examination of the relationship between personal
and ambient PM concentrations without the large influences caused by smoking, cooking, and
other indoor particle-generating activities. The EPA has a research program focused on
understanding PM exposure characteristics and relationships. Within the program, longitudinal
studies are being conducted on elderly participants with underlying heart and lung disease
(COPD, patients with cardiac defibrillator, and myocardial infarction), an elderly environmental
justice cohort, and asthmatics. These studies are being conducted in several cities throughout the
United States and over several seasons. Only preliminary data are currently available, and results
are not reported in this document.
A series of studies by Phillips et al. (1994, 1996, 1997a,b, 1998a,b, 1999) examined
personal ETS exposure in several European cities. Participants varied by age and occupation.
Respirable Particulate Matter (RSP) concentrations were reported. These studies are not
included in Table 5-4 because of their focus on ETS exposure, which is not the focus of this
chapter. A small personal exposure study in Zurich, Switzerland, was reported by Monn et al.,
(1997) for PM10. This study also is not listed in Table 5-4 because indoor and outdoor
measurements were not taken simultaneously with the personal measurements, and other details
of the study were not published.
5.4.2.2 Microenvironmental Data
Usually, personal PM monitoring is conducted using integrated measurements over a 12- or
24-h period. As such, total PM exposure estimates based on PEM measurements do not capture
data from individual microenvironments. Recent studies have examined PM concentrations in
various microenvironments using a number of different types of instruments ranging from filter-
based to continuous particle monitors. Details on the instruments used, measurements collected,
and findings of these studies according to microenvironment (residential indoor, nonresidential
March 2001
5-23
DRAFT-DO NOT QUOTE OR CITE
-------
1 indoor, and traffic-related) are summarized in Table 5-5. Studies that collected
2 microenvironmental data as part of a personal exposure monitoring study are summarized in
3 Table 5-4. In general, the studies listed in Table 5-5 are relatively small, purposeful studies
4 designed to provide specific data on the factors that effect microenvironmental concentration of
5 PM from both ambient and nonambient sources.
6 Recently published studies have used various types of continuous monitors to examine
7 particle concentrations in specific microenvironments and resulting from specific activities.
8 Continuous particle monitors such as the SMPS, APS, and Climet have been used to measure
9 particle size distributions in residential microenvironments (Abt et al., 2000a; Long et al., 2000a;
10 Wallace et al., 1997; Wallace, 2000a; McBride et al., 1999; Vette et al., 2001). These studies
11 have been able to assess penetration efficiency for ambient particles and microenvironments
12 indoors as well as penetration factors and deposition rates. Continuous instruments are also a
13 valuable tool for assessing the impact of particle resuspension caused by human activity.
14 A semi-quantitative estimate of PM exposure can be obtained using personal nephelometers that
15 measure PM using light-scattering techniques. Recent PM exposure studies have used personal
16 nephelometers (1 min avg time) to measure PM continuously (Howard-Reed et al., 2000;
17 Quintana et al., 2000) in various microenvironments. These data have been used to identify the
18 most important ambient and nonambient sources of PM, to provide an estimate of source
19 strength, and to compare modeled time activity data and PEM 24-h mass data to nephelometer
20 measurements (Rea et al., 2001). Several studies also have examined PM exposure in vehicles
21 using both continuous and filter-based techniques.
22
23 5.4.2.3 Interpretation of Participate Matter Exposure Data
24 Papers that have reanalyzed and interpreted the data collected in previous PM exposure
25 studies are summarized in Table 5-6. These analyses are directed towards understanding the
26 personal cloud, the variability in total PM exposure, and the personal exposure-to-ambient
27 concentration relationships for PM. Results are highlighted here and given in more detail in
28 Section 5.4.3. Brown and Paxton (1998) determined that the high variability in personal
29 exposure to PM makes the personal-to-ambient PM relationship difficult to predict. Wallace
30 (2000b) used data from a number of studies to test two hypotheses: elderly COPD patients have
31 (1) smaller personal clouds and (2) higher correlations between personal exposure and ambient
March 2001
5-24
DRAFT-DO NOT QUOTE OR CITE
-------
c»
fi
5ASUREMENT STUD
ROENVIRONMENTAL MI
Q.
^
H
r^
(s
L
Pi
Pi
5
&
^
g
v!
i
m
|
H
a
Notes/Findii
i
•s
fr
C3
00
0
l|
UH
'tn'
i
g
J
g
'a* H.
l|
Q
1
1
2 h'')
GO •<
1°"»
o § J?
CN CN VO
1?
g
§1
*e3 ^
II
s
|o S
•H '5 w
1 M § J
Sources of fine particles: cookinj
Sources of coarse particles: cook
activities. 50% of particles by vol
indoor events were ultrafine partii
Continuous PM distributions and size
distributions obtained for indoor and
outdoor air
o
a?
o\
o%
^<
•a >
t; ~-
§J
JQ
W
•s
.™
•£
e ®
5|
*l
il
hidoor PM concentrations were k
conditioning (AC) than non-air-ci
24 h mean: Regional air 20.2 ±9.9 n^rtf
;n = 50); Outdoor homes 21. 8 ± 14.8 fig/m3
;n = 43); Indoor w/ AC 18.7 ± 13.2 /ig/m3
;n = 49); Indoor w/o AC 2 1 . 1 ±7.5 ^g/m3
;n = 9)
o
$ i
1 ~^
Is*
si^
g^-g
u 73 t>
ll|
§ l •,?
0.3- to 0.5-/im particles linked to
frying, broiling; 0.5- to 2.5-yum pj
cooking events; >2.5-i*m particle!
movement.
Fime activity data, whole-house air
;xchange rates
Continuous carbon monoxide: descriptive
lata for monitored pollutants; size profiles
"or six indoor particle sources
«•
1 1 1 _,_ |
*tq § S O Tf
vo _• o O O
g
•S
cS
•*-* u
§
u «
M§
II
•o 8 J1
|'5-g
•S£ 1
a o «
« -o "2
•o « §
111
Time-series plots of personal nepl
that each participant's PM exposi
of short-term peaks, imposed on a
ambient PM concentrations.
Continuous (15-min avg) PM and time
ictivity data; 24-h PM mass; participants
rom Baltimore and Fresno PM panel
itudies.
Descriptive statistics from each study for
we microenvironments
o
— «n
O CN
S
"^ § x-v
•I*!!! w
£-• -ZZs *Z CU
OT
"c
>M
"o
ra
O.
•/^
|
§.
•a
s a
**? S
^02
1 = 1
111
"8
^i ^
i^ l-s
O CO o ^3
8~ g-C
(D C S °
2 S'^s'Sl
as IT
K S £ •*
54 ± 3 1 % of average daily PM2.5 e
indoor residences, where participz
their time. A significant portion o
occurred where participants spent
Continuous (15-min avg) PM and time
ictivity data; 24-h PM mass;
vlodeled PM mass and time activity data to
ipportion time spent in a location. Good
:omparison with nephelometer mass (6-
!0%)
o
s i
—* *n
O
-------
H
CO
§
^
«?
«n
S
I
.Si
'&
D-. t,
11
•i-81
•a* §
2S2SS
O OO
'
I
.5
•a!.
II
^ o
22
11
•s
— o. o. «
•» o
ea oo
if
o-^
1°
S.S
a
'3;
0-.
I
3 2 0
M flj r-~.
"
112
-4
^ r. ••
"e 713
3. "5*00
o «> *o
111
«§<§•§
•S « !
cu a, i
^3
I:
09
Q e i
-a
i-liil
i 111 E
o
o
ci
15 O
* o
u c
33 s
> £
8>1
§ O
5 I
11
•§"=
^ J
(U ft,
•z:
cd w
I
•§
CQ
March 2001
5-26
DRAFT-DO NOT QUOTE OR CITE
-------
fc
Pi
tt
pt
B
§
«3
I
3
•o
1
!
M
1
t»
f
Di £
a-g S-a
lill
Z ^ c o.
o ^
§,0
S|
55 .§
.tg2
< S
_• 8
M .g
•S g
(GQ C3
'53 [i,
tcj
&o
73
g i
i§
II
M
§i
t3 3
£ O
"e4
4 ?
V
E1?
2
lg
II
'}•>
< i
V ^ "S»
1 O e
ts*
S3
a oa
i c
II
11
5 s.s
§.g .a
•g i •s
F
I O
- CQ
•O b"
2<^ §
co c->
21
§ 00
Jl
March 2001
5-27
DRAFT-DO NOT QUOTE OR CITE
-------
CO
W
CO
H
Z
ps
Z
O
P
V)
5
H
°!S
gs
II
iSS
3 1
IS-§
h
SB
sS
> m
• QJ
I 2
73 S)
o o
•s s
•^•n
•3
•£%• S .1
1 E.E 6
11
ill
II
oo g .2
-Ni
IS
•- &
JS 5
E
S
o S
U1 .g
^
^.-^
8
(8 ^
M . .
S g
a. > «
ed vj
4 81
2.1^
II"
> > (N
-5-5^
* U *—' ^ *"*
|3l^
5 09 — w oo
§
o
00 I
•g
§11
•g -S s?
§1
I §
March 2001
5-28
DRAFT-DO NOT QUOTE OR CITE
-------
h-
O
w
g
5
A
>
ta
PH
O
g
£
e
ci
c£
vo
•A
S
£ §
4s |.
-c S e
S g
% M S -O -2
Jsl?*8.!
l^llpl
111^1
Pltltfe"
ill=a?g:
1§^11211
±: »vo «,£:—• cxu
k et al,
r 1988
o 2 .
":•_- -3
-3f «
ioy etal.,
COPD (B
m COPD (
__ o\ ^-*
•S
AM (
., 1989
., 1990
ES(Li
ville
sterd
1
Bos
1
1
I
•§
.s«r
"S o
"^-l<^
o-SZ^-
•-^ IK ta «
^l§l
I|S| ffi
Sg s s =~
f- O. M CO
e
I
la
CQ ^,
^B
March 2001
5-29
DRAFT-DO NOT QUOTE OR CITE
-------
1 concentrations, compared to healthy elderly, children, and the general population. The analysis
2 by Wallace (2000a) and three subsequent longitudinal studies (Williams 2000a,b,c; Ebelt et al.,
3 2000; Samat et al., 2000) support hypothesis 1 but not hypothesis 2. Ozkaynak and Sperigler
4 (1996) show that at least 50% of personal PM10 exposure for the general population is because of
5 ambient particles that significantly contribute to inhaled particles. Wilson and Suh (1997)
6 conclude that fine and coarse particles should be treated as separate classes of pollutants because
7 of differences in characteristics and potential health effects. Wilson et al. (2000) give a review of
8 what they call the "exposure paradox" and determine that personal PM needs to be divided into
9 different classes according to source type, and that correlations between personal and ambient
10 PM will be higher when nonambient sources of PM are removed from the personal PM
11 concentration. Mage (1998) conducted analysis using the PTEAM data and showed that on
12 average a person is exposed to >75% of ambient PM2 5 and >64% of ambient PM10. Mage et al.
13 (1999) use an algorithm to fill in missing data and outliers to analyzed data sets and show that
14 variation in daily personal exposures for subjects with similar activity patterns and no ETS
15 exposure are driven by variation in ambient PM concentrations.
16
17 5.4.3 Factors Influencing and Key Findings on Particulate Matter Exposures
18 5.4.3.1 Correlations of Personal/Microenvironmental Particulate Matter with Ambient
19 Particulate Matter
20 The relationship between measured personal PM exposure and PM concentrations
21 measured at ambient sites has been of interest to exposure analysts. Many of the studies,
22 summarized above in Table 5-4, have analyzed this relationship using measurements of personal
23 PM exposures and ambient PM concentrations. The statistical correlation between these
24 measurements for the various personal exposure studies is discussed in this section. .
25
26 5.4.3.1.1 Types of Correlations
27 The three types of correlation data that will be discussed in this section are longitudinal,
28 "pooled", and daily-average correlations. Longitudinal correlations are calculated when data
29 from a study includes measurements over multiple days for each subject (longitudinal study
30 design). Longitudinal correlations describe the temporal relationship between daily personal PM
31 exposure and daily ambient PM concentration for each individual subject. The longitudinal
March 2001
5-30
DRAFT-DO NOT QUOTE OR CITE
-------
1 correlation coefficient, r, may differ for each subject, and an analysis of the variability in r across
2 subjects can be performed with this type of data. Typically, the median r is reported along with
3 the range across subjects in the study. Pooled correlations are calculated when a study involves
4 one or only a few measurements per subject and different subjects are studied on subsequent days
5 (sometimes called a "cross-sectional" study design). The different subject/different day data are
6 combined, pooled, for the correlation calculation. Pooled correlations describe the relationship
7 between daily personal PM exposure and daily ambient PM concentration across all subjects in
8 the study. This type of correlation is sometimes called cross-sectional, but will be called pooled
9 in this chapter because only a limited number of participants are monitored on any given day.
10 Daily-average correlations are calculated using the average exposure across subjects for each
11 day. Daily-average correlations describe the relationship between daily community-averaged
12 personal PM exposure and daily ambient PM concentration. This type of correlation could be
13 called cross-sectional, but given that the pooled correlation also is referred to as cross-sectional,
14 the term daily average is used here.
15 Studies that have reported longitudinal correlations also typically have reported pooled
16 correlations. However, pooling of the data for the correlation has been handled differently across
17 the various studies. For some studies, the multiple days of measurements for each subject were
18 assumed to be independent (after autocorrelation and sensitivity analysis) and combined together
19 in the correlation calculation (Ebelt et al., 2000). In other studies, daily averages across subjects
20 were calculated and the correlation determined from the daily averages (Williams et al., 2000b).
21 A third approach also was used in other studies to simulate a cross-sectional study design
22 (Janssen et al., 1997, 1998a, 1999c). In this approach, a random-sampling procedure was used to
23 select a random day from each subject's measurements to use for the correlation. This procedure
24 was repeated many times, and statistics such as the mean and standard deviation of the pooled
25 correlation coefficient were reported.
26 The type of correlation analysis can have a substantial effect on the resulting correlation
27 coefficient. Mage et al. (1999) mathematically demonstrated that very low correlations between
28 personal exposure and ambient concentrations could be obtained when people with very different
29 nonambient exposures are pooled, even though their individual longitudinal correlations are high.
30 The longitudinal studies conducted by Tamura et al. (1996a) and Janssen et al. (1997, 1998a,
31 1999c) determined that the longitudinal correlations between personal exposure and ambient PM
March 2001
5-31
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
concentrations were much higher than the correlations obtained from a pooled data set. Wallace
(2000a) reviewed a number of longitudinal studies found that the median longitudinal correlation
coefficient was much higher than the pooled correlation coefficient for the same data (see
Tables 1 and 2 Wallace, 2000a). Williams et al. (2000a,b) and Evans et al. (2000) reported
higher correlation coefficients for daily-average correlations compared to longitudinal
correlations.
5.4.3.1.2 Correlation Data from Personal Exposure Studies
Measurement data and correlation coefficients for the personal exposure studies described
hi Section 5.4.2.1 are summarized in Table 5-7. All data are based on mass measurements. The
studies are grouped by the type of study design, longitudinal or pooled. For each study in
Table 5-7, summary statistics for the total personal PM exposure measurements are presented,
as well as statistics for residential indoor, residential outdoor, and ambient PM concentrations,
where available. The correlation coefficient (r) between total personal PM exposures and
ambient PM concentrations also are presented and classified as longitudinal or pooled
correlations. When reported,/?-values for the correlation coefficients are included. Correlation
coefficients between personal, indoor, outdoor, and ambient also are reported, when available.
5.4.3.1.3 Correlations Between Personal Exposures, Indoor, Outdoor, and Ambient
Measurements
Longitudinal and pooled correlations between personal exposure and ambient or outdoor
PM concentrations varied considerably between study and study subjects. Most studies report
longitudinal correlation coefficients that range from <0 to ~ 1, indicating that an individual's
activities and residence type may have a significant effect on total personal exposure to PM.
General population studies tend to show lower correlations because of the higher variation in the
levels of PM generating activities. In contrast, the absence of indoor sources for the populations
in several of the longitudinal studies resulted in high correlations between personal exposure and
ambient PM within subjects over time for these populations. But even for these studies,
correlations varied by individual, depending on then: activities and the microenvironments that
they occupied.
March 2001
5-32
DRAFT-DO NOT QUOTE OR CITE
-------
>
£?H
0
HH
3
H
1
0
U
Q
9
1
..
&H
H H
^ij fr.
g w
w o
i-l H
^ o
PARTIC
TION C<
g|
PY- {•••
co 5
S5
P Q
co 5
o
(2j
S
o
H
g
0
§
1
CO
PS
P*
t-"
uo
J
PH
H
^
CO
.1
'3
0
1
a
£
... r*
11
fi
< o
.2 U
rt -.
c §
OH 2^
U
1="
•a
1
J3
1
i
§
•a
i
S
S
"a
§
I
1
z
x-v
0>
CO
c
s»- •*
1
>
^
ft
1
1
*M ^
-a "C3
1°
"3
18
*a *a
"S -^
c^
_ ,
1
(§
«
1^
o
"^
1
fl
<£=
g
"1
.2
1
"«
_g
1
I
0?C?
99
O oo
— •" OO
•**• O Tl O fvj rO
O O O O O O
JJ
X t I I * *
cL-or cuSc S S
a a a
8 g 1 a 1 a S 1 2
I ft 10 P a! 1 f I
8 ,| | 111 ll 1| 111
I II """!! < «> «"iao Ido louQ Q
Sft/l^Q 2P S'^CflS1 t2o5°° I^OOOOOT^W^1
J J: 'f ? "f
§ S f 'S S f S Is 1 S • S S
1 §• 5 J r
"3 2 u o S
1 # 1 # i i 1 i I i^ s3 ss
^ CU > CU CUcCU gCU R P-i P-i fU
March 2001
5-33
DRAFT-DO NOT QUOTE OR CITE
-------
Q
g
5
NATIONS
1
Z
&M
o
z
o
u
o
P-")
in
sl
il
co O
S U
fi 5
S^
Sr H
M S
MONITOK
CORI
M
^j*
0
CO
'"•^
AH
o
a
0
~—^
f^
V)
W
S
H
a
.1
& *5
j3 -P
O c
1
S
„ 13
11
lo ts
Personal-Am
orrelation Coef
o
ion Levels O'g/m3)
S
c
o
1
*o
s
s
•eT
U)
&_
0
[ra
^
1
S
z
1
¥
1
1
^D
1
1 h
Ii
"rt
11
1
"3
S
cu
"|7§
n CO
.£2
1
s
U)
oi|
r-
CM * r^ '
en Os — CM
00 SO
rf en
-H u-j ~H CM
OO OO so —
^ t ^1
en m CN —
Si CM
§
O *> ^** Q u
-H J * -H J
^
^" "3 CM
CM ™ —
a
293
S S S
1* fc
pfe
o o
JJ
•a -d
SS
(0.0-0.72)'
3
Median L
r- .
oo
CM r*
CM
•n
en CM
oo
rf
-H «
CM CM
r^ '
en OS
s
Q «*
co S1
-H §
^
CM
o
^ S
O O
JJ
•a -d
U U
b S3 -
u i | E
S o3 § s ^
E .S f. ~. «
77-S^ *i
c c 12 S .3 >
C- ON OO
<*. 99
0 — 00
S Tl
s. 2.S
O VO u^ OS wi O *f
fn f- CN oo r-- «n TT
o o o o o o o
_] -J -1
.9 .1 .i
S ' S 2 a. a. a. a.
2 §
jch
oo
^
^r ^
3S 25
-H — -H -H
NO ' 1^ to
•X « NO* 06
— T (S —
IN " «
a «> ,. Q Q
tn "" £ ») en
-H c*j ° 41 -H
is
S
rsi 0^3-
2
J i ~f
.£ J 0-
^ erf ^ od
Sz Sz
CO ^ CO ^
ff ft
9' ° ^ ^
i-E 11
SK =S
o o" o o
J _] -J-J
ii ii
CM Ji ^ 3
en t"* 06 T
" i2 o os
-H -H "* ^
en oo ^ ^
en CM r~- os
KS £S
a a a a
-H -H -H -H
IX IX IX IX
J. J.
-------
co
§
»
s
,—,
CA
g
1
| 8
.0
1
3
'^^
~ 2
|j
2 3
in '-C
""3
£~-
S
^
•3
«j
S
1
I
1
S
e
a
£
:>
"l
1
"g
S
"u
^s w
ii
.2 "3
(2°
*rt
1 0
-------
Q
Z
^t
1
CH^
S
H
W
U
Z
o
u
Q
w
5
^J<
W }
^•J p"^
. . z
5 H
OH U
O fa
fy[ jy
OJ O
w
^^ ^^1
^J ^1
f-t H
CO ^
^ T
61
5ONALMON
(
K
PM
"S
q
V)
H
,— .
J
o
S3 g
sl
o
I
^-.
^
"11
< s
1 i
Jj-g
o
U
I
1 Concentration Levels
1
M
c
£
3
"3
1
1
"M
J
"l
^
»
|j
—
11
1°
is
s-g
I-5
c
g
cu
«
1"
s
on
j>l
8*5ra
o, v *2 S
.2 II II II II II II S I o ~ "3 § g
> a o s ra o •" o«
a.ti gwUwneuOw
^>;O> h- ^i«r4rn^*tovovo
g
2
f~
.
"2
_g
cd
B!
fe
^
March 2001
5-36
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Probability Studies
In the Toronto study (Pellizzari et al., 1999), pooled correlations were derived for personal,
indoor, outdoor, and fixed site ambient measurements. This study was conducted in Toronto on
a probability sample of 732 participants who represented the general population, 16 years and
older. The study included between 185 and 203 monitoring periods with usable PM data for
personal, residential indoor, and outdoor measurements. For PM10, measurements, the mean
concentrations were 67.9 /ug/m3 for personal, 29.8 /J-g/m? for indoor air, and 24.3 ,ug/m3 for
outdoor air samples. For PM2 5, the mean concentrations were 28.4 //g/m3 for personal,
21.1 yUg/m3 for indoor air, and 15.1 yug/m3 for outdoor air samples. A low but significant
correlation (r = 0.23, p < 0.01) was reported between personal exposure and ambient
measurements. The correlations between indoor concentrations and the various outdoor
measurements of PM25 ranged from 0.21 to 0.33. The highest correlations were for outdoor
measurements at the residences with the ambient measurements made at the roof site (0.88) and
the other fixed site (0.82). Pellizzari et al. (1999) state that much of the difference among the
data for personal/indoor/outdoor PM
... can be attributed to tobacco smoking, since all variables reflecting smoking... were found to be
highly correlated with the personal (and indoor) particulate matter levels, relative to other variables that
were measured... none of the outdoor concentration data types (residential or otherwise) can
adequately predict personal exposures to particulate matter, (p. 729)
Santos-Burgoa et al. (1998) describe a 1992 study of personal exposures and indoor
concentrations to a randomly sampled population near Mexico City. The sample of 66 monitored
subjects included children, students, office and industrial workers, and housewives. None of the
people monitored were more than 65 years old. The mean 24-h personal exposure and indoor
concentrations were 97 ± 44 (SD) and 99 ± 50 ^g '3, respectively, with an rPersonal/Ambient = 0.26
(p = 0.099). Other correlations of interest were rPeisonal/Indoor = 0.47 (p = 0.002) and rIndoor/Ambient =
0.23 (p = 0.158). A strong statistical association was found between personal exposure and
socioeconomic class (p = 0.047) and a composite index of indoor sources at the home
(p = 0.039).
March 2001
5-37
DRAFT-DO NOT QUOTE OR CITE
-------
1 Correlation analysis for personal exposure has not yet been reported for EXPOLIS. Some
2 preliminary results (Jantunen et al., 2000) show that in Basel and Helsinki, a single ambient
3 monitoring station was sufficient to characterize the ambient PM2 5 concentration in each city.
4 Using microenvironmental concentration data collected while the subjects were at home, at work,
5 and outdoors, they calculated the sum of the time-weighted-averages of these data and found the
6 results closely match the personal PM2 5 exposure data collected by the monitors carried by most
7 of the subjects, with a few subjects, mostly smokers, being noticeable exceptions.
8
9 Longitudinal Studies
10 A number of longitudinal studies using a purposeful sampling design have been conducted
11 and reported in the literature since 1996. A number of these studies (Janssen et al., 1998a,
12 1999b, 2000; Williams et al., 2000b; Evans et al., 2000) support the previous work by Janssen
13 et al. (1995) and Tamura et al. (1996a) and demonstrate that, for individuals with little exposure
14 to nonambient sources of PM, correlations between total PM exposure and ambient PM
15 measurements are high. Other studies (Ebelt et al., 2000; Samat et al., 2000) show strong
16 correlations for the SO4"2 component of PM2 5 but poorer correlations for PM2 5 mass. Still other
17 studies show only weak correlations (Rojas-Bracho et al., 2000; Linn et al., 1999; Bahadori et al.,
18 2001). Even when strong longitudinal correlations are demonstrated for individuals in a study,
19 the variety of living conditions may lead to variations in the fraction of ambient PM contributing
20 to personal exposure. Groups with similar living conditions, especially if measurements are
21 conducted during one season, may have similar a and, therefore, very high correlations between
22 personal exposure and ambient concentrations. However, when a panel contains subjects with
23 homes of very different ventilation characteristics or covers more than one season, variations in a
24 can be high across subjects.
25 Elderly Subjects. Janssen et al. (2000) continued their longitudinal studies with
26 measurements of personal, indoor, and outdoor concentrations of PM25 for elderly subjects with
27 doctor-diagnosed angina pectoris or coronary heart disease. Studies were conducted in
28 Amsterdam and Helsinki, Finland, in the winter and spring of 1998 and 1999. In the Amsterdam
29 study, with 338 to 417 observations, the mean concentrations were 24.3, 28.6, and 20.6 //g/m3 for
30 personal, indoor, and outdoor samples, respectively. If the measurements with ETS in the home
31 were excluded, the mean indoor concentration dropped to 16 Aig/m3, which was lower than
March 2001
5-38
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
outdoors. In the Helsinki study, the mean PM2 5 concentrations were 10.8 yug/m3 for personal,
11.0 //g/m3 for indoor air, and 12.6 yug/m3 outdoor air samples. The authors note that for this
group of subjects, personal exposure, indoor concentrations, and ambient concentrations of PM25
were highly correlated within subjects over time. Median Pearson's correlation coefficients
between personal exposure and outdoor concentrations were 0.79 in Amsterdam and 0.76 in
Helsinki. The median Pearson's r for the indoor/outdoor relationship was 0.85 for the
Amsterdam study, excluding homes with ETS. The correlation for indoors versus outdoors was
0.70 for all homes.
A series of PM personal monitoring studies involving elderly subjects was conducted in
Baltimore County, MD, and Fresno, CA. The first study was a 17-day pilot (January-February
1997) to investigate daily personal and indoor PM15 concentrations, and outdoor PM2 5 and
PM2.s-io concentrations experienced by nonsmoking elderly residents of a retirement community
located near Baltimore (Liao et al, 1999; Williams et al., 2000c). The 26 residents were aged
65 to 89 (mean = 81), and 69% of them reported a medical condition, such as hypertension or
coronary heart disease. In addition, they were quite sedentary; less than 5 h day"1, on average,
was spent on ambulatory activities. Because most of the residents ate meals in a communal
dining area, the average daily cooking time in the individual apartments was only 0.5 h (range 0
to 4.5 h). About 96% of the residents' time was spent indoors (Williams et al., 2000c). Personal
monitoring, conducted for five subjects, yielded longitudinal correlation coefficients between
ambient concentrations and personal exposure ranging from 0.00 to 0.90.
Subjects with COPD. Linn et al. (1999) describe a 4-day longitudinal assessment of
personal PM2 5 and PMIO exposures (on alternate days) in 30 COPD subjects aged 56 to 83;
concurrent indoor and outdoor monitoring were conducted at their residences. This study
occurred in the summer and autumn of 1996 in the Los Angeles area. PM10 data from the nearest
fixed-site monitoring station to each residence also was obtained. Pooled correlations for
personal exposure to outdoor measurements were 0.26 and 0.22 for PM2 5 and PM10, respectively.
Day-to-day changes in PM2 5 and PM10 measured outside the homes tracked concurrent PM10
measurements at the nearest ambient monitoring location, with R2 values of 0.22 and 0.44,
respectively. Day to day changes in PM mass measured indoors also tracked outdoors at the
homes with R2 values of 0.27 and 0.19 for PM10 and PM25 respectively.
March 2001
5-39
DRAFT-DO NOT QUOTE OR CITE
-------
1 Personal, indoor, and outdoor PM2 5? PM,0i and PM2.5.10 correlations were reported by
2 Rojas-Bracho et al. (2000) for a study conducted in Boston, MA, on 18 individuals with COPD.
3 Both the mean and median personal exposure concentrations were higher than the indoor
4 concentrations, which were higher than outdoor concentrations for all three PM measurement
5 parameters. Geometric mean indoor/outdoor ratios were 1.4 ± 1.9 for PM10, 1.3 ± 1.8 for PM2 5,
6 and 1.5 ± 2.7 for PM2.s.10. Median longitudinal R2s between personal exposure and ambient PM
7 measurements were 0.12 for PM10, 0.37 for PM25 and 0.07 for PM2.5.I0. The relationship between
8 the indoor and outdoor concentrations was strongest for PM25, with a median R2 of 0.55 and
9 11 homes having significant R2 values. For PM10 the median R2 value was 0.25, with significant
10 values for eight homes. Only five homes had significant indoor/outdoor associations for PM2 5.10,
11 with an insignificant median R2 value of 0.04.
12 Bahadori et al. (2001) report a pilot study of the PM exposure of 10 nonrandomly chosen
13 chronic obstructive pulmonary disease (COPD) patients in Nashville, TN, during the summer of
14 1995. Each subject alternately carried a personal PM25 or PM10 monitor for a 12-h daytime
15 period (8 a.m. to 8 p.m.) for 6 consecutive days. These same pollutants were monitored
16 simultaneously indoors and outdoors at their homes. All of the homes were air-conditioned and
17 had low air exchange rates (mean = 0.57 hf1), which may have contributed to the finding that
18 mean indoor PMZ5 was 66% of the mean ambient PM2 5. This can be contrasted with the
19 PTEAM study in Riverside, CA, where no air conditioners were in use and the mean indoor
20 PM2-S was 98% of the mean ambient PM2.5 (Clayton et al., 1993). Data sets were pooled for
21 correlation analysis. Resulting pooled correlations between personal and outdoor concentrations
22 were r= 0.09 for PM2.5 and r=-0.08 for PM10.
23
24 5.4.3.1.4 A Correlation Between a Daily-Average Exposure and Ambient Concentrations
25 A recent biostatistical analysis (Zeger et al., 2000) suggests that the community mean
26 exposure is the appropriate parameter for analyzing exposure error in community time-series
27 epidemiology. Ott et al. (2000) suggest that the correlation of the community mean exposure
28 with ambient concentrations will approach 1.0 for a large community and demonstrated this
29 using data from the PTEAM study. Mage et al. (1999) calculated the daily-average exposure for
30 three earlier studies with sufficient data and found that the coefficients for the correlation of daily
March 2001
5-40
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
averages with the ambient concentrations were high (R2 = 0.9). Two new studies have obtained
sufficient data to calculate a daily-average correlation.
The 1997 Baltimore pilot study was followed up in July and August 1998 by a more
extensive study in which identical samplers were used for personal, indoor, and outdoor
measurements. The participants, aged 72 to 93 (mean 81) years, included healthy members, as
well as subjects with COPD and cardiovascular disease. The participants lived in an 18-story
retirement facility that provided a self-contained living environment. There was a central HVAC
system for common areas but each apartment had an individually controlled HVAC system. The
subjects had limited exposures to indoor-generated sources of PM because of their low
frequency/duration of activities like cooking, cleaning, or interacting with tobacco smokers
(Williams et al., 2000a,b). As a result, the daily-average correlation coefficient was very high
(r = 0.89) between personal exposure and ambient concentrations of PM2 5. Median longitudinal
correlations were also high (r = 0.81; range = 0.38 to 0.98).
Evans et al. (2000) report two panel studies in Fresno, CA, with daily-average correlation
coefficients of 0.41 and 0.84.
5.4.3.1.5 Correlations Using Sulfate as a Surrogate for Personal Exposure to Ambient
Particulate Matter
A study, conducted in Vancouver, involving sixteen COPD patients aged 54 to 86, reported
low median longitudinal (r = 0.48) and pooled (r = 0.15) correlation coefficients between
personal exposure and ambient concentrations of PM25 (Ebelt et al., 2000). However, the
correlation between personal exposure and ambient concentrations of SO42" was much higher.
The results for PM2 5 and sulfate are compared in Figure 5-1. Ebelt et al. (2000) conclude the
following.
We found SO42' to be a good measure of exposure to accumulation mode PM of ambient
origin. Personal and ambient measures of SO42" were highly correlated over time, unlike the
moderate correlation found for PM2 5. The individual correlations demonstrated that ambient
SO42' was a consistently strong predictor across all individuals and all levels of exposure, whereas
for PM2.5 correlations varied by individual and were dependent upon the level of personal
exposure. Although indoor sources likely contribute to personal exposures of PM2 5, accounting
for such variables did not lead to models with the same predictive power as found for SO42".
March 2001
5-41
DRAFT-DO NOT QUOTE OR CITE
-------
Correlation Coefficient for Individuals
1.00 -
0.75 -
0.50 -
0.25 -
f 0.00 -
-0.25 -
-0.50 -
-0.75 -
_-i nn -
Ebelt et al., 2000
Pearson's r
P=.
PM2.S 5
I
Sulfate
Percentile
90th Percentile
75th Percentile
Median
25th Percentile
1 0th Percentile
Sarnatetal., 2000
Spearman's r
PM25
&
^
I
T_
~
I
NSSSSN Surr
i i Win
gL
T
Sulfate
imer
ter
PM25 Sulfate
PM25 Sulfate
Figure 5-1. Comparison of correlation coefficients for longitudinal analysis of personal
exposure versus ambient concentrations for individual subjects for PM2 5 and
sulfate.
1
2
3
4
5
6
7
Similarly, we found that accounting for spatial variability in ambient levels did not improve the
relationship between ambient concentrations and measured personal exposures. Overall, we have
shown that a personal measure of exposure to outdoor source PM is highly related to variation in
ambient levels of PM.
Another study conducted in Baltimore, MD, involved 15 nonsmoking adult subjects
(>64 years old) who were monitored for 12 days during summer 1998 and winter 1999 (Sarnat
March 2001
5-42
DRAFT-DO NOT QUOTE OR CITE
-------
1 et al., 2000). All subjects (nonrandom selection) were retired, physically healthy, and lived in
2 nonsmoking private residences. Each residence, except one, was equipped with central
3 air-conditioning; however, not all residences used air-conditioning throughout the summer. The
4 average age of the subjects was 75 years (±6.8 years). Sarnat et al. (2000) reported higher
5 longitudinal and pooled correlations for PM2 5 during summer than winter. Similar to Ebelt et al.
6 (2000), Sarnat et al. (2000) reported stronger associations between personal exposure to SO42"
7 and ambient concentrations of SO42". The ranges of correlations are shown in Figure 5-1 along
8 with similar data from Ebelt et al. (2000).
9 The study conducted by Sarnat et al. (2000) also illustrates the importance of ventilation on
10 personal exposure to PM. During the summer, subjects recorded the ventilation status of every
11 visited indoor location (e.g., windows open, air-conditioning use). As a surrogate for the
12 air-exchange rate, personal exposures were classified by the fraction of time the windows were
13 open while a subject was in an indoor environment (Fv). Sarnat et al. (2000) report regression
14 analyses for personal exposure on ambient concentration for total PM2 s and for sulfate for each
15 of the three ventilation conditions. Personal exposure to sulfate may be taken as a surrogate for
16 personal exposure to ambient accumulation-mode PM in the absence of indoor sulfate sources.
17 Figure 5-2 shows a comparison of the regressions and indicates how the use of a sulfate tracer as
18 a surrogate for PM of ambient origin improves the correlation coefficient. The improvement is
19 especially pronounced for the lowest ventilation conditions. For the lowest ventilation condition,
20 R2 improves from 0.25 to 0.72.
21 The Ebelt et al. (2000) and Sarnat et al. (2000) studies did not use their sulfate data to
22 develop relationships between personal exposure to ambient PM and ambient PM concentrations
23 for individual subjects, as suggested by Wilson et al. (2000). However, the higher correlation
24 coefficients and the narrower range of the correlation coefficient for sulfate suggest that
25 removing indoor-generated and personal activity PM from total personal PM would result in a
26 higher correlation with ambient concentrations. However, the variation in ventilation status (and
27 thus in the attenuation coefficient a) still would cause variations between ambient concentrations
28 of PM and personal exposure to ambient PM, especially if the study continued long enough to
29 extend through more than one season.
30
March 2001
5-43
DRAFT-DO NOT QUOTE OR CITE
-------
60'
g 40-
Q. 30-
"g 20-
IX 10-
Well Ventilated Indoor Environment
35-
R2=0.80
30
25
20
15
10
5
R2 = 0.88
60'
^ 50-
1
g- 40-
8. 30-
"S 20-
(? 10-
Moderately Vented Indoor Environment
35
R2 = 0.57
/•
/
<*'•'• '
30'
25
20
15
10
5
R2 = 0.73
60
50-
g- 40-
§. 30-
1 20-
Poorly Ventilated Indoor Environment
35
R2 = 0.25
10 20 30 40 50
PM2.5
30-
25-
20-
15-
10-
5-
0
R2 = 0.72
60 0
10 15 20 25 30 40
SO4
Ambient Concentration (M9/m3)
Figure 5-2. Personal exposure versus ambient concentrations for PM2 s and sulfate. (Slope
estimated from mixed models).
Source: Sarnat et al. (2000).
March 2001
5-44
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
5.4.3.1.6 Correlations Between Personal Exposure to Ambient and Nonambient
Particulate Matter
The utility of treating personal exposure to ambient PM, Eag, and personal exposure to
nonambient PM, Enonag, as separate and distinct components of total personal exposure to PM, Et,
was pointed out by Wilson and Suh (1997). The PTEAM study measured, in addition to indoor,
outdoor, and personal PM, the air exchange rate for each home and collected information on the
time spent in various indoor and outdoor #e. This information is available for 147, 12-h daytime
periods. With this information, it is possible to estimate the daytime Eag and Enonag as described
in Section 5.3.2.3.1. Various examples of this information have been reported (Mage et al.,
1999; Wilson et al., 2000). Graphs showing the relationships between ambient concentration and
the various components of personal exposure (Et, Eag, and Enonag) are shown in Figure 5-3. The
correlation coefficient for the pooled data set improves from r = 0.377 for Et versus Ca
(Figure 5-3a) to r = 0.856 for Eag versus Ca (Figure 5-3b) because of the removal of the Enonag ,
which, as shown in Figure 5-3c, is highly variable and independent of Ca. The correlation
between Eag and Ca is less than 1 because of the day-to-day variation in ait. The regression
analysis with E, total PM gives O~= 0.711 and N = 81.6 ^g/m3. The regression analysis with Eag
gives a = 0.625. The regression with Enonag gives N = 79.2 Aig/m3. The finite intercept in the
regression with Eag must be attributed to bias or error in some of the measurements. No studies,
other than PTEAM, have provided the quantity of data on Et, Ca, Ci5 and a required to conduct
an analysis comparable to that shown in Figure 5-3.
The higher correlations found between daily-average personal exposures and ambient PM
concentrations, as opposed to lower correlations found between individual exposures and
ambient PM levels, recently have been attributed to statistical rather than physical causes. Ott
et al. (2000), using their Random Component Superposition (RCS) model, solely attribute this to
the averaging process. Because personal exposures also include contributions from ambient
concentrations, the correlation between personal exposure and ambient concentrations increases
as the number of subjects measured daily increases. Based on theory, Ott et al. (2000) predict
expected correlations above 0.9 if 25 subjects had been studied during the PTEAM study and
above 0.70 in the New Jersey study reported by Lioy et al. (1990).
March 2001
5-45
DRAFT-DO NOT QUOTE OR CITE
-------
250
5150-1
S ^
0) en
|u!10°-
fig
I! 5°-
«|
i 0'
-SO'
r» 0.051
R'» 0.0026
N=39.6+0.086C
*=147
*
* *
^M^T^
; *' V T~»
50 100 150 200
Ambient Concentration, \ig/m3
250
Figure 5-3. Regression analyses of aspects of daytime personal exposure to PM10 estimated
using data from the PTEAM study, (a) Total personal exposure to PM, E,,
regressed on ambient concentration, Ca. (b) Personal exposure to ambient PM,
Eag regressed on Ca. (c) Personal exposure to nonambient PM, Enonag regressed
onCa.
Source: Data taken from Clayton et al. (1993).
March 2001
5-46
DRAFT-DO NOT QUOTE OR CITE
-------
1 The RCS model introduced by Ott et al.(2000) presents a modeling framework to determine
2 the contribution of ambient PM10 and indoor-generated PM10 on personal exposures in large
3 urban metropolitan areas. The model has been tested using personal, indoor and outdoor PM10
4 data from three urban areas (Riverside, CA; Toronto; and Phillipsburg, NJ). Results suggest that
5 it is possible to separate the ambient and nonambient PM contributions to personal exposures on
6 a community-wide basis. However, as discussed in the paper, the authors make some
7 assumptions that require individual consideration in each-city specific application of the model
8 for exposure or health effects investigations. Primarily, housing factors, air-conditioning,
9 seasonal differences, and complexities in time-activity profiles specific to the cohort being
10 studied have to be taken into account prior to adopting the model to a given situation.
11
12 5.4.3.2 Factors That Affect Correlations
13 A number of factors will affect the relationship between personal exposure and PM
14 measured at ambient-site community monitors. Spatial variability in outdoor microenvironments
15 and penetration into indoor microenvironments will influence the relationship for ambient-
16 generated PM, air-exchange rates and decay rates in indoor microenvironments will influence the
17 relationship for both ambient-generated and total PM, whereas personal activities will influence
18 the relationship for total PM but not ambient-generated PM. Information on these effects is
19 presented in detail in the following section.
20
21 5.4.3.2.1 Spatial Variability and Correlations Over Time
22 Chapter 3 (Section 3.2.3) presents information on the spatial variability of PM mass and
23 chemical components at fixed-site ambient monitors; for purposes of this chapter, this spatial
24 variability is called an "ambient gradient". The data presented in Section 3.2.3 indicate that
25 ambient gradients of PM and its constituents exist in urban areas to a greater or lesser degree.
26 This gradient, and any that may exist between a fixed-site monitor and the outdoor //e near where
27 people live, work, and play, obviously affects the exposure. The purpose of this section is to
28 review the available data on ambient monitor-to-outdoor microenvironmental concentration
29 gradients, or relationships, that have been measured by researchers since 1996. A few outdoor-
30 to-outdoor monitoring studies also are included to highlight relationships among important yue
31 categories. To assess spatial variability or gradients, the spatial correlations in the data are
March 2001
5-47
DRAFT-DO NOT QUOTE OR CITE
-------
1 usually analyzed. However, it should be noted that high temporal correlation between two
2 monitoring locations does not imply low spatial variability or low ambient gradients. High
3 temporal correlation between two sites indicates that changes in concentrations at one site can be
4 estimated from data at another site.
5 Oglesby et al. (2000), in a paper on the EXPOLIS-EAS study, conclude that very little
6 spatial variability exists in Basel, Switzerland, between PM levels measured at fixed site
7 monitors and the participant's outdoor //e. The authors report a high correlation between home
8 outdoor PM2.5 levels (48-h measurements beginning and ending at 8:00 a.m.) and the
9 corresponding 24-h average PM4 (time-weighted values calculated from midnight to midnight)
10 measured at a fixed monitoring station ^ = 38,^ = 0.96, p< 0.001). They considered each
11 home outdoor monitor as a temporary fixed monitor and concluded that "the PM2 5 level
12 measured at home outdoors ... represents the fine particle level prevailing in the city of Basel
13 during the 48-h measuring period...."
14 In a study conducted in Helsinki, Finland, Buzorius et al. (1999) conclude that a single
15 monitor may be used to adequately describe the ambient gradient across the metropolitan area.
16 Particle size distributions were measured using a differential mobility particle sizer (DMPS;
17 Wintlmayer) coupled with a condensation particle counter (CPC TSI3010, 3022) at four
18 locations including the official air monitoring station, which represented a "background" site.
19 The monitoring period varied between 2 weeks and 6 mo for the sites and data were reported for
20 10-min and 1-, 8-, and 24-h averages. As expected, temporal variation decreased as the
21 averaging time increased. The authors report that particle number concentration varied in
22 magnitude with local traffic intensity. Linear correlation coefficients computed for all possible
23 site-pairs and averaging times showed that the correlation coefficient improved with increasing
24 averaging time. Using wind speed and direction vectors, lagged correlations were calculated and
25 were generally higher than the "raw" data correlations. Weekday correlations were higher than
26 weekend correlations as "traffic provides relatively uniform spatial distribution of particulate
27 matter" (p. 565). The authors conclude that, even for time periods of 10 min and 1 h, sampling at
28 one station can describe changes across relatively large areas of the city with a correlation
29 coefficient >0.7.
30 Dubowsky et al. (1999) point out that, although the variation of PM2 5 mass concentration
31 across a community may be small, there may be significant spatial variations of specific
March 2001
5-48
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
components of the total mass on a local scale. An example is given of a study of concentrations
of polycyclic aromatic hydrocarbons (PAH) at three indoor locations in a community; an urban
and a semi-urban site separated by 1.6 km, and a suburban site located further away. The authors
found the geometric mean PAH concentrations at these three locations varied respectively as
31:19:8 ng/m3, and suggest that the local variations hi traffic density were responsible for this
gradient. Note that these concentrations are 1,000 times lower than the total PM mass
concentration, so that such a small gradient would not be detectable for total PM2 5 mass
measurements on the order of 25 //g m"3.
Leaderer et al. (1999) monitored 24-h PM10, PM2 5, and sulfates during the summers of
1995 and 1996 at a regional site in Vinton, VA (6 km from Roanoke, VA). One similar 24-h
measurement was made outdoors at residences in the surrounding area, at distances ranging from
1 km to > 175 km from the Vinton site, at an average separation distance of 96 km. The authors
reported significant correlations for PM2 5 and sulfates between the residential outdoor values and
those measured at Vinton on the same day. In addition, the mean values of the regional site and
residential site PM2 5 and sulfates showed no significant differences in spite of the large distance
separations and mountainous terrain intervening in most directions. However, for the
concentrations of PM2 5_10, estimated as PM,0-PM25, no significant correlation among these sites
was found (n = 30, r = -0.20).
Lillquist et al. (1998) found no significant gradient in PMIO concentrations in Salt Lake
City, UT, when levels were low, but a gradient existed when levels were high. PMIO
concentrations were measured outdoor at three hospitals using a Minivol 4.01 sampler
(Airmetrics, Inc.) operating at 5 L min'1 and at the Utah Department of Air Quality (DAQ)
ambient monitoring station located between 3 and 13 km from the hospitals for a period of about
5 mo.
Pope et al. (1999) monitored ambient PMIO concentrations in Provo, UT (Utah Valley),
during the same time frame the following year and reported nearly identical concentrations at
three sites separated by 4 to 12 km. Pearson correlation coefficients for the data were between
0.92 and 0.96. The greater degree of variability in the Salt Lake City PM10 data relative to the
Provo data maybe related to the higher incidence of wind-blown crustal material in Salt Lake
City. Pope et al. (1999) reported that increased health effects in the Utah Valley were associated
with stagnation and thermal inversions trapping anthropogenically derived PM10, whereas, no
March 2001
5-49
DRAFT-DO NOT QUOTE OR CITE
-------
1 increases in health effects were observed when PMIO levels were increased during events of wind
2 blown crustal material.
3 Vakeva et al. (1999) found significant vertical gradients in submicron particles existed in
4 an urban street canyon of Lahti, Finland. Particle number concentrations were measured using a
5 TSI screen diffusion battery and a condensation particle counter at 1.5 and 25 m above the street
6 at rooftop level. The authors found a fivefold decrease in concentration between the two
7 sampling heights and attributed the vertical gradient to dilution and dispersion of pollutants
8 emitted at street level.
9 White (1998) suggests that the higher random measurement error for the coarse PM
10 fraction compared to the error for the fine PM fraction may be responsible for a major portion of
11 the apparent greater spatial variability of coarse ambient PM concentration compared to fine
12 ambient PM concentration in a community (e.g., Burton et al., 1996; Leaderer et al., 1999).
13 When PM2.5 and PM10 are collected independently, and the coarse fraction is obtained by
14 difference (PM2 S_IO = PMi0-PM2 5), then the expected variance in the coarse fraction is the sum of
15 the variances of the PM10 and PM25 measurements. When a dichotomous sampler collects PM25
16 and PM2 5_10 on two separate filters, the coarse fraction also is expected to have a larger error than
17 the fine fraction. There is a possible error caused by loss of mass below the cut-point size and a
18 gain of mass above the cut-point size that is created by the asymmetry of the product of the
19 penetration tunes PM concentration about the cut-point size. Because a dichotomous PM
20 sampler collects coarse mass using an upper and lower cut-point, it is expected to have a larger
21 variance than for the fine mass collected using the same lower cut-point.
22 Wilson and Suh (1997) conclude that PM25 and PMi0 concentrations are correlated more
23 highly across Philadelphia than are PM2^.10 concentrations. Ambient monitoring data from 1992
24 to 1993 was reviewed for PM2 5, PM2 5.10, and PM10, as well as for PM2 5 and PM2 5_10 dichotomous
25 data for 212 site-years of information contained in the AIRS database. The authors also observed
26 that PM10 frequently was correlated more highly with PM2 5 than with PM2 5.10. The authors note
27 that PM2iS constitutes a large fraction of PM10, and that this is the likely reason for the strong
28 agreement between PM2 5 and PM10. Similar observations were made by Keywood et al. (1999)
29 in six Australian cities. The authors reported that PM10 was more highly correlated with PM2 5
30 than with coarse PM (PM2 5_10), suggesting that "variability in PM10 is dominated by variability in
31 PM2.5."
March 2001
5-50
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Lippmann et al. (2000) examined the site-to-site temporal correlations in Philadelphia
(1981 to 1994) and found the ranking of median site-to-site correlation was O3 (0.83), PM10
(0.78), TSP (0.71), N02 (0.70), CO (0.50), and SO2 (0.49). The authors explain that O3 and a
fraction of TSP and PM10 (e.g., sulfate) are secondary pollutants that would tend to be distributed
spatially more uniformly within the city than primary pollutants such as CO and SO2, which are
more likely to be influenced by local emission sources. Lippman et al. (2000) conclude: "Thus,
spatial uniformity of pollutants may be due to area-wide sources, or to transport (e.g., advection)
of fairly stable pollutants into the urban area from upwind sources. Relative spatial uniformity of
pollutants would therefore vary from city to city or region to region."
5.4.3.2.2 Physical Factors Affecting Indoor Microenvironmental Particulate Matter
Concentrations
Several physical factors affect ambient particle concentrations in the indoor /j.e, including
air exchange, penetration, and particle deposition. Combined, these factors are critical variables
that describe ambient particle dynamics in the indoor #e and, to a large degree, significantly
affect an individual's personal exposure to ambient-generated particles while indoors. The
relationship between ambient outdoor particles and ambient particles that have infiltrated indoors
is given by
(5-10)
where Cai and Cao are the concentration of ambient indoor and outdoor particles, respectively;
P is the penetration factor; a is the air exchange rate; and k is the particle deposition rate (as
discussed in Section 5.3.2.3.1, use of this model assumes equilibrium conditions and assumes
that all variables remain constant). Particle penetration is a dimensionless quantity that describes
the fraction of ambient particles that effectively penetrates the building shell. "Air exchange" is
a term used to describe the rate at which the indoor air in a building or residence is replaced by
outdoor air. The dominant processes governing particle penetration are air exchange and
deposition of particles as they traverse through cracks and crevices and other routes of entry into
the building. Although air-exchange rates have been measured in numerous studies, very few
field data existed prior to 1996 to determine size-dependent penetration factors and particle
March 2001
5-51
DRAFT-DO NOT QUOTE OR CITE
-------
1 deposition rates. All three parameters (P, a, and K) may vary substantially depending on building
2 type, region of the country, and season. In the past several years, researchers have made
3 significant advancements in understanding the relationship between particle size and penetration
4 factors and particle deposition rates. This section will highlight the studies that have been
5 conducted to better understand physical factors affecting indoor particle dynamics.
6
7 Air-Exchange Rates
8 The air-exchange rate, a, in a residence varies depending on a variety of factors, including
9 geographical location, age of the building, the extent to which window and doors are open, and
10 season. Murray and Burmaster( 1995) used measured values of a from households throughout
11 the United States to describe empirical distributions and to estimate univariate parametric
12 probability distributions of air-exchange rates. Figure 5-4 shows the results classified by season
13 and region. In general, a is highest in the warmest region and increases from the coldest to the
14 warmest region during all seasons. Air-exchange rates also are quite variable within and between
15 seasons, as well as between regions (Figure 5-4). Data from the warmest region in summer
16 should be viewed cautiously as many of the measurements were made in Southern California in
17 July when windows were more likely to be open than in other areas of the country where
18 air-conditioning is used. Use of air-conditioning generally results in lowering air-exchange rates.
19 In a separate analyses of these data, Koontz and Rector (1995) suggested that a conservative
20 estimate for air exchange in residential settings would be 0.18 h'1 (1 Oth percentile) and a typical
21 air exchange would be 0.45 h"1 (50th percentile).
22 These data provide reasonable experimental evidence that a varies by season in locations
23 with distinct seasons. As a result, infiltration of ambient particles may be more efficient during
24 warmer seasons when windows are likely to be opened more frequently and air-exchange rates
25 are higher. This suggests that the fraction of ambient particles present in the indoor //e would be
26 greater during warmer seasons than colder seasons. For example, in a study conducted in
27 Boston, MA, participants living in non-air-conditioned homes kept the windows closed except
28 during the summer (Long et al., 2000a). This resulted in higher and more variable air-exchange
29 rates in summer than during any other season (Figure 5-5). During nighttime periods, when
30 indoor sources are negligible, the indoor/outdoor concentration ratio or infiltration factor may be
31 used to determine the relative contribution of ambient particles in the indoor y.e. Particle data
March 2001
5-52
DRAFT-DO NOT QUOTE OR CITE
-------
O»
WD
S
3 -
_
2H
•* 1 H
o
Coldest Region
Colder Region
Warmer Region
Warmest Region
Winter Spring Summer
Season
Fall
Figure 5-4. Air-exchange rates measured in homes throughout the United States. Climatic
regions are based on heating-degree days: Coldest region £ 7000, Colder
region = 5500 to 6999, Warmer region = 2500 to 4999, and Warmest region
<. 2500 heating-degree days.
Based on data from Murray and Burmaster (1995).
1
2
3
4
5
collected during this study (Figure 5-6) shows the indoor/outdoor concentration ratios by particle
size. Data show that for these nine homes in Boston, the fraction of ambient particles penetrating
indoors is higher during summer when air exchange rates were higher than fall (Long et al.,
2000b).
March 2001
5-53
DRAFT-DO NOT QUOTE OR CITE
-------
8
7 -
=S 6H
U)
03
O)
§ 6
I 4
OJ
CO
01
£
co
i
£
0 -
95%
.90%
Median
Fall
Winter
Spring
Season
Summer
Figure 5-5. Box plots of hourly air-exchange rates stratified by season in Boston, MA,
during 1998.
Source: Long et al. (2000a).
1 Particle Deposition Rates and Penetration Factors
2 Physical factors affecting indoor particle concentrations, including particle deposition rates,
3 k, and penetration factors, P, are possibly the most uncertain and variable quantities. Although k
4 can be modeled with some success, direct measurements are difficult and results often vary from
5 study to study. Particle deposition rates vary considerably depending on particle size because of
6 the viscous drag of air on the particles hindering their movement to varying degrees. The nature
7 and composition of particles also affect deposition rates. Surface properties of particles, such as
8 their electrostatic properties, can have a significant influence on deposition rates. In addition,
9 thennophoresis can also affect k, but probably to a lesser degree in the indoor /we because
March 2001
5-54
DRAFT-DO NOT QUOTE OR CITE
-------
a) Home NEW2
CO
LL
c
.o
13
<*=
1.1
1.0 -
0.9 -
0.8 -
0.7 -
0.6 -
0.5 -
0.4 -
0.3 -
0.2 -
0.1 -
0.0
0.1
Summer Fall
eo
o
I
o
P 8
9 9
8 S
S*- in
. o T
S S 9
Poo
8 ° 5
o
—I
co
o
in
o
CO
c\i
•^ m
CO 4
o
S
o o o
Particle Size (|jm)
Figure 5-6. Geometric mean infiltration factor (indoor/outdoor ratio) for hourly nighttime,
nonsource data for two seasons. Box plots of air exchange rates are shown as
inserts for each plot. (Boston, 1998)
Source: Long et al. (2000b).
1 temperatures generally vary over a small range. Combined, these effects can produce order of
2 magnitude variations in k between particles of different size and, in the case of electrophoresis
3 and thermophoresis, particles of the same size.
4 Particle penetration efficiency into the indoor /we depends on particle size and air exchange
5 rates. Penetration varies with particle size because of the size-dependent deposition of particles
6 caused by impaction, interception, and diffusion of particles onto surfaces as they traverse
7 through cracks and crevices. Penetration also is affected by air exchange rates. When air
8 exchange rates are high, P approaches unity because the majority of ambient particles have less
9 interaction with the building shell. In contrast, when air exchange rates are low, P is governed by
10 particle deposition as particles travel through cracks and crevices.
March 2001
5-55
DRAFT-DO NOT QUOTE OR CITE
-------
1 Significant advancements have been made in the past few years to better characterize
2 particle deposition rates and penetration factors. Several new studies, including two in which
3 semi-continuous measurements of size distributions were measured indoors and outdoors, have
4 produced new information on these quantities, which are key to understanding the contributions
5 of ambient PM to indoor PM concentrations (Equation 5-10).
6 Studies involving semi-continuous measurements of indoor and outdoor particle size
7 distributions have been used to estimate k and P as a function of particle size (Vette et al., 2001;
8 Long et al., 2000b; Abt et al., 2000b). These studies each demonstrated that the indoor/outdoor
9 concentration ratios (Cao/C in Equation 5-11) were highest for accumulation mode particles and
10 lowest for ultrafine and coarse-mode particles. Various approaches were used to estimate size-
11 specific values for k and P. Vette et al. (2001) and Abt et al. (2000b) estimated k by measuring
12 the decay of particles at times when indoor levels were significantly elevated. Vette et al. (2001)
13 estimated P using measured values of k and indoor/outdoor particle measurements during
14 nonsource nighttime periods. Long et al. (2000b) used a physical-statistical model, based on
15 Equation 5-10, to estimate k and P during nonsource nighttime periods. The results for k
16 reported by Long et al. (2000b) and Abt et al. (2000b) are compared with other studies in
17 Figure 5-7. Although not shown in Figure 5-7, the results for k obtained by Vette et al. (2001)
18 were similar to the values of £ reported by Abt et al. (2000b) for particle sizes up to 1 /u.m.
19 Results for P by Long et al. (2000b) show that penetration was highest for accumulation-mode
20 particles and decreased substantially for coarse-mode particles (Figure 5-8). The results for
21 P reported by Vette et al. (2001) show similar trends, but are lower than those reported by Long
22 et al. (2000b). This likely is because of lower air-exchange rates in the Fresno, CA, residence
23 (a ~ 0.5 h"1; Vette et al., 2001) than the Boston, MA, residences (a > 1 h'1; Long et al., 2000b).
24 These data for P and k illustrate the role that the building shell may provide in increasing the
25 concentration of particles because of indoor sources and reducing the concentration of indoor
26 particles from ambient sources, especially for homes with low air-exchange rates.
27
28 Compositional Differences Between Indoor-Generated and Ambient-Generated
29 Particulate Matter
30 Wilson et al. (2000) discuss the differences in composition between particles from indoor
31 and outdoor sources. They note that, because of the difficulty in separating indoor PM into
March 2001
5-56
DRAFT-DO NOT QUOTE OR CITE
-------
10-
^_
J
S 1-
0.1-
-
0.
•
31
o Byrne (1992)
0 Foghefa/. (1997)a
v Ligockiefa/. (1990)"
v Ozkaynak ef a/. (1 996a)c
n Sinclair ef al. (1 988, 1 990)b
0 Thatcher and Layton (1 995)"
O Wallace ef al. (1997)d
T Abtefa/. (2000b)d
• Longefa/. (2000b)d'e
• • • Lai and Nazaroff (2000)'
f -
• """"•--.. * *
0.1
f-
* * i
' ,
' i
«
'f \s$]
r
i
error bar
hdudes 0
2,
^»
ifj
error bar
IndudesO
5
.-'
.*
.-•*
f
IT 1
."f ,
_.'*
f
J.
r
i
10
Particle size (|jm)
'Decay rates represent Summary Estimates from the four houses examined.
"Decay rates are based on sulfate and are presented as <2.5 um.
Estimates were computed using a surface-to-volume ratio of 2 m-1 (Koutrakis ef at, 1992).
'Data represent PM^j
"Particle sizes are the midpoint of the ranges examined.
•Decay rates presented are estimates of k for nightiy average data from all nine study homes.
Decay rates are theorectically modeled deposition values for smooth indoor surfaces and homogeneous and isotropically turbulent air flow.
Presented curves assume typical room dimensions (3 m x 4 m x S m) and a friction velocity of 1.0 cm/s.
Figure 5-7. Comparison of deposition rates from this study with literature values (adapted
from Abt et al., 2000b). Error bars represent standard deviations for same-
study estimates.
Source: Long et al. (2000b).
1
2
3
4
5
6
1
8
ambient and nonambient PM, there is little direct experimental information on the composition
differences between the two. Although experimental data are limited, Wilson et al. (2000)
suggest the following.
Photochemistry is significantly reduced indoors; therefore, most secondary sulfate [H2SO4,
NH4HSO4, and (NH4)2SO4] and nitrate (NH4NO3) found indoors come from ambient sources.
Primary organic emissions from incomplete combustion may be similar, regardless of the source.
However, atmospheric reactions of polyaromatic hydrocarbons and other organic compounds
March 2001
5-57
DRAFT-DO NOT QUOTE OR CITE
-------
£
•o
UJ
c
1.2
1.1
1.0
0.9
0.8
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.2
1.1
1.0
0.9
0.8 ^
0.7 jiT
£
8.
1
o"
o
o
9
o
9 ^
o a
s
0.6
0.5 g
s-
0.4 Q
0.3
0.2
0.1
0.0
Size Interval (pm)
Figure 5-8. Penetration efficiencies and deposition rates from models of nightly average
data. Error bars represent standard errors. (Boston, 1998, winter and
summer)
Source: Long et al. (2000b).
1 produce highly oxygenated and nitrated products, so these species are also of ambient origin.
2 Gasoline, diesel fuel, and vehicle lubricating oil all contain naturally present metals or metal
3 additives. Coal and heavy fuel oil also contain more metals and nonmetals, such as selenium and
4 arsenic, than do materials such as wood or kerosene burned inside homes. Environmental
5 tobacco smoke (ETS), however, with its many toxic components, is primarily an indoor-generated
6 pollutant
7
8 Particles generated indoors may have different chemical and physical properties than those
9 generated by anthropogenic ambient sources. Siegmann et al. (1999) have demonstrated that
10 elemental carbon in soot particles generated indoors have different properties than in those
11 generated outdoors by automotive or diesel engines. In the United States, combustion-product
March 2001
5-58
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
PM in the ambient/outdoor air generally is produced by burning fossil fuels (e.g., coal, gasoline,
fuel oil) and wood, whereas combustion-product PM from indoor sources is produced by
biomass burning (e.g., tobacco, wood, foods, etc.). However, some indoor sources of PM (such
as cigarette smoking, meat cooking, and coal burning) occur both indoors and outdoors and may
constitute an identifiable portion of measured ambient PM (Cha et al., 1996; Kleeman and Cass,
1998).
Indoor Air Chemistry
Gas- and aerosol-phase chemical reactions in the indoor jj,e are responsible for secondary
particle formation and modification of existing particles. Homogeneous gas phase reactions
involving ozone and terpenes (specifically d-limonene, a-terpinene, and a-pinene) have been
identified as an important source of submicron particles (Weschler and Shields, 1999). Terpenes
are present in several commonly available household cleaning products and d-limonene has been
identified in more than 50% of the buildings monitored in the BASE study (Hadwen et al., 1997).
Long et al. (2000a) found that when PineSol (primary ingredient is a-pinene) was used indoors,
indoor PM2 5 mass concentrations increased by 3 to 32 //g m'3 (indoor ozone concentrations
unknown, but ambient ozone concentrations were 44 to 48 ppb). Similarly, a 10-fold increase in
number counts of 0.1 to 0.2 //m particles was observed in an experimental office containing
supplemented d-limonene and normally encountered indoor ozone concentrations (< 5 to
45 ppb), resulting in an average increase in particle mass concentration of 2.5 to 5.5 //g m"3
(Weschler and Shields, 1999). Ozone appears to be the limiting reagent as particle number
concentration varied proportionally to ozone concentrations (Weschler and Shields, 1999). Other
studies showed similar findings (e.g., Jang and Kamens, 1999; Wainman et al., 2000).
Indoor Sources of Particles
The major sources of indoor PM in nonsmoking residences and buildings include
suspension of PM from bulk material, cooking, cleaning, and the use of combustion devices,
such as stoves and kerosene heaters. Human and pet activities also lead to PM detritus
production (from tracked-in soil, fabrics, skin and hair, home furnishings, etc.), which is found
ubiquitously in house dust deposited on floors and other ulterior surfaces. House dust and lint
particles may be resuspended indoors by agitation (cleaning) and turbulence (HVAC systems,
March 2001
5-59
DRAFT-DO NOT QUOTE OR CITE
-------
1 human activities, etc.). Ambient particles that have infiltrated into the indoor /we also may be
2 resuspended after deposition to indoor surfaces. Typically, resuspension of particles from any
3 source involves coarse-mode particles (>1 Aim); particles of smaller diameter are not resuspended
4 efficiently. On the other hand, cooking produces both fine- and coarse-mode particles, whereas
5 combustion sources typically produce fine-mode particles.
6 Environmental tobacco smoke (ETS) is also a major indoor source of PM. It is, however,
7 beyond the scope of this chapter to review the extensive literature on ETS. A number of articles
8 provide source strength information for cigarette or cigar smoking (e.g., Daisey et al. [1998] and
9 Nelson etal. [1998]).
10 A study conducted on two homes in the Boston metropolitan area (Abt et al., 2000a)
11 showed that indoor PM sources predominate when air exchange rates were <1 h"1, and outdoor
12 sources predominate when air exchange rates were >2 h"1. The authors attributed this to the fact
13 that when air-exchange rates were low (<1 h'1), particles released from indoor sources tend to
14 accumulate because particle deposition is the mechanism governing particle decay and not air
15 exchange. Particle deposition rates are generally <1 h'1, especially for accumulation-mode
16 particles. When air-exchange rates were higher (>2 h'1), infiltration of ambient aerosols and
17 exfiltration of indoor-generated aerosols occur more rapidly, reducing the impact of indoor
18 sources on indoor particle levels. The study also confirmed previous findings that the major
19 indoor sources of PM are cooking, cleaning, and human activity. They discuss the size
20 characteristics of these ubiquitous sources and report the following.
21
22 The size of the particles generated by these activities reflected their formation processes.
23 Combustion processes (oven cooking, toasting, and barbecuing) produced fine particles and
24 mechanical processes (sauteing, firing, cleaning, and movement of people) generated coarse
25 particles. These activities increased particle concentrations by many orders of magnitude higher
26 than outdoor levels and altered indoor size distributions. (Abt et al., 2000a; p. 43)
27
28 They also note that variability in indoor PM for all size fractions was greater than for outdoor
29 PM, especially for short averaging times (2 to 33 times higher).
30 In a separate study conducted in nine nonsmoking homes in the Boston area, Long et al.
31 (2000a) concluded that the predominant source of indoor fine particles was infiltration of outdoor
March 2001
5-60
DRAFT-DO NOT QUOTE OR CITE
-------
1 particles, and that cooking activities were the only other significant source of fine particles.
2 Coarse particles, however, had several indoor sources, such as cooking, cleaning, and various
3 indoor activities. This study also concluded that more than 50% of the particles (by volume)
4 generated during indoor events were ultrafine particles. Events that elevated indoor particle
5 levels were found to be brief, intermittent, and highly variable, thus requiring the use of
6 continuous instrumentation for their characterization. Table 5-8 provides information on the
7 mean volume mean diameter (VMD) for various types of indoor particle sources. The
8 differences in mean VMD confirm the clear separation of source types and suggest that there is
9 very little resuspension of accumulation-mode PM. In addition, measurements of organic and
10 elemental carbon indicated that organic carbon had significant indoor sources, whereas elemental
11 carbon was primarily of ambient origin.
12 Vette et al. (2001) found that resuspension was a significant indoor source of particles
13 >1 /zm, whereas fine- and accumulation-mode particles were not affected by resuspension.
14 Figure 5-9 shows the diurnal variability in the indoor/outdoor aerosol concentration ratio from an
15 unoccupied residence in Fresno. The study was conducted in the absence of common indoor
16 particle sources such as cooking and cleaning. The data in Figure 5-9 show the mean
17 indoor/outdoor concentration ratio for particles >1 /zm increased dramatically during daytime
18 hours. This pattern was consistent with indoor human activity levels. In contrast, the mean
19 indoor/outdoor concentration ratio for particles <1 //m (fine- and accumulation-mode particles)
20 remain fairly constant during both day and night.
21
22 5.4.3.2.3 Time/Activity Patterns
23 Total exposure to PM is the sum of various microenvironmental exposures that an
24 individual encounters during the day and will depend on the microenvironments occupied.
25 As discussed previously, PM exposure in each microenvironment is the sum of exposures from
26 ambient sources (Eag), indoor sources (Eig), and personal activities (Epact). Eag and Eig are
27 determined by the microenvironments in which an individual spends time; whereas Epact is
28 determined by the personal activities that he/she conducts while in those microenvironments.
29 Determining microenvironments and activities that contribute significantly to human
30 exposure begins with establishing human activity pattern information for the general population,
31 as well as subpopulations. Personal exposure and time activity pattern studies have shown that
March 2001
5-61
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-8. VOLUME MEAN DIAMETER (VMD) AND MAXIMUM PM2 5
CONCENTRATIONS OF INDOOR PARTICLE SOURCES a'b
Size Statistics
Particle Source
Cooking
Baking (Electric)
Baking (Gas)
Toasting
Broiling
Sauteing
Stir-Frying
Frying
Barbecuing
Cleaning
Dusting
Vacuuming
Cleaning with Pine Sol
General Activities
Walking Vigorously (w/Carpet)
Sampling w/Carpet
Sampling w/o Carpet
Burning Candles
N
8
24
23
4
13
3
20
2
11
10
5
15
52
26
7
Indoor Activity
Mean VMD
Cum)
0.1 89f
0.1 07f
0.138f
0.1 14f
0.184f,3.48g
0.135f
0.173f
0.159f
5.388
3.86g
0.097f
3.96g
4.25g
4.28B
0.311f
Background3'11
Mean VMD
Cum)
0.22 lf
0.224f
0.222f
0.236f
0.223f, 2.93s
0.277f
0.223f
0.205f
3.53E
2.79B
0.238f
3.18g
2.63B
2.93g
0.224f
Maximum
Mean
14.8
101.2
54.9
29.3
65.6
37.2
40.5
14.8
22.6
6.5
11.0
12.0
8.0
4.8
28.0
PM2.5
Concentration0'11
SD
7.4
184.9
119.7
43.4
95.4
31.4
43.2
5.2
22.6
3.9
10.2
9.1
6.6
3.0
18.0
Notes:
"All concentration data corrected for background particle levels.
Includes only individual particle events that were unique for a given time period and could be detected above
background particle levels.
CPM concentrations in yug/m3.
dMaximum concentrations computed from 5-min data for each activity.
'Background data are for time periods immediately prior to the indoor event.
fSize statistics calculated for PV0 02.0 5 using SMPS data.
8Size statistics calculated for PV0 7.10 using APS data.
Source: Long et al. (2000a).
1 different populations have varying time activity patterns and, accordingly, different personal PM
2 exposures. Both characteristics will vary greatly as a function of age, health status, ethnic group,
3 socioeconomic status, season, and region of the country. Collecting detailed time activity data
4 can be very burdensome on participants but is clearly valuable in assessing human exposure and
March 2001
5-62
DRAFT-DO NOT QUOTE OR CITE
-------
o
1
o
o
1
O
o
I
2.0'
0.0
Time
Figure 5-9. Mean hourly indoor/outdoor particle concentration ratio from an unoccupied
residence in Fresno, CA, during spring 1999.
Source: Vette et al. (2001).
1
2
3
4
5
6
7
8
9
10
11
microenvironments. For modeling purposes, human activity data frequently come from general
databases that are discussed below.
The gathering of human activity information, often called "time-budget" data, started in the
1920s; however, their use for exposure assessment purposes only began to be emphasized in the
1980s. Many of the largest U.S. human activity databases have been consolidated by EPA's
National Exposure Research Laboratory's (NERL) into one comprehensive database containing
over 22,000 person-days of 24-h activity known as the Consolidated Human Activity Database,
or CHAD (Glen et al., 1997). The information in CHAD will be accessible for constructing
population cohorts of people with diverse characteristics that are useful for analysis and
modeling (McCurdy, 2000). See Table 5-2 for a summary listing of human activity studies in
CHAD. Most of the databases in CHAD are available elsewhere, including the National Human
March 2001
5-63
DRAFT-DO NOT QUOTE OR CITE
-------
1 Activity Pattern Survey (NHAPS), California's Air Resources Board (CARB), and the University
2 of Michigan's Institute for Survey Research data sets.
3 Although CHAD provides a very valuable resource for time and location data, there is less
4 information on PM generating personal activities. In addition, very few of the time-activity
5 studies have collected longitudinal data within a season or over multiple seasons. Such
6 longitudinal data are important in understanding potential variability in activities and how they
7 impact correlations between PM exposure and ambient site measurements for both total PM and
8 PM of ambient origin.
9
10 5.4.3.3 Impact of Ambient Sources on Exposures to Particulate Matter
11 Different sources may generate ambient PM with different aerodynamic and chemical
12 characteristics, which may, in turn, result in different health responses. Thus, to fully understand
13 the relationship between PM exposure and health outcome, exposure from difference sources
14 should be identified and quantified. Source apportionment techniques provide a method for
15 determining personal exposure to PM from specific sources. Daily contributions from sources
16 that have no indoor component can be used as tracers to generate exposure to ambient PM of
17 similar aerodynamic size or directly as exposure surrogates in epidemiologic analyses. The
18 recent EPA PM Research Needs Document (U. S. Environmental Protection Agency, 1998)
19 recommended use of source apportionment techniques to determine daily time-series of source
20 categories for use in community, time-series epidemiology.
21 A number of epidemiological studies (discussed more fully in Chapter 6) have evaluated
22 the relationship between health outcomes and sources of particulate matter determined from
23 measurements at a community monitor. These studies suggest the importance of examining
24. sources and constituents of indoor, outdoor, and personal PM. Ozkaynak and Thurston (1987)
25 evaluated the relationship between particulate matter sources and mortality in 36 Standard
26 Metropolitan Statistical Areas (SMSAs). Particulate matter samples from EPA's Inhalable
27 Particle (IP) Network were analyzed for SO42" and NCy by automated colorimetry, and elemental
28 composition was determined with X-ray fluorescence (XRF). Mass concentrations from five
29 particulate matter source categories were determined from multiple regression of absolute factor
30 scores on the mass concentration: (1) resuspended soil, (2) auto exhaust, (3) oil combustion,
31 (4) metals, and (5) coal combustion.
March 2001
5-64
DRAFT-DO NOT QUOTE OR CITE
-------
1 Mar et al. (2000) applied factor analysis to evaluate the relationship between PM
2 composition (and gaseous pollutants) in Phoenix. In addition to daily averages of PM2 5 elements
3 from XRF analysis, they included in their analyses organic and elemental carbon in PM2 s and
4 gaseous species emitted by combustion sources (CO, NO2, and SO2). They identified five factors
5 classified as (1) motor vehicles, (2) resuspended soil, (3) vegetative burning, (4) local SO2, and
6 (5) regional sulfate.
7 Laden et al. (2000) applied specific rotation factor analysis to particulate matter
8 composition (XRF) data from six eastern cities (Ferris et al., 1979). Fine particulate matter was
9 regressed on the recentered scores to determine the daily source contributions. Three main
10 sources were identified: (1) resuspended soil (Si), (2) motor vehicle (Pb), and (3) coal
11 combustion (Se).
12 Source apportionment or receptor modeling has been applied to the personal exposure data
13 to understand the relationship between personal and ambient sources of particulate matter.
14 Application of source apportionment to ambient, indoor, and personal PM composition data is
15 especially useful in sorting out the effects of particle size and composition. If a sufficient
16 number of samples are analyzed with sufficient compositional detail, it is possible to use
17 statistical techniques to derive source category signatures, identify indoor and outdoor source
18 categories and estimate their contribution to indoor and personal PM.
19 . Positive Matrix Factorization (PMF) has been applied to the PTEAM database by
20 Yakovleva et al. (1999). The authors utilize mass and XRF elemental composition data from
21 indoor and outdoor PM25 and personal, indoor, and outdoor PM10 samples. PMF is an advance
22 over ordinary factor analysis because it allows measurements below the quantifiable limit to be
23 used by weighting them by their uncertainty. This effectively increases the number of species
24 that can be used in the model. The factors used by the authors correspond to general source
25 categories of PM, such as outdoor soil, resuspended indoor soil, indoor soil, personal activities,
26 sea-salt, motor vehicles, nonferrous metal smelters, and secondary sulfates. PMF, by identifying
27 not only the various source factors but also apportioning them among the different monitor
28 locations (personal, indoor, and outdoor), was able to quantify an estimate of the contribution of
29 resuspended indoor dust to the personal cloud (15% from indoor soil and 30% from resuspended
30 indoor soil). Factor scores for these items then were used in a regression analysis to estimate
31 personal exposures (Yakovleva etal., 1999).
March 2001
5-65
DRAFT-DO NOT QUOTE OR CITE
-------
1 The most important contributors to PMIO personal exposure were indoor soil, resuspended
2 indoor soil, and personal activities; these accounted for approximately 60% of the mass
3 (Yakovleva et al., 1999). Collectively, they include personal cloud PM, smoking, cooking, and
4 vacuuming. For both PM2.5 and PM10, secondary sulfate and nonferrous metal operations
5 accounted for another 25% of PM mass. Motor vehicle exhausts, especially starting a vehicle
6 inside of an attached garage, accounted for another 10% of PM mass. The authors caution that
7 these results may not apply to other geographic areas, seasons of the year, or weather conditions.
8 Simultaneous measurement of personal (PM10) and outdoor measurements (PM25 and
9 PMIO) were evaluated as a three-way problem with PMF, which allowed for differentiation of
10 source categories based on their variation in time and type of sample, as well as their variation in
11 composition. By use of this technique, it was possible to identify three sources of coarse-mode,
12 soil-type PM. One was associated with ambient soil, one was associated with indoor soil
13 dispersed throughout the house, and one was associated with soil resulting from the personal
14 activity of the subject.
15 Two other source apportionment models have been applied to ambient measurement data
16 and can be used for the personal exposure studies. The effective variance weighted Chemical
17 Mass Balance (CMB) receptor model (Watson et al., 1984, 1990, 1991) solves a set of linear
18 equations that incorporate the uncertainty in the sample and source composition. CMB requires
19 the composition of each potential source of PM and the uncertainty for the sources and ambient
20 measurements. Source apportionment with CMB can be conducted on individual samples,
21 however, composition of each of the sources of PM must be known. An additional source
22 apportionment model, UNMIX (Henry et al., 1994) is a multivariate source apportionment
23 model. UNMIX is similar to PMF, but does not use explicitly the measurement uncertainties.
24 Because measurement uncertainties are not used, only species above the detection limit are
25 evaluated in the model. UNMIX provides the number of sources and source contributions and
26 requires a similar number of observations as PMF.
27 The Yakovleva et al. (1999) study demonstrates that source apportionment techniques also
28 could be very useful in determining parameters needed for exposure models and for determining
29 exposure to ambient-generated PM. Exposure information, similar to that obtained hi the
30 PTEAM study, but including other PM components useful for definition of other source
31 categories (e.g., elemental [EC] and organic carbon [OC]; organic tracers for elemental carbon
March 2001
5-66
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
from diesel vehicle exhaust, gasoline vehicle exhaust, and wood combustion; nitrate; Na; Mg and
other heavy metal tracers; and, also, gas-phase pollutants) would be useful as demonstrated in the
use of EC/OC and gas-phase pollutants by Mar et al. (2000).
5.4.3.4 Correlations of Particulate Matter with Other Pollutants
Several epidemiological studies have included the gaseous pollutants CO, NO2, SO2, and
O3 along with PM10 or PM2 5 in the analysis of the statistical association of health responses with
pollutants. In a recent study, the personal exposure to O3 and NO2 were determined, as well as
that to PM25 and PM25.10 for a cohort 15 elderly subjects in Baltimore, MD. Sarnat et al. (2000)
conclude that the potential for confounding of PM2 5 by O3, NO2, or PM 2 5.10 appears to be
limited, because, despite significant correlations observed among ambient pollutant
concentrations, the correlations among personal exposures were low. Spearman correlations for
14 subjects in summer and 14 subjects in winter are given in Table 5-9 for relationships between
personal PM2 5 and ambient concentrations of PM25, PM2 5.10, O3, and NO2. In contrast to ambient
concentrations, neither personal exposure to total PM2 5 nor PM2 5 ambient origin was correlated
significantly with personal exposures to the co-pollutants, PM25.10, nonambient PM25, O3, NO2,
and SO2. Personal-ambient associations for PM2 5.10, O3, NO2, and SO2 were similarly weak and
insignificant. It should be noted that measured personal exposures to O3, NO2, and SO2 were
below their respective LOD for 70% of the samples.
A newly developed Roll-Around System (RAS) was used to evaluate the hourly
relationship between gaseous pollutants (CO, O3, NO2, SO2, and VOCs) and PM (Chang et al.,
2000). Exposures were characterized over a 15-day period for the summer and winter in
Baltimore, based on scripted activities to simulate activities performed by older adults (65+ years
of age). Spearman rank correlations were reported for PM2 5, O3, CO, and toluene for both the
summer and winter and the correlations are given for each microenvironment in Table 5-10:
indoor residence, indoor other, outdoor near roadway, outdoor away from road, and in vehicle.
No significant relationships (p < 0.05) were found between hourly PM2 5 and O3. Significant
relationships were found between hourly PM2 5 and CO: indoor residence, winter; indoor other,
summer and winter; and outdoor away from roadway, summer. Significant relationships also
were found between hourly PM2 5 and toluene: indoor residence, winter; indoor other, winter;
and in vehicle, winter. The significant relationships between CO and PM2 5 in the winter may be
March 2001
5-67
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-9. CORRELATIONS BETWEEN PERSONAL PM2.5 AND AMBIENT
POLLUTANT CONCENTRATIONS1
Personal PM2.5
vs. Ambient:
SUMMER
WINTER
Median
Median
Subject
SA1
SA2
SA5
SB1
SB2
SB3
SB4
SB5
SB6
SCI
SC2
SC3
SC4
SC5
WAI
WA2
WA4
WAS
WB1
WB2
WB3
WB4
WC1
WC2
WC3
WC4
WC5
WC6
Summer
Winter
PM2.5 03
0.55
0.55
0.59
0.65
-0.21
0.52
0.75
0.73
0.53
• 0.95
0.78
0.85
0.78
0.55
0.22
-0.38
-0.18
0.22
0.50
0.62
0.55
-0.12
0.74
0.79
0.28
0.19
0.57
0.01
0.76
0.25
'Correlations represent Spearman'
0.15
0.31
0.18
0.40
-0.62
0.55
0.62
0.45
0.15
0.75
0.65
0.75
0.66
0.51
-0.18
-0.07
0.67
-0.43
-0.84
-0.32
-0.45
-0.01
-0.62
-0.88
-0.42
-0.84
-0.62
-0.03
0.48
-0.43
NO2
0.38
0.66
0.52
-0.15
0.57
-0.14
-0.34
-0.42
-0.38
0.66
0.36
0.73
0.59
0.32
-0.26
-0.36
-0.22
0.67
0.77
0.59
0.62
0.34
-0.15
0.17
0.03
0.50
0.08
0.65
0.37
0.26
PM2.5.10
-0.12
0.57
0.64
0.38
0.15
-0.04
-0.12
0.23
0.12
0.65
0.51
0.65
0.70
0.43
-0.05
-0.70 .
-0.29
0.50
0.41
0.09
0.04
-0.10
0.44
0.77
0.57
0.45
0.57
0.37
0.41
0.39
Personal PM2 5
of Ambient Origin vs. Ambient:
03
0.27
0.21
0.33
0.59
0.26
0.52
0.45
0.36
-0.03
0.55
0.66
0.69
0.50
0.34
-0.75
-0.15
-0.33
-0.72
-0.57
-0.76
-0.77
-0.50
-0.64
-0.57
-0.77
-0.72
-0.76
-0.75
0.41
-0.76
s r values; italicized values indicate significance at the
NO2
0.77
0.64
0.57
-0.74
0.08
-0.20
-0.29
-0.48
-0.57
0.65
0.65
0.77
0.50
0.33
-0.04
-0.15
0.20
-0.09
0.53
0.59
0.56
0.65
0.02
0.25
0.30
0.22
0.05
0.19
0.42
0.21
a = 0.05 level.
PM2.5.10
0.15
0.68
0.79
-0.03
0.33
0.00
-0.14
0.33
0.32
0.57
0.76
0.50
0.51
0.27
-0.24
0.02
0.00
0.40
0.66
0.59
0.60
0.48
0.69
0.77
-0.45
0.67
0.42
-0.45
0.33
0.45
Source: Samat et al. (2000).
March 2001
5-68
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-10. CORRELATIONS BETWEEN HOURLY PERSONAL PM25 AND
GASEOUS POLLUTANTS
Indoor
Residence
PM^vs.
Summer
Winter
PM^vs.
Summer
Winter
PM25vs.
Summer
Winter
N
03
35
56
CO
41
59
Toluene
46
66
rs
0.29
0.05
0.25
0.43a
0.23
0.38a
Indoor Other
N
16
37
19
39
21
47
rs
-0.14
-0.06
0.59a
0.62"
-0.14
0.44a
Outdoor Near
Roadway
N
10
11
13
13
14
17
rs
0.05
-0.28
0.14
0.37
0.26
0.40
Outdoor Away
from road
N
12
7
12
8
14
8
rs
0.45
0.04
0.62
0.41
0.02
0.48
In Vehicle
N
37
34
46
37
48
42
rs
0.21
-0.10
0.23
0.10
0.12
0.43a
"Correlations represent Spearman's r values; italicized values indicate significance at the a = 0.05 level.
Source: Chang et al. (2000).
1 caused by reduced air-exchange rates that could allow them to accumulate (Chang et al., 2000).
2 Although no significant correlation was found between in vehicle PM2 5 and CO, toluene, which
3 is a significant component of vehicle exhaust (Conner et al., 1995), was correlated significantly
4 to PM2 5 in the winter.
5 Carrer et al. (1998) present data on the correlations among personal and
6 microenvironmental PM10 exposures and concentrations and selected environmental chemicals
7 that were monitored simultaneously (using a method that was not described). These chemicals
8 were nitrogen oxides (NOJ, carbon monoxide (CO), and total volatile organic compounds
9 (TVOC), benzene, toluene, xylene, and formaldehyde. The Kendall T correlation coefficient was
10 used; only results significant at p < 0.05 are mentioned here. Significant associations were found
11 only between the following pairs of substances (T shown in parentheses): personal PM10 (24 h)
12 and NOX (0.34), CO (0.34), TVOC (0.18), toluene (0.19), and xylene (0.26); office PM10 and NOX
13 (0.31); home PM10 and NOX (0.24), CO (0.24), toulene (0.17), and xylene (0.25). Surprisingly,
March 2001
5-69
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
because most of the chemical substances are associated with motor vehicular emissions, there
was no significant correlation between "commuting PM10" and any of the substances (Carrer
etal., 1998).
5.5. SUMMARY OF PARTICULATE MATTER CONSTITUENT DATA
5.5.1 Introduction
Atmospheric PM contains a number of chemical constituents that may be of significance
with respect to the human exposure and health effects. These constituents may be either
components of the ambient particles or bound to the surface of particles. They may be elements,
inorganic species, or organic compounds. A limited number of studies have collected data on
concentrations of elements, acidic aerosols, and polycyclic aromatic hydrocarbons (PAHs) in
ambient, personal, and microenvironmental PM samples. But, there have not been extensive
analyses of the constituents of PM in personal or microenvironmental samples. Data from
relevant studies are summarized in this section. The summary does not address bacteria,
bioaerosols, viruses, or fungi (e.g., Owen et al., 1992; Ren et al., 1999).
5.5.2 Monitoring Studies That Address Particulate Matter Constituents
A limited number of studies have measured the constituents of PM in personal or
microenvironmental samples. Relevant studies published in recent years are summarized in
Tables 5-11 and 5-12 for personal exposure measurements of PM and microenvironmental
samples, respectively. Studies that measured both personal and microenvironmental samples are
included in Table 5-11.
The largest database on personal, microenvironmental, and outdoor measurements of PM
elemental concentrations is the PTEAM study (Ozkaynak et al., 1996b). The results are
highlighted in the table and discussed below. The table shows that a number of studies have
measured aerosol acidity, sulfate, ammonia, and nitrate concentrations. Also, a number of
studies have measured PAHs, both indoors and outdoors. Other than the PAHs, there is little
data on organic constituents of PM.
March 2001
5-70
DRAFT-DO NOT QUOTE OR CITE
-------
W2
CE
•
P.
|
£E
5
x
Summary of Results
8
"3
CO
'o
o
CO
e
o
JS
S
Q.
g
U
O
S
CO
0
1
tju
o
S:
ea
2
CO
§
c
6
s
a.
«•§
llllj
S •= T3 3 °i
Illia
a s ° ^ v, •
lir was the major source for mo
iroviding 70 to 100% of the obi
tions for 12 of the 15 elements.
ts for central monitoring site v<
nces were 0.98 for sulfur and 0
[except copper).
g 01- S .2 •§ -2
| s s ;! s s
111 §11
a
"3
•5
oo
•<
0
u"
T3
'e
.S
OS
"«
•s
•i
0
1
s
„
§c
_
w
o
door or outdo
omparison.
\s and Pb levels higher than in<
3 community ambient site for ci
1--
W *fl>
£ S
8
S
r-
»n
§
•a
S
1
1
^
S
o
?.
J3
OT
^
1
£
e
«
"^
S^
loors. ButPti
cations than s
2.5 Mn higher outdoors than ind
ntrations higher at two fixed lo<
ts' homes.
-------
C/3
Ct
ged from 0.7 to 0.9 for three PM
4J (4-1
i 2
g a
t
•g 2
ill
i
S
o
5
i
3
a •§
,2
TS
t>
P
U
2
1
o
o
<
u
o-
§
J2o
.g
"a
Q
(X 0.
March 2001
5-72
DRAFT-DO NOT QUOTE OR CITE
-------
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
5.5.3 Key Findings
5.5.3.1 Correlations of Personal and Indoor Concentrations with Ambient Concentrations
of Participate Matter Constituents
The elemental composition of PM in personal samples was measured in the PTEAM study,
the first probability-based study of personal exposure to particles. A number of important
observations, made from the PTEAM data collected in Riverside, CA, are summarized by
Ozkaynak et al. (1996b). Population-weighted daytime personal exposures averaged
150 ± 9 Aig/m3, compared to concurrent indoor and outdoor concentrations of 95 ± 6 yUg/m3. The
personal exposure measurements suggested that there was a "personal cloud" of particles
associated with personal activities. Daytime personal exposures to 14 of the 15 elements
measured in the samples were considerably greater than concurrent indoor or outdoor
concentrations, with sulfur being the only exception.
The PTEAM data also showed good agreement between the concentrations of the elements
measured outdoors at the backyard of the residences with the concentrations measured at the
central site in the community. The agreement was excellent for sulfur. Although the particle and
element mass concentrations were higher in personal samples than for indoor or outdoor samples,
a nonlinear mass-balance method showed that the penetration factor was nearly 1 for all particles
and elements.
Similarly to the PTEAM results, recent measurements of element concentrations in
NHEXAS showed elevated concentrations of As and Pb in personal samples relative to indoor
and outdoor samples (Clayton et al., 1999b). The elevated concentrations of As and Pb were
consistent with elevated levels of PM in personal samples (median particle exposure of
101 //g/m3), compared to indoor concentrations (34.4 ^g/m3). There was a strong association
between personal and indoor concentrations and indoor and outdoor concentrations for both As
and Pb. However, there were no central site ambient measurements for comparison to the
outdoor or indoor measurements at the residences.
Manganese (Mn) concentrations were measured in PM2 5 samples collected in Toronto
(Crump, 2000). The mean PM2 5 Mn concentrations were higher outdoors than indoors. But the
outdoor concentrations measured at the participant's homes were lower than those measured at
two fixed locations. Crump (2000) suggested that the difference in the concentrations may have
March 2001
5-73
DRAFT-DO NOT QUOTE OR CITE
-------
1 been because the fixed locations were likely closer to high-traffic areas than were the
2 participant's homes.
3 Studies of acidic aerosols and gases typically measure strong acidity (H+), SO42', NH4+, and
4 NO3". The relationship between the concentrations of these ions and the relationship between
5 indoor and outdoor concentrations have been addressed in a number of studies during which
6 personal samples, microenvironmental, and outdoor samples have been collected, as shown in
7 Tables 5-11 and 5-12. Key findings from these studies include those shown below.
8 • Acid aerosol concentrations measured at the residences in the Uniontown, PA, study were
9 significantly different from those measured at a fixed ambient site located 16 km from the
10 community. But, Leaderer et al. (1999) reported that the regional ambient air monitoring
11 site in Vinton, VA, provided a reasonable estimate of indoor and outdoor sulfate
12 measurements during the summer at homes without tobacco combustion.
13 • Approximately 75% of the fine aerosol indoors during the summer was associated with
14 outdoor sources based on I/O sulfate ratios measured in the Leaderer et al. (1999) study.
15 • Personal exposures to strong acidity (ET) were lower than corresponding outdoor levels
16 measured in studies by Brauer et al. (1989,1990) and Suh et al. (1992). But the personal
17 exposure levels measured by Suh et al. (1992) were higher than the indoor
18 microenvironmental levels.
19 • Personal exposures to NH4+, and NO3' were reported by Suh et al. (1992) to be lower than
20 either indoor or outdoor levels.
21 • Personal exposures to SO42" were also lower than corresponding outdoor levels, but
22 higher than the indoor microenvironmental levels (Suh et al., 1992; 1993a,b), as shown in
23 Table 5-13.
24 The fact that the personal and indoor H* concentrations were substantially lower than
25 outdoor concentrations suggests that a large fraction of aerosol strong acidity is neutralized by
26 ammonia. Ammonia is emitted in relatively high concentrations in exhaled breath and sweat.
27 The difference between indoor and outdoor H+ concentrations in the Suh et al. (1992, 1993a,b)
28 studies was also much higher than the difference for indoor and outdoor SO42", indicative of
29 neutralization of the H+. Results of the Suh et al. (1992, 1993a,b) studies also showed substantial
30 interpersonal variability of H+ concentrations that could not be explained by variation in outdoor
31 concentrations.
March 2001
5-74
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-13. SUMMARY STATISTICS FOR PERSONAL, INDOOR, AND
OUTDOOR CONCENTRATIONS OF SELECTED AEROSOL COMPONENTS IN
TWO PENNSYLVANIA COMMUNITIES
Aerosol
State College
NO3-
SO42'
NH/
H+
Uniontown
SO42'
NH/
H+
Home Type
A/C Homes0
Non-A/C
A/C Homes
Non-A/C
All Homes'*
All Homes
A/C Homes
Non-A/C
All Homes0
All Homes'
All Homes0
All Homes0
Sample Site
(In/Out)a
53/71
254/71
56/75
259/75
214/76
314/155
28/74
230/74
163/75
91/46
91/44
91/46
Indoor (12 h)
GM ± GSDb
2.1 ±2.7
3.2 ±2.3
61.8 ±2.5
96.7 ± 2.5
69.1 ±2.6
154.7 ±2.8
4.2 ± 4.3
11.2±3.1
9.1 ±3.5
87.8 ±2.1
157.2 ±2.8
13.7 ±2.5
Concentration
(nmol m"3)
Outdoor (24 h)
GM ± GSDb
1.4 ±2.1
1.4±2.1
109.4 ± 2.4
109.4 ± 2.4
91.0 ±2.5
104.4 ±2.3
82.5 ±2.6
82.5 ± 2.6
72.4 ± 2.9
124.9 ±1.9
139.4 ±2.1
76 6 ± 2 7
Personal (12 h)
GM ± GSDb
— •
71. 5 ±2.4
—
18.4 ±3.0
110.3 ±1.8
167.0 ±2.0
42 8 ± 2 2
"In/Out = Indoor sample site/outdoor sample site.
bGM ± GSD = Geometric mean ± geometric standard deviation.
°A/C Homes = Homes that had air-conditioning (A/C); this does not imply that it was on during the entire
sampling period.
Non-A/C = Homes without air conditioning.
dThe sample size (n) for the personal monitoring = 209.
°n = 174 for personal monitoring.
Source: Suhetal. (1992,1993a,b).
1
2
3
4
5
6
Similar results for ammonia were reported by Waldman and Liang (1993). They reported
that levels of ammonia in institutional settings that they monitored were 10- to 50- times higher
than outdoors, and that acid aerosols were largely neutralized. Leaderer et al. (1999) reported
that ammonia concentrations during both winter and summer in residences were an order of
magnitude higher indoors than outdoors, consistent with results of other studies and the presence
of sources of ammonia indoors.
March 2001
5-75
DRAFT-DO NOT QUOTE OR CITE
-------
1 Sulfate aerosols appear to penetrate indoors effectively. Waldman et al. (1990) reported
2 I/O ratios of 0.7 to 0.9 in two nursing care facilities and a day-care center. Sulfate I/O ratios were
3 measured for three particle size fractions in 12 residences in Birmingham, England, by Jones
4 et al. (2000). The sulfate I/O ratios were 0.7 to 0.9 for PM < 1.1 ywm, 0.6 to 0.8 for PM 1.1 to
5 2.1 /zm, and 0.7 to 0.8 for PM 2.1 to 10 //m. Suh et al. (1993b) reported that personal and
6 outdoor sulfate concentrations were highly correlated, as depicted in Figure 5-10.
600
Q.
0
100 200 300 400 500
Outdoor Sulfate (nmoles/m3)
600
Figure 5-10. Personal versus outdoor SO4= in State College, PA. Open circles represent
children living in air conditioned homes; the solid line is the 1:1 line.
Source: Suh et al. (1993b).
March 2001
5-76
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Indoor/outdoor relationships were measured for a number of PM25 components and related
species in Lindon, UT, during January and February of 1997 by Patterson and Eatough (2000).
Outdoor samples were collected at the Utah State Air Quality monitoring site. Indoor samples
were collected in the adjacent Lindon Elementary School. The infiltration factors, Cai/Cao, given
by the slope of the regression lines (Table 5-14), were low (0.27 for sulfate and 0.12 for PM2 5),
possibly because of removal of particles in the air heating and ventilation system. The authors
concluded that the data indicate that indoor PM2 5 mass may not always be a good indicator of
exposure to ambient combustion material caused by the influence of indoor sources of particles.
However, ambient sulfate, SO2, nitrate, soot, and total particulate number displayed strong
correlations with indoor exposure. Ambient PM2 5 mass was not a good indicator of indoor PM2 5
mass exposure.
TABLE 5-14. STATISTICAL CORRELATION OF OUTDOOR (x) VERSUS INDOOR
(y) CONCENTRATION FOR MEASURED SPECIES
(Units are nmol m'3, except for soot and metals, which are /
and absorption units m"3, respectively.)3
Species
SO2 All Samples
SO2 Day Samples
SO2 Night Samples
Sulfate All Samples
Sulfate Day Samples
Sulfate Night Samples
Nitrate All Samples
Nitrate Day Samples
Nitrate Night Samples
Soot Day Samples
Soot Night Samples
Total Acidity All Samples
Metals All Samples
Slope
0.0272 ± 0.0023
0.0233 ± 0.0037
0.0297 ± 0.0029
0.267 ±0.024
0.261 ± 0.034
0.282 ± 0.035
0.0639 ± 0.0096
0.097 ± 0.0096
0.047 ±0.0 11
0.43 ± 0.25
0.33 ±0.13
0.04 ± 0.73
0.10 ±0.30
Intercept
0.34 ±0.1 3
0.75 ± 0.26
0.099 ± 0.075
-0.14 ±0.48
0.40 ± 0.66
-0.84 ±0.68
0.9 ±1.5
-0.4 ± 1.4
1.5 ±1.8
3.5 ±1.7
0.00 ± 0.55
0.42 ± 0.23
0.0014 ± 0.0042
r2
0.73
0.62
0.82
0.70
0.71
0.70
0.54
0.88
0.44
0.43
0.69
0.00
0.01
Average
Outdoors
38
56
20
16
16
16
134
126
139
6
4
0.2
0.0042
"Linden Elementary School, Lindon, UT, January and February 1997.
Source: Patterson and Eatough (2000).
March 2001
5-77
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Oglesby et al. (2000) conducted a study to evaluate the validity of fixed-site fine particle
concentration measurements as exposure surrogates for air pollution epidemiology. Using 48-h
EXPOLIS data from Basel, Switzerland, they investigated the personal exposure/outdoor
concentration relationships for four indicator groups: (1) PM2 5 mass, (2) sulfur and potassium
for regional air pollution, (3) lead and bromine for traffic-related particles, and (4) calcium for
crustal particles. The authors reported that personal exposures to PM2 5 mass were not correlated
to corresponding home outdoor levels (n = 44, r = 0.07). In the study group reporting neither
relevant indoor sources nor relevant activities, personal exposures and home outdoor levels of
sulfur were highly correlated (n = 40, r = 0.85). Oglesby et al. (2000) concluded that
"for regional air pollution, fixed-site fine particle levels are valid exposure surrogates. For source-
specific exposures, however, fixed-site data are probably not the optimal measure. Still, in air pollution
epidemiology, ambient PM2^ levels may be more appropriate exposure estimates than total personal
exposure, since the latter reflects a mixture of indoor and outdoor sources."
PAHs have been measured in studies by EPA and the California Air Resources Board.
PAH results from a probability sample of 125 homes in Riverside are discussed in reports by
Sheldon et al. (1992a,b) and Ozkaynak et al. (1996b). Data for two sequential 12-h samples were
reported for PAHs by ring size (3 to 7) and for individual phthalates. The results are summarized
below.
• The particulate-phase 5- to 7-ring species had lower relative concentrations than the more
volatile 3- to 4-ring species.
• The 12-h indoor/outdoor ratios for the 5- to 7-ring species ranged from 1 . 1 to 1 .4 during
the day and from 0.64 to 0.85 during the night (Sheldon et al., 1993a).
• An indoor air model used to calculate indoor "source strengths" for the PAHs showed
that smoking had the strongest effect on indoor concentrations.
Results from a larger PAH probability study in 280 homes in Placerville and Roseville
(Sheldon et al., 1993b,c) were similar to the 125-home study. The higher-ring, particle-bound
PAH's had lower indoor and outdoor concentrations than the lower-ring species. For most
PAHs, the I/O ratio was greater than 1 for smoking and smoking/fireplace homes and less than
1 for fireplace-only, wood stove, wood stove/gas heat, gas heat, and "no source" homes.
March 2001
5-78
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
A study of PAHs in indoor and outdoor air was conducted in 14 inner-city and 10 rural
low-income homes near Durham, NC, in two seasons (winter and summer) in 1995 (Chuang
et al., 1999). Fine-particle-bound PAH concentrations measured with a real-time monitor were
usually higher indoors than outdoors (2.47 ± 1.90 versus 0.53 ± 0.58 yug/m3). Higher indoor
levels were seen in smoker's homes compared with nonsmoker's homes, and higher outdoor and
indoor PAH levels were seen in urban areas compared with rural areas.
In a study reported by Dubowsky et al. (1999), the weekday indoor PAH concentrations
attributable to traffic (indoor source contributions were removed) were 39 ± 25 ng/m3 in a
dormitory that had a high air exchange rate because of open windows and doors, 26 ± 25 ng/m3
in an apartment, and 9 ± 6 ng/m3 in a suburban home. The study showed that both
outdoor—especially motor vehicular traffic—and indoor sources contributed to indoor PAH
concentrations.
5.5.4 Factors Affecting Correlations Between Ambient Measurements and
Personal or Microenvironmental Measurements of Particulate Matter
Constituents
The primary factors affecting correlations between personal exposure and ambient air PM
measurements have been discussed in Section 4.3.2. These include air-exchange rates, particle
penetration factors, decay rates and removal mechanisms, indoor air chemistry, and indoor
sources. The importance of these factors varies for different PM constituents. For acid aerosols,
indoor air chemistry is particularly important as indicated by the discussion of the neutralization
of the acidity by ammonia, which is present at higher concentrations indoors because of the
presence of indoor sources. For SVOCs, including PAHs and phthalates, the presence of indoor
sources will impact substantially the correlation between indoor and ambient concentrations
(Ozkaynak et al., 1996b). Penetration factors for PM will impact correlations between indoors
and outdoors for most elements, except Pb, which may have significant indoor sources in older
homes. Indoor air chemistry, decay rates, and removal mechanisms may affect soot and organic
carbon. These factors must be fully evaluated when attempting to correlate ambient, personal,
and indoor PM concentrations.
March 2001
5-79
DRAFT-DO NOT QUOTE OR CITE
-------
1 5.5.5 Limitations of Available Data
2 The previous discussion demonstrates that there is very limited data available that can be
3 used to compare personal, microenvironmental, and ambient air concentrations of PM
4 constituents. Because of resource limitations, PM constituents have not been measured in many
5 studies of PM exposure. Although there is some data on acid aerosols, the comparisons between
6 the personal and indoor data generally have been with outdoor measurements at the participant's
7 residences, not with community ambient air measurement sites. The relationship between
8 personal exposure and indoor levels of acid aerosols is not clear because of the limited database.
9 The exception is sulfate, for which there appears to be a strong correlation between indoor and
10 ambient concentrations.
11 With the exception of PAHs, there are practically no data available to relate personal or
12 indoor concentrations with outdoor or ambient site concentrations of SVOCs, which may be
13 generated from a variety of combustion and industrial sources. The relationship between
14 exposure and ambient concentrations of particles from specific sources, such as diesel engines,
15 has not been determined.
I g Although there is an increasing amount of research being performed to measure PM
17 constituents in different PM size fractions, the current data are inadequate to adequately assess
18 the relationship between indoor and ambient concentrations of most PM constituents.
19
20
21 5.6. IMPLICATIONS OF USING AMBIENT PARTICULATE MATTER
22 CONCENTRATIONS IN EPIDEMIOLOGIC STUDIES OF
23 PARTICULATE MATTER HEALTH EFFECTS
24 m this section, the exposure issues that relate to the interpretation of the findings from
25 epidemiologic studies of PM health effects are examined. This section examines the errors that
26 may be associated with using ambient PM concentrations in epidemiologic analyses of PM health
27 effects. Fkst, implications of associations found between personal exposure and ambient PM
28 concentrations are reviewed. This is discussed separately in the context of either community
29 time-series studies or long-term, cross-sectional studies of chronic effects. Next, the role of
30 compositional and spatial differences hi PM concentrations are discussed and how these may
31 influence the interpretation of findings from PM epidemiology. Finally, using statistical
March 2001
5-80
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
methods, an evaluation of the influence of exposure measurement errors on PM epidemiology
studies is presented.
5.6.1 Potential Sources of Error Resulting from Using Ambient Particulate
Matter Concentrations in Epidemiologic Analyses
Measurement studies of personal exposures to PM are still few and limited in spatial,
temporal, and demographic coverage. Consequently, with the exception of a few longitudinal
panel studies, most epidemiologic studies of PM health effects rely on ambient community
monitoring data giving 24-h average PM concentration measurements. Moreover, because of
limited sampling for PM2 5, many of these epidemiologic studies had to use available PM10or in
some instances had to rely on historic data on other PM measures or indicators, such as TSP,
SO4=, IP15, RSP, COH, KM, etc. A critical question often raised in the interpretation of results
from acute or chronic epidemiologic community-based studies of PM is whether the use of
ambient stationary site PM concentration data influences or biases the findings from these
studies. Because the health outcomes are measured on individuals, the epidemiologists might
prefer to use personal exposure measurements (total, ambient, or nonambient) instead of
surrogates, such as ambient PM concentration measurements collected at one or more ambient
monitoring sites in the community. Use of ambient concentrations could lead to
misclassification of individual exposures and to errors in the epidemiologic analysis of pollution
and health data depending on the pollutant and on the mobility and lifestyles of the population
studied. Ambient monitoring stations can be some distance away from the individuals and can
represent only a fraction of all likely outdoor microenvironments that individuals come in contact
with during the course of their daily lives. Furthermore, most individuals are quite mobile and
move through multiple microenvironments (e.g., home, school, office, commuting, shopping,
etc.) and engage in diverse personal activities at home (e.g, cooking, gardening, cleaning,
smoking). Some of these microenvironments and activities may have different sources of PM
and result in distinctly different concentrations of PM than that monitored by the fixed-site
ambient monitors. Consequently, exposures of some individuals will be classified incorrectly if
only ambient monitoring data are used to estimate individual level exposures to PM. Thus, bias
or loss of precision in the epidemiologic analysis may result from improper assessment of
exposures using data routinely collected by the neighborhood monitoring stations.
March 2001
5-81
DRAFT-DO NOT QUOTE OR CITE
-------
1 Because individuals are exposed to particles in a multitude of indoor and outdoor
2 microenvironments during the course of a day, concern over error introduced in the estimation of
3 PM risk coefficients using ambient, as opposed to personal, PM measurements has received
4 considerable attention recently from exposure analysts, epidemiologists, and biostatisticians.
5 Some exposure analysts contend that, for community time-series epidemiology to yield
6 information on the statistical association of a pollutant with a health response, there must be an
7 association between personal exposure to a pollutant and the ambient concentration of that
8 pollutant because people tend to spend around 90% time indoors and are exposed to both indoor
9 and outdoor-generated PM (cf. Wallace, 2000b; Brown and Paxton, 1998; Ebelt et al., 2000).
10 Consequently, numerous findings reported in the epidemiologic literature on significant
11 associations between ambient PM concentrations and various morbidity and mortality health
12 indices, in spite of the low correlations between ambient PM and concentrations and measures of
13 personal exposure, has been described by some exposure analysts as an exposure paradox
14 (Lachenmyer and Hidy, 2000, Wilson et al., 2000).
15 To resolve the so-called exposure paradox several types of analyses need to be considered.
16 The first type of analysis has to examine the correlations between ambient PM concentrations
17 and personal exposures that are relevant to most of the existing PM epidemiology studies using
18 either pooled, daily-average, or longitudinal exposure data. The second approach has to study the
19 degree of correlations between the two key components of personal PM exposures (i.e.,
20 exposures caused by ambient-generated PM and exposures caused by nonambient PM) with
21 ambient or outdoor PM concentrations, for each of the three types of exposure study design.
22 In addition, several factors influencing either the exposure or health response characterization of
23 the subjects have to be addressed. These include such factors as
24 »spatial variability of PM components,
25 • health or sensitivity status of subjects,
26 • variations of PM with other co-pollutants,
27 • formal evaluation of exposure errors in the analysis of health data, and
28 • how the results may depend on the variations in the design of the epidemiologic study.
29 To facilitate the discussion of these topics, a brief review of concepts pertinent to exposure
30 analysis issues in epidemiology is presented.
31
March 2001
5-82
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
5.6.2 Associations Between Personal Exposures and Ambient Particulate
Matter Concentrations
As defined earlier in Sections 5.3 and 5.4, personal exposures to PM result from an
individual's exposures to PM in many different types of microenvironments (e.g., outdoors near
home, outdoors away from home, indoors at home, indoors at office or school, commuting,
restaurants, malls, other public places, etc.). Total personal exposures (Et) that occur in these
indoor and outdoor microenvironments can be classified as those resulting from PM of outdoor
origin (Eag) and those primarily generated by indoor sources and personal activities (Enonag=
Eig+Epact). The associations between personal exposures and ambient PM concentrations that
have been reported from various personal exposure monitoring studies under three broad
categories of study design: (1) longitudinal, (2) daily-average, or (3) pooled exposure studies are
summarized below.
In the previous Sections 5.4.3.1.2 and 5.4.3.1.3, some of the recent studies conducted
primarily in the United States, involving children, elderly, and subjects with COPD were
reviewed, and they indicated that both intra- and interindividual variability in the relationships
between personal exposures and ambient PM concentrations were observed. A variety of
different physical, chemical, and personal or behavioral factors were identified by the original
investigators that seem to influence the magnitude and the strength of the associations reported.
Clearly, for cohort studies in which individual daily health response are obtained,
individual longitudinal PM personal exposure data (including ambient-generated and nonambient
components) provide the appropriate indicators. In this case, health responses of each individual
can be associated with the total personal exposure, the ambient-generated exposure, or the
nonambient exposure of each individual. Also, the relationships of personal exposure indicators
with ambient concentration can be investigated. In the case of community time-series
epidemiology, however, it is not feasible to obtain experimental measurements of personal
exposure for the millions of people over time periods of years that are needed to investigate the
relationship between air pollution and infrequent health responses such as deaths or even hospital
admissions. The epidemiologist must work with the aggregate number of health responses
occurring each day and a measure of the ambient concentration that is presumed to be
representative of the entire community. The relationship of PM exposures of the potentially
susceptible groups to monitored ambient PM concentrations depends on their activity pattern and
March 2001
5-83
DRAFT-DO NOT QUOTE OR CITE
-------
1 level, residential building and HVAC factors (which influence the infiltration factor), status of
2 exposure to ETS, amount of cooking or cleaning indoors, and seasonal factors, among others.
3 Average personal exposures of these special subgroups to ambient-generated PM are correlated
4 well with ambient PM concentrations regardless of individual variation in the absence of major
5 microenvironmental sources.
6 There seem to be clear differences in the relationships of ambient (Eag) and nonambient
7 (Enonjg) exposure with ambient concentration (CJ. Various researchers have shown that Enonag is
8 independent of Ca, but that Eag is a function of Ca. Wilson et al. (2000) explains the difference
9 based on different temporal patterns that effect PM concentrations. "Concentrations of ambient
10 PM are driven by meteorology and by changes in the emission rates and locations of emission
11 sources, while concentrations of nonambient PM are driven by the daily activities of people."
12 Ott et al. (2000) also discuss the reasons for assuming that Enonag is independent of Eag and
13 Ca. They show that the nonambient component of total personal exposure is uncorrelated with
14 the outdoor concentration data. Ott et al. (2000) show the Enonag is similar for three population-
15 based exposure studies, including two large probability-based studies, the PTEAM study
16 conducted in Riverside (Clayton et al., 1993; Thomas et al., 1993; Ozkaynak et al., 1996a,b) and
17 a study in Toronto (Pelizzarri et al., 1999; Clayton et al., 1999a), as well as a nonprobability-
18 based study, conducted in Phillipsburg (Lioy et al., 1990). Based on these three studies, they
19 conclude that Enonag and the distribution of(Enonag)it can be treated as constant from city to city.
20 Dominici et al. (2000) examined a larger database consisting of five different PM exposure
21 studies and concluded that Enonag can be treated as relatively constant from city to city.
22 If (Enonag), were constant, this would imply that it would have a zero correlation with (Ca)t.
23 However, this hypothesis of constant (Enonag)it has not been established fully because only a few
24 studies have obtained the data needed to estimate (Enonag)it. Although Enonag is independent of
25 Ca, it may not be independent of a. Sarnat et al. (2000) show that Enonag goes up as the
26 ventilation rate (and a) goes down. Lachenmeyer and Hidy (2000) also show, by comparing
27 winter and summer regression equations, that as the slope (a) goes down, the intercept (Enonag)
28 goes up.
29 Mage et al. (1999) assume that the PM!0 concentration component from indoor sources,
30 such as smoking, cooking, cleaning, burning candles, and so on, is not correlated with the
March 2001
5-84
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
outdoor concentration. They indicate that this lack of correlation is expected, because people are
unaware of ambient concentrations and do not necessarily change their smoking or cooking
activities as outdoor PM10 concentrations vary, an assumption supported by other empirical
analyses of personal exposure data. For the PTEAM data set, Mage et al. (1999) have shown that
Eig and Ca have r near zero (R2 = 0.005). Wilson et al. (2000) have shown the Cai and Cig also
have r near zero (R2 = 0.03). Figure 5-11 shows the relationship of estimated (Enonag)it and Enonag
with Ca (calculated by EPA from PTEAM and THEES data).
Based on these results it is reasonable to assume that ordinarily Enonag has no relationship
with Ca. Therefore, in linear nonthreshold models of PM health effects, Enonag is not expected to
contribute to the relative risk determined hi a regression of health responses on Ca. Furthermore,
in time-series analysis of pooled or daily heath data, it is expected that Eag rather than E, will have
the stronger association with Ca.
5.6.3 Role of Compositional Differences in Exposure Characterization for
Epidemiology
The majority of the available data on PM exposures and relationships with ambient PM
have come from a few large-scale studies, such as PTEAM, or longitudinal studies on selected
populations, mostly the elderly. Consequently, for most analyses, exposure scientists and
statisticians had to rely on PM10 or PM25 mass data, instead of elemental or chemical
compositional information on individual or microenvironmental samples. In a few cases,
researchers have examined the factors influencing indoor outdoor ratios or penetration and
deposition coefficients using elemental mass data on personal, indoor, and outdoor PM data (e.g.,
Ozkaynak et al. 1996a,b; Yakovleva et al. 1999). These results have been informative in terms
of understanding relative infiltration of different classes of particle sizes and sources into
residences (e.g., fossil fuel combustion, mobile source emissions, soil-derived, etc.). Clearly, in
the accumulation-mode, particles associated with stationary or mobile combustion sources have
greater potential for penetration into homes and other microenvironments than do crustal
material. The chemical composition of even these broad categories of source classes may have
distinct composition and relative toxicity. Moreover, when particles and reactive gases are
present indoors in the presence of other pollutants or household chemicals, they may react to
form additional or different compounds and particles with yet unknown physical, chemical, and
March 2001
5-85
DRAFT-DO NOT QUOTE OR CITE
-------
200
50
100
150
200
250
«£ 60'
gl
0) O)
Q.S
izf
>
50-
40-
il ^
20'
(b)
50 100 150
Ambient Concentration, pg/m3
200
Figure 5-11. Plots of nonambient exposure to PM10, (a) daytime individual values from
PTEAM data and (b) daily-average values from TREES data.
Source: Data taken from (a) Clayton et al. (1993) and (b) Lioy et al. (1990).
1 toxic composition (Isukapalli et al. 2000). Thus, if indoor-generated and outdoor-generated PM
2 were responsible for different types of health effects, or had significantly different toxicities on a
3 per unit mass basis, it would be then be important that Eag and Enonag should be separated and
4 treated as different species, much like the current separation of PM10 into PM25 and PM10.2 5.
5 These complexities in personal exposure profiles may introduce nonlinearities and other
March 2001
5-86
DRAFT-DO NOT QUOTE OR CITE
-------
1 statistical challenges in the selection and fitting of concentration-response models.
2 Unfortunately, PM health effects models have not yet been able to meaningfully consider such
3 complexities. The relationships of toxicity to the chemical and physical properties of PM are
4 discussed in Chapter 7.
5 It is important also to note that individuals spend time in places other than their homes and
6 outdoors. Many of the interpretations reported in the published literature on factors influencing
7 personal PM]0 exposures, as well as in this chapter, come from the PTEAM study. The PTEAM
8 study was conducted 10 years ago in one geographic location in California, during one season,
9 and most residences had very high and relatively uniform air-exchange rates. Nonhome indoor
10 microenvironments were not monitored directly during the PTEAM study. Commuting
11 exposures from traffic or exposures in a variety of different public places or office buildings
12 could not be assessed directly. Nonresidential buildings may have lower or higher ambient
13 infiltration rates depending on the use and type of the mechanical ventilation systems employed.
14 Because the source and chemical composition of particulate matter effecting personal exposures
15 in different microenvironments vary by season, day-of-the-week, and time of day, it is likely that
16 some degree of misclassification of exposures to PM toxic agents of concern will be introduced
17 when health effects models use only daily-average mass measures such as PM10 or PM2 5.
18 Because of the paucity of currently available data on many of these factors, it is impossible to
19 ascertain at this point the magnitude and severity of these more complex exposure
20 missclassification problems in the interpretation of results from PM epidemiology.
21
22 5.6.4 Role of Spatial Variability in Exposure Characterization for
23 Epidemiology
24 Chapter 3 (Section 3.2.3) and Chapter 5 (Section 5.3) present information on the spatial
25 variability of PM mass and chemical components at fixed-site ambient monitors; for purposes of
26 this chapter, this spatial variability is called an "ambient gradient." Any gradient that may exist
27 between a fixed-site monitor and the outdoor /we near where people live, work, and play,
28 obviously affects the concentration profile actually experienced by people as they go about their
29 daily lives.
30 However, the evidence so far indicates that PM concentrations, especially fine PM (mass
31 and sulfate), generally are distributed uniformly in most metropolitan areas. This reduces the
March 2001
5-87
DRAFT-DO NOT QUOTE OR CITE
-------
1 potential for exposure misclassification because of outdoor spatial gradients when a limited
2 number of ambient PM monitors are used to represent population average ambient exposures in
3 time-series or cross-sectional epidemiologic studies of PM. This topic is further discussed below
4 in Section 5.6.5. However, as discussed earlier, the same assumption is not necessarily true for
5 different components of PM, because source-specific and other spatially nommiform pollutant
6 emissions could alter the spatial profile of individual PM components in a community.
7 For example, particulate and gaseous pollutants emitted from motor vehicles tend to be higher
8 near roadways and inside cars. Likewise, acidic and organic PM species may be location- and
9 time-dependent. Furthermore, human activities are complex, and if outdoor PM constituent
10 concentration profiles are either spatially or temporally variable, it is likely that exposure
11 misclassification errors could be introduced in the analysis of PM air pollution and health data.
12
13 5.6.5 Analysis of Exposure Measurement Error Issues in Particulate Matter
14 Epidemiology
15 The effects of exposure misclassification on relative risk estimates of disease using
16 classical 2x2 contingency design (i.e., exposed/nonexposed versus diseased/nondiseased) have
17 been studied extensively in the epidemiologic literature. It has been shown that the magnitude of
18 the exposure-disease association (e.g., relative risk) because of either misclassification of
19 exposure or disease alone (i.e., nondifferential misclassification) biases the effect results toward
20 the null, and differential misclassification (i.e, different magnitudes of disease misclassification
21 in exposed and nonexposed populations) can bias the effect measure toward or away from the
22 null value relative to the true measure of association (Shy et al., 1978; Gladen and Rogan, 1979;
23 Copeland et al., 1977; Ozkaynak et al., 1986). However, the extension of these results from
24 contingency analysis design to multivariate (e.g., log-linear regression, Poissson, logit) models
25 typically used in recent PM epidemiology has been more complicated. Recently, researchers
26 have developed a framework for analyzing measurement errors typically encountered in the
27 analysis of time-series mortality arid morbidity effects from exposures to ambient PM (cf. Zeger
28 et al., 2000; Dominici et al., 2000; Samet et al., 2000). Some analysis in the context of cross-
29 sectional epidemiology have also been conducted (e.g. Navidi et al., 1999).
30 The appropriateness of using ambient PM concentration as an exposure metric in the
31 context of epidemiologic analysis of health effects associated with exposure to PM recently has
March 2001
5-88
DRAFT-DO NOT QUOTE OR CITE
-------
1 been examined by a number of investigators (cf. Zeger et al., 2000; Dominici et al., 2000; Navidi
2 et al., 1999; Ozkaynak and Spengler, 1996). hi the following section, the error analysis model
3 framework developed in Zeger et al. (2000) will be discussed in the context of time-series
4 epidemiology. After which, issues and implications of exposure errors to findings from long-
5 term/chronic or cross-sectional epidemiology will be discussed briefly.
6
7 5.6.5.1 Analysis of Exposure Measurement Errors in Time-Series Studies
8 Zeger et al. (2000) provide a useful framework for analyzing exposure error in community
9 time-series epidemiology. This framework, coupled with results from recent exposure studies,
10 makes it possible to clarify some important questions regarding relationships among the three
11 aspects of personal exposure (1) total personal, (2) personal caused by ambient PM, and
12 (3) personal resulting from nonambient PM and ambient concentration. Consider the regression
13 of a health response (i.e., mortality rate on day t, Yt, against the ambient concentration of PM on
14 day t, C,). In analyzing pollution-level data on mortality and air pollution, log-linear regressions
15 of the form:
16
17
18
19
20
21
22
23
24
25
26
27
(5-11)
are fit, where Yj is the expected mortality rate; s(t) is an arbitrary but smooth function of time,
introduced to control for the confounding of longer trends and seasonality; Ct, is the average of
multiple monitor measurements of ambient pollution measurement for day t; and u, are other
possible confounders such as temperature and dew point on the same or previous day. Each
coefficient, P, in Equation 5-11 gives the expected change in the health response, Y, because of a
unit change in its corresponding variable.
However, instead of Equation 5-11, Zeger et al. (2000) suggest that the analyst would like
to know the corresponding relationship for personal exposure rather than ambient concentration,
Yt - exp|>0) + EtPE + utfiu ]. (5-12)
Zeger et al. (2000) do not differentiate among the three aspects of personal or community
exposure. To understand the error in P caused by using ambient concentrations instead of
March 2001
5-89
DRAFT-DO NOT QUOTE OR CITE
-------
1 personal exposure in the regression analysis, it is necessary to examine the relationship between
2 Pc, based on a unit change in the ambient concentration, C, and PE, based on a unit change in one
3 of the three aspects of personal exposure, E. In considering the consequences for Pc, as an
4 estimate of PE, of having a measure of ambient pollution C,, rather than actual personal exposure
5 Eit, it is convenient to express the desired pollution measurement, Eit, as Ct plus three error terms:
6
8
9 Here Et represents the daily, community-average personal exposure. The first term,
10 (Eit — Et), is the error resulting from having only aggregated or community-averaged exposure
11 rather than individual-level exposure data. The second term, (Et - Ct), is the difference
12 between the average personal exposure and the true ambient pollutant level, and the third term,
13 (C*, -C,), represents the difference between the true and the measured ambient concentration.
14 In the evaluation of these error terms, two types of measurement error often are considered
15 in the context of epidemiology. The classical error model assumes that measurement error,
16 (Q-Et), depends on ambient measurements [simply referred to as Ct here instead of (Ca)J. The
17 Berkson error model assumes that the measurement error is dependent on the true value or the
18 personal exposure (E,). The regression coefficient (Pc), estimated from the health effects model
19 in the Berkson error case, gives an unbiased estimate of PE. In the classical error case, Pc is a
20 biased estimate of pE, and the degree of bias depends on the correlation between the
21 measurement error and Ct. The measurement error analysis of Zeger et al. (2000) includes three
22 components: (1) an individual's deviation from the risk-weighted average personal exposure;
23 (2) the difference between the average personal exposure and the true ambient level; and (3) the
24 difference between the measured and the true ambient levels, which include the spatial variation
25 of outdoor PM and instrument sampling error. Zeger et al. (2000) conclude that the first and
26 third components are of the Berkson type and, therefore, are likely to have smaller effects on the
27 relative risk estimates for PM. However, the second component can be a source of substantial
28 bias if, for example, there are short-term associations of the contributions of indoor sources with
29 ambient concentrations. However, recent analysis of PTEAM data (Mage et al., 2000) and
30 theoretical considerations (Ott et al., 2000) indicate that it is unlikely that nonambient exposures
March 2001
5-90
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
will be correlated with the ambient concentration. Therefore, this type of bias is unlikely.
However, if the community average exposure to ambient PM is less than the ambient
concentration, the risk regression coefficient, pc, will be biased low. According to Carrol (1995),
Pc = cc PE, where Pc is the percentage increase in risk because of a unit increase in ambient
concentration, and PE is the estimated percentage increase in risk because of a unit increase in the
community-average personal exposure to ambient PM. Both Zeger et al. (2000) and Dominici
et al. (2000) examine the nature of error with this second component. Both of these analyses
conclude that the error introduced because of measured differences between the average personal
exposure and ambient levels can bias the regression coefficients. In both cases they find the Pc is
close to a PE.
This framework analysis demonstrates the importance of the daily community-average
exposure, Et, in community time-series epidemiology. It is Et, not the random, pooled values of
Ej t, that need to have a statistically significant correlation with Ct for proper interpretation of
community time-series epidemiology studies based on ambient monitoring data, as discussed
further in Wilson et al. (2000) and Mage et al. (1999).
A critical assumption in the above analysis is that the risk varies linearly with C or E (i.e.,
Pc and PE are constant). This assumption does not permit a threshold (a concentration below
which there is no effect). It also includes the assumption that the appropriate metric for
determination of a health response is the 24-h average PM mass concentration. Zeger et al.
(2000) show that the likely consequence of using ambient concentrations instead of the risk-
weighted average personal exposure measures is to underestimate the pollution effects.
According to Zeger et al. (2000) the largest biases in inferences about the mortality-personal
exposure relative risk will occur because of more complex errors between ambient concentration
and daily-average personal exposure measures. It is important to note that both the Zeger et al.
(2000) and the Dominici et al. (2000) error analyses used personal PM10data from the PTEAM
study data. However, effects of measurement error estimates may differ by particle size and
composition. It is possible that PM25 , ultrafine particle measures, or another component of PM,
may better reflect personal exposures to PM of outdoor origin. Finally, the seasonal or temporal
variations in the measurement errors and correlations between different PM concentration
measures and co-pollutants (e.g. SO2, CO, NO2, O3) could influence the error analysis results
reported by the investigators cited above.
March 2001
5-91
DRAFT-DO NOT QUOTE OR CITE
-------
1 5.6.5.2 Analysis of Exposure Measurement Errors in Long-Term Epidemiology Studies
2 The Six Cities (Dockery et al., 1993) and ACS (Pope et al., 1995) studies have played an
3 important role in assessing the health effects from long-term exposures to particulate pollution.
4 Even though these studies often have been considered as chronic epidemiologic studies, it is not
5 easy to differentiate the role of historic exposures from those of recent exposures on chronic
6 disease mortality. In the Six Cities study, fine particles and sulfates were measured at the
7 community level, and the final analysis of the database used six city-wide average ambient
8 concentration measurements. This limitation also applies to the ACS study but has less impact
9 because of the larger number of cities considered in that study. In a HEI-sponsored reanalysis of
10 the Six Cities and the ACS data sets, Krewski et al. (2000) attempted to examine some of the
11 exposure misclassification issues either analytically or through sensitivity analysis of the
12 aerometric and health data. The HEI reanalysis project also addressed exposure measurement
13 error issues related to the Six Cities study. For example, the inability to account for exposures
14 prior to the enrollment of the cohort, hampered accurate interpretation of the relative risk
15 estimates in terms of acute versus chronic causes. Although the results seem to suggest past
16 exposures are more strongly associated with mortality than recent exposures, the measurement
17 error for long-term averages could be higher, thus influencing these interpretations. For example,
18 Krewski et al. (2000), using the individual mobility data available for the Six Cities cohort,
19 analyzed the mover and nonmover groups separately. The relative risk of fine particle effects on
20 all-cause mortality was shown to be higher for the nonmover group than for the mover group,
21 suggesting the possibility of higher exposure misclassification biases for the movers. The issue
22 of using selected ambient monitors in the epidemiologic analyses also was investigated by the
23 ACS and Six Cities studies reanalysis team. Krewski et al.(2000) presented the sensitivity of
24 results to choices made in selecting stationary or mobile-source-oriented monitors. For the ACS
25 study, reanalysis of the sulfate data using only those monitors designated as residential or urban,
26 and excluding sites designated as industrial, agricultural, or mobile did not change the risk
27 estimates appreciably. On the other hand, application of spatial analytic methods designed to
28 control confounding at larger geographic scales (i.e., between cities) caused changes in the
29 particle and sulfate risk coefficients. Spatial adjustment may account for differences in pollution
30 mix or PM composition, but many other cohort-dependent risk factors will vary across regions or
31 cities in the United States. Therefore, it is difficult to interpret these findings solely in terms of
March 2001
5-92
DRAFT-DO NOT QUOTE OR CITE
-------
1 spatial differences in pollution composition or relative PM toxicity until further research is
2 concluded.
3 Another study that has examined the influence of measurement errors in air pollution
4 exposure and health effects assessments is the one reported by Navidi et al. (1999). This study
5 developed techniques to incorporate exposure measurement errors encountered in long-term air
6 pollution health effects studies and tested them on the data from the University of Southern
7 California Children's Health Study conducted in 12 communities in California. These
8 investigators developed separate error analysis models for direct (i.e., personal sampling) and
9 indirect (i.e., microenvironmental) personal exposure assessment methods. These models were
10 generic to most air pollutants, but a specific application was performed using a simulated data set
11 for studying ozone health effects on lung function decline in children. Because the assumptions
12 made in their microenvironmental simulation modeling framework were similar to those made in
13 estimating personal PM exposures, it is useful to consider the conclusions from Navidi et. al.
14 (1999). According to Navidi et al. (1999), neither the microenvironmental nor the personal
15 sampler method produces reliable estimates of the exposure-response slope (for O3) when
16 measurement error is uncorrected. Because of nondifferential measurement error, the bias was
17 toward zero under the assumptions made in Navidi et al. (1999) but could be away from zero if
18 the measurement error was correlated with the health response. A simulation analysis indicated
19 that the standard error of the estimate of a health effect increases as the errors in exposure
20 assessment increase (Navidi et al., 1999). According to Navidi et al. (1999), when a fraction of
21 the ambient level in a microenvironment is estimated with a standard error of 30%, the standard
22 error of the estimate is 50% higher than it would be if the true exposures were known. It appears
23 that errors in estimating ambient PM indoor/ambient PM outdoor ratios have much more
24 influence on the accuracy of the microenvironmental approach than do errors in estimating time
25 spent in these microenvironments.
26
27 5.6.5.3 Conclusions from Analysis of Exposure Measurement Errors on Particulate Matter
28 Epidemiology
29 Personal exposures to PM are influenced by a number of factors and sources of PM located
30 in both indoor and outdoor microenvironments. However, PM resulting from ambient sources
31 does penetrate into indoor environments, such as residences, offices, public buildings, etc., in
March 2001
5-93
DRAFT-DO NOT QUOTE OR CITE
-------
1 which individuals spend a large portion of their daily lives. The correlations between total
2 personal exposures and ambient or outdoor PM concentrations can vary depending on the relative
3 contributions of indoor PM sources to total personal exposures. Panel studies of both adult and
4 young subjects have shown that, in fact, individual correlations of personal exposures with
5 ambient PM concentrations could vary person to person, and even day to day, depending on the
6 specific activities of each person. Separation of PM exposures into two components,
7 ambient-generated PM and nonambient PM, would reduce uncertainties in the analysis and
8 interpretation of PM health effects data. Nevertheless, because ambient-generated PM is an
9 integral component of total personal exposures to PM, statistical analysis of cohort-average
10 exposures are strongly correlated with ambient PM concentrations when the size of the
11 underlying population studied is large. Using the PTEAM study data, analysis of exposure
12 measurement errors, in the context of tune-series epidemiology, also has shown that errors or
13 uncertainties introduced by using surrogate exposure variables, such as ambient PM
14 concentrations, could lead to biases in the estimation of health risk coefficients. These then
15 would need to be corrected by suitable calibration of the PM health risk coefficients.
16 Correlations between the PM exposure variables and other covariates (e.g., gaseous
17 co-pollutants, weather variables, etc.) also could influence the degree of bias in the estimated PM
18 regression coefficients. However, most time-series regression models employ seasonal or
19 temporal detrending of the variables, thus reducing the magnitude of this cross-correlation
20 problem (Ozkaynak and Spengler 1996).
21 Ordinarily, exposure measurement errors are not expected to influence the interpretation of
22 findings from either the cross-sectional or time-series epidemiologic studies that have used
23 ambient concentration data if they include sufficient adjustments for seasonality and key
24 confounders. Clearly, there is no question that better estimates of exposures to components of
25 PM of health concern are beneficial. Composition of PM may vary in different geographic
26 locations and different exposure microenvironments. Compositional and spatial variations could
27 lead to further errors in using ambient PM measures as surrogates for exposures to PM. Even
28 though the spatial variability of PM (PM2 5 in particular) mass concentrations in urban
29 environments seems to be small, the same conclusions drawn above regarding the influence of
30 measurement errors may not necessarily hold for all of the PM toxic components. Again, the
31 expectation based on statistical modeling considerations is that these exposure measurement
March 2001
5-94
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
errors or uncertainties will most likely reduce the statistical power of the PM health effects
analysis, making it difficult to detect a true underlying association between the correct exposure
metric and the health outcome studied. However, until more data on exposures to toxic agents of
PM become available, existing studies on PM exposure measurement errors must be relied on;
these indicate that use of ambient PM concentrations as a surrogate for exposures is not expected
to change the principal conclusions from PM epidemiologic studies, utilizing community average
health and pollution data.
5.7 SUMMARY OF KEY FINDINGS AND LIMITATIONS
Exposure Definitions and Components
• Personal exposure (E) to PM mass or its constituents results when individuals come in contact
with particulate pollutant concentrations (C) in locations or microenvironments (jj,e) that they
frequent during a specific period of time. Various PM exposure metrics can be defined
according to its source (i.e., ambient, nonambient) and the microenvironment where exposure
occurs.
• Personal exposure to PM results from an individual's exposure to PM in many different types
of microenvironments (e.g., outdoors near home, outdoors away from home, indoors at home,
indoors at office or school, commuting, restaurants, malls, other public places, etc.). Thus, total
daily exposure to PM for a single individual (E,) can be expressed as the sum of various
microenvironmental exposures that the person encounters during the course of a day.
• In a given fte, particles may originate from a wide variety of sources. In an indoor
microenvironment, PM may be generated from within as a result of PM generating activities
(e.g., cooking, cleaning, smoking, resuspending PM from PM resulting from both indoor and
outdoor sources that had settled out), from outside (outdoor PM entering through cracks and
openings in the structure), and from the chemical interaction of pollutants from outdoor air with
indoor-generated pollutants.
• The total daily exposure to PM for a single individual (Et) also can be expressed as the sum of
contributions of ambient-generated (Eag) and nonambient-generated (Enonag) PM (i.e.,
E = Eag + Enonag). Enonag, in turn, is composed of PM generated by indoor sources (Eig ) and PM
generated by personal activities (Epact) (i.e., Enonag = Eig + Epact). Eag is composed of exposures to
March 2001 5-95 DRAFT-DO NOT QUOTE OR CITE
-------
1 ambient PM concentrations while outdoors, Ca Afa, and ambient PM that has infiltrated
/
2 indoors, JJ CflI.Af, while indoors (i.e., Eag = JJ CaAffl + £ q,,Af, ).
» / /
3 * Exposure models are useful tools for examining the importance of sources, microenvironments,
4 and physical and behavioral factors that influence personal exposures to PM. However,
5 development and evaluation of population exposure models for PM and its components has
6 been limited. Improved modeling methodologies and new model input data are needed.
7
8 Factors Affecting Concentrations and Exposures to Participate Matter
9 • Concentrations of PM indoors are affected by several factors and mechanisms: ambient
10 concentrations outdoors; air exchange rates; particle penetration factors; particle production
11 from indoor sources and indoor air chemistry; and indoor particle decay rates and removal
12 mechanisms caused by physical processes or resulting from mechanical filtration, ventilation or
13 air-conditioning devices.
14 • Average personal exposures to PM mass and its constituents are influenced by
15 microenvironmental PM concentrations and by how much tune is spent by each individual in
16 these various indoor and outdoor microenvironments. Nationwide, individuals, on average,
17 spend nearly 90% of their time indoors (at home and in other indoor locations) and about 6% of
18 their tune outdoors.
19 • The relative size of personal exposure to ambient-generated PM relative to nonambient-
20 generated PM depends on the ambient concentration, the infiltration rate of outdoor PM into
21 indoor microenvironments, the amount of PM generated indoors (e.g., ETS, cooking and
22 cleaning emissions), and the amount of PM generated by personal activity sources. Infiltration
23 rates primarily depend on air-exchange rate, size-dependent particle penetration across the
24 building membrane, and size-dependent removal rates. All of these factors vary over time and
25 across subjects and building types.
26 • The relationship between PM exposure and health outcome could depend on the concentration,
27 composition, and toxicity of the PM originating from different sources. Application of source
28 apportionment techniques to ambient, indoor, and personal PM composition data have
29 identified the following general source categories of importance: outside soil, resuspended
30 indoor soil, indoor soil, personal activities, sea-salt, motor vehicles, nonferrous metal smelters,
31 and secondary sulfates.
March 2001 5-96 DRAFT-DO NOT QUOTE OR CITE
-------
1 • There have been only a limited number of studies that have measured the physical and
2 chemical constituents of PM in personal or microenvironmental samples. Available data on
3 PM constituents indicate that
4 - personal and indoor sulfate measurements often are correlated highly with outdoor and
5 ambient sulfate concentration measurements;
6 - for acid aerosols, indoor air chemistry is particularly important because of the
7 neutralization of the acidity by ammonia, which is present at higher concentrations
8 indoors because of the presence of indoor sources of ammonia;
9 - for S VOCs, including PAHs and phthalates, the presence of indoor sources will
10 substantially impact the relation betwebn indoor and ambient concentrations;
11 - penetration and decay rates are a functions of size and will cause variations in the
12 attenuation factors as a function of particle size; infiltration rates will be higher for PM,
13 and PM2 5 than for PM10, PM10.2 5 or ultrafine particles; and
14 - Indoor air chemistry may increase indoor concentrations of organic PM.
15 • Even though there is an increasing amount of research being performed to measure PM
16 constituents in different PM size fractions, with few exceptions (i.e., sulfur or sulfates), the
17 current data are inadequate to adequately assess the relationship between personal, indoor, and
18 ambient concentrations of most PM constituents.
19
20 Correlations Between Personal Exposures, Indoor, Outdoor, and Ambient Measurements
21 • Most of the available personal data on PM measurements and information on the relationships
22 between personal and ambient PM come from a few large-scale studies, such as the PTEAM
23 study, or the longitudinal panel studies, which have been conducted on selected populations,
24 such as the elderly.
25 • Panel and cohort studies that have measured PM exposures and concentrations typically have
26 reported their results in terms of three types of correlations: (1) longitudinal, (2) pooled, and
27 (3) daily-average correlations between personal and ambient or outdoor PM.
28 • The type of correlation analysis performed can have a substantial effect on the resulting
29 correlation coefficient. Low correlations with ambient concentrations could result when people
30 with very different nonambient exposures are pooled, even though temporally, their individual
31 personal exposures may be correlated highly with ambient concentrations.
March 2001
5-97
DRAFT-DO NOT QUOTE OR CITE
-------
1 • Recent studies conducted by EPA of the elderly subjects living in a retirement facility in
2 Baltimore and a group of elderly living in Fresno produced higher correlation coefficients
3 between personal and ambient PM for daily-average correlations compared to longitudinal
4 correlations. This supports earlier analyses showing the daily-average correlations are higher
5 than pooled correlations.
6 • Longitudinal and pooled correlations between personal exposure and ambient or outdoor PM
7 concentrations reported by various investigators varied considerably among the different
8 studies and in each study between the study subjects. Most studies report longitudinal
9 correlation coefficients that range from close to zero to near one, indicating that individual's
10 activities and residence type may have a significant effect on total personal exposures to PM.
11 • Longitudinal studies that measured sulfate found high correlations between personal and
12 ambient sulfate.
13 • In general, probability-based population studies tend to show low pooled correlations because
14 of the high differences in levels of nonambient PM generating activities from one subject to
15 another. In contrast, the absence of indoor sources for the populations in several of the
16 longitudinal panel studies resulted in high correlations between personal exposure and ambient
17 PM within subjects over time for these populations. But even for these studies, correlations
18 varied by individual depending on their activities and on the microenvironments that they
19 occupied.
20
21 Potential Sources of Error Resulting from Using Ambient Particulate Matter
22 Concentrations in Epidemiologic Analyses
23 • There is, as yet, no clear consensus among exposure analysts as to how well ambiently
24 measured PM concentrations represent a surrogate for personal exposure to total PM or to
25 ambient-generated PM.
26 • Measurement studies of personal exposures to PM are still few and limited in spatial, temporal,
27 and demographic coverage. Consequently, with the exception of a few longitudinal panel
28 studies, most epidemiologic studies on PM health effects have relied on daily-average PM
29 concentration measurements obtained from ambient community monitoring data as a surrogate
30 for the exposure variable.
March 2001
5-98
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
• Because individuals are exposed to particles in a multitude of indoor and outdoor
microenvironments during the course of a day, concerns over error introduced in the estimation
of PM risk coefficients using ambient, as opposed to personal PM measurements, have been
raised.
• Total personal exposures to PM could vary from person to person, and even day to day,
depending on the specific activities of each person. Separation of PM exposures into two
components, ambient-generated PM and nonambient-generated PM, would reduce potential
uncertainties in the analysis and interpretation of PM health effects data.
• Available data indicate that PM mass concentrations, especially fine PM, typically are
distributed uniformly in most metropolitan areas, thus reducing the potential for exposure
^classification because of spatial variability when a limited number of ambient PM monitors
are used to represent population average ambient exposures in community time-series or
long-term, cross-sectional epidemiologic studies of PM.
• Even though the spatial variability of PM (in particular, PM2.5) mass concentrations in urban
environments seems to be small, the same conclusions drawn above regarding the influence of
measurement errors may not necessarily hold for all of the PM components.
• There are important differences in the relationship of ambient PM concentrations (CJ with
exposures to ambient PM (Eag), and with exposures to nonambient PM (Enonag). Various
researchers have shown that Eag is a function of Ca, and that concentrations of ambient PM are
driven by meteorology, by changes in source emission rates, and in locations of emission
sources relative to the measurement site. However, Enonag is independent of Ca, because
concentrations of nonambient PM are driven by the daily activities of people.
• Because personal exposures also include a contribution from ambient concentrations, the
correlation between daily-average personal exposure and the daily-average ambient
concentration increases as the number of subjects measured daily increases. An application of
a Random Component Superposition (RCS) model has shown that the contributions of ambient
PM10 and indoor-generated PM10 to community mean exposure can be decoupled in modeling
urban population exposure distributions.
• If linear nonthreshold models are assumed in time-series analysis of daily-average ambient PM
concentrations and community health data, Enonag is not expected to contribute to the relative
risk estimates determined by regression of health responses on Ca.
March 2001
5-99 DRAFT-DO NOT QUOTE OR CITE
-------
1 • Using the PTEAM study data, analysis of exposure measurement errors in the context of
2 time-series epidemiology has shown that errors or uncertainties introduced by using surrogate
3 exposure variables, such as ambient PM concentrations, could lead to biases in the estimation
4 of health risk coefficients.
5 - Because sources and chemical composition of particulate matter affecting personal exposures in
6 different microenvironments vary, by season, day-of-the-week, and time of day, it is likely that
7 some degree of misclassification of exposures to PM toxic agents of concern will be introduced
8 when health effects models use only daily-average mass measures such as PM10 or PM2 5.
9 Because of the paucity of currently available data on many of these factors, it is impossible to
10 ascertain at this point the significance of these more complex exposure misclassification
11 problems in the interpretation of results from PM epidemiology.
12 - Exposure measurement errors may depend on particle size and composition. PM2.5 better
13 reflects personal exposure to PM of outdoor origin than PM10. It is possible that various
14 ultrafme particle measures, or other components of PM may be better exposure indicators for
15 epidemiologic studies.
16 • Seasonal or temporal variations in the measurement errors and their correlations between
17 different PM concentration measures and co-pollutants (e.g., SO2, CO, NO2, O3) could
18 influence the error analysis results but not likely the interpretation of current findings.
19 • Ordinarily, PM exposure measurement errors are not expected to influence the interpretation of
20 findings from either the community time-series or long-term epidemiologic studies that have
21 used ambient concentration data if they include sufficient adjustments for seasonality and key
22 personal and geographic confounders.
23 • To reduce exposure misclassification errors hi PM epidemiology, conducting new cohort
24 studies of sensitive populations with better real-time techniques for exposure monitoring and
25 further speciation of indoor-generated, ambient, and personal PM mass are essential.
26 • Based on statistical modeling considerations, it is expected that existing PM exposure
27 measurement errors or uncertainties most likely will reduce the statistical power of the PM
28 health effects analysis, thus making it difficult to detect a true underlying association between
29 the correct exposure metric and the health outcome studied.
March 2001
5-100
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
Currently available studies on PM exposure measurement errors indicate that use of ambient
PM concentrations as a surrogate for personal exposures is not expected to change the key
conclusions derived from most of the recent epidemiologic studies on PM health effects.
March 2001
5-101
DRAFT-DO NOT QUOTE OR CITE
-------
1
REFERENCES
2 Abt, E.; Suh, H. H.; Allen, G.; Koutrakis, P. (2000a) Characterization of indoor particle sources: a study conducted
3 in the metropolitan Boston area. Environ. Health Perspect. 108: 35-44.
4 Abt, E.; Suh, H. H.; Catalano, P.; Koutrakis, P. (2000b) The relative contribution of outdoor and indoor particle
5 sources to indoor concentrations. Environ. Sci. Technol. 34: 3579-3587.
6 Akland, G. G.; Hartwell, T. D.; Johnson, T. R.; Whitmore, R. W. (1985) Measuring human exposure to carbon
7 monoxide in Washington, D.C., and Denver, Colorado, during the winter of 1982-1983. Environ. Sci.
8 Technol. 19:911-918.
9 Al-Raheem, M. Y. A.; El-Desouky, M.; Mage, D. T.; Kollander, M.; Wallace, L. A.; Spengler, J. D. (2000)
10 Personal exposures to PM and gaseous air pollutants at the end of the 1991 Kuwait oil fires. Presented at:
11 PM2000: particulate matter and health—the scientific basis for regulatory decision-making, specialty
12 conference & exhibition; January; Charleston, SC. Pittsburgh, PA: Air & Waste Management Association.
13 Aim, S.; Jantunen, M. J.; Vartiainen, M. (1999) Urban commuter exposure to particle matter and carbon monoxide
14 inside an automobile. J. Exposure Anal. Environ. Epidemiol. 9: 237-244.
15 Anuszewski, J.; Larson, T. V.; Koenig, J. Q. (1998) Simultaneous indoor and outdoor particle light-scattering
16 measurements at nine homes using a portable nephelometer. J. Exposure Anal. Environ. Epidemiol.
17 8:483-493.
18 Back, S.-O.; Kim, Y.-S.; Perry, R. (1997) Indoor air quality in homes, offices and restaurants in Korean urban
19 areas-indoor/outdoor relationships. Atmos. Environ. 31: 529-544.
20 Bahadori, T.; Suh, H. H.; Rojas-Bracho, L.; Koutrakis, P. (2001) Personal exposure to particulate matter of
21 individuals with chronic obstructive pulmonary disease (COPD). J. Expos. Anal. Environ. Epidemiol.:
22 Accepted.
23 Beyea, J.; Hatch, M. (1999) Geographic exposure modeling: a valuable extension of geographic information
24 systems for use in environmental epidemiology. Environ. Health Perspect. Suppl. 107(1): 181-190.
25 Bohadana, A. B.; Massin, N.; Wild, P.; Toamain, J.-P.; Engel, S.; Goutet, P. (2000) Symptoms, airway
26 responsiveness, and exposure to dust in beech and oak wood workers. Occup. Environ. Med. 57: 268-273.
27 Brauer, M.; 't Mannetje, A. (1998) Restaurant smoking restrictions and environmental tobacco smoke exposure.
28 Am. J. Public Health 88:1834-1836.
29 Brauer, M.; Koutrakis, P.; Spengler, J. D. (1989) Personal exposures to acidic aerosols and gases. Environ. Sci.
30 Technol. 23:1408-1412.
31 Brauer, M.; Koutrakis, P.; Keeler, G. J.; Spengler, J. D. (1990) Indoor and outdoor concentrations of acidic aerosols
32 and gases. In: Indoor air '90: precedings of the 5th international conference on indoor air quality and climate,
33 volume 2, characteristics of indoor air; July-August; Toronto, ON, Canada. Ottawa, ON, Canada:
34 International Conference on Indoor Air Quality and Climate, Inc.; pp. 447-452.
35 Brauer, M.; Bartlett, K.; Regalado-Pineda, J.; Perez-Padilla, R. (1996) Assessment of particulate concentrations
36 from domestic biomass combustion in rural Mexico. Environ. Sci. Technol. 30: 104-109.
37 Brauer, M.; Hirtle, R. D.; Hall, A. C.; Yip, T. R. (1999) Monitoring personal fine particle exposure with a particle
38 counter. J. Exposure Anal. Environ. Epidemiol. 9:228-236.
39 Brown, K. G.; Paxton, M. B. (1998) Predictability of personal exposure to airborne particulate matter from ambient
40 concentrations in four communities. In: Vostal, J. J., ed. Health effects of particulate matter in ambient air.
41 Proceedings of an international conference; 1997; Prague, Czech Republic. Pittsburgh, PA: Air & Waste
42 Management Association; pp. 333-350. (A&WMA publication VIP-80).
43 Burke, J. M.; Zufall, M. J.; Ozkaynak, H. A.; Xue, J.; Zidek, J. V. (2000) Predicting population exposures to PM:
44 the importance of microenvironmental concentrations and human activities. Presented at: PM2000:
45 particulate matter and health~the scientific basis for regulatory decision-making, specialty conference &
46 exhibition; January; Charleston, SC. Pittsburgh, PA: Air & Waste Management Association.
47 Burke, J.; Zufall, M.; Ozkaynak, H. (2001) A population exposure model for particulate matter: case study results
48 for PM2.5 in Philadelphia, PA. J. Expos. Anal. Environ. Epidemiol.: submitted.
49 Burton, R. M.; Suh, H. H.; Koutrakis, P. (1996) Spatial variation in particulate concentrations within metropolitan
50 Philadelphia. Environ. Sci. Technol. 30: 400-407.
51 Buzorius, G.; Hameri, K.; Pekkanen, J.; Kulmala, M. (1999) Spatial variation of aerosol number concentration in
52 Helsinki city. Atmos. Environ. 33: 553-565.
53 Byrne, M. A.; Lange, C.; Goddard, A. J. H.; Roed, J. (1992) Indoor aerosol deposition studies using
54 neutron-activatable tracers. J. Aerosol Sci. 23(suppl. 1): S543-S546.
March 2001
5-102
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Calder, K. L. (1957) A numerical analysis of the protection afforded by buildings against B W aerosol attack.
Fort Detrick, MD: Office of the Deputy Commander for Scientific Activities; BWL tech. study no. 2.
Carrer, P.; Alcini, D.; Cavallo, D.; Bollini, D.; Ghalandar, R.; Lovato, L.; Ploia, E.; Vercelli, F.; Visigalli, F.;
Maroni, M. (1998) Personal exposure to PM10 of subjects living in Milan. In: Vostal, J. J., ed. Health effects
of paniculate matter in ambient air. Proceedings of an international conference; 1997; Prague, Czech
Republic. Pittsburgh, PA: Air & Waste Management Association; pp. 359-365. (A&WMA publication •
VIP-80).
Carroll, R. J.; Ruppert, D.; Stefanski, L. A. (1995) Measurement error in nonlinear models. London, United
Kingdom: Chapman & Hall. (Cox, D. R.; Hinkley, D. V.; Keiding, N.; Reid, N.; Rubin, D. B.; Silverman,
B. W., eds. Monographs on statistics and applied probability: v. 63).
Cha, Q.; Chen, Y.-Z.; Du, Y.-X. (1996) Influence of indoor air pollution on quality of outdoor air in city of
Guangzhou, China. In: Indoor air '96: the 7th international conference on indoor air quality, volume I; July;
Nagoya, Japan. Tokyo, Japan: Institute of Public Health; pp. 1131 -1135.
Chang, L.-T.; Koutrakis, P.; Catalano, P. J.; Suh, H. H. (2000) Hourly personal exposures to fine particles and
gaseous pollutants - results from Baltimore, Maryland. J. Air Waste Manage. Assoc. 50: 1223-1.235.
Chuang, J. C.; Callahan, P. J.; Lyu, C. W.; Wilson, N. K. (1999) Polycyclic aromatic hydrocarbon exposures of
children in low-income families. J. Exposure Anal. Environ. Epidemiol. 2: 85-98.
Clayton, C. A.; Perritt, R. L.; Pellizzari, E. D.; Thomas, K. W.; Whitmore, R. W.; Wallace, L. A.; Ozkaynak, H.;
Spengler, J. D. (1993) Particle total exposure assessment methodology (PTEAM) study: distributions of
aerosol and elemental concentrations in personal, indoor, and outdoor air samples in a southern California
community. J. Exposure Anal. Environ. Epidemiol. 3: 227-250.
Clayton, C. A.; Pellizzari, E. D.; Rodes, C. E.; Mason, R. E.; Piper, L. L. (1999a) Estimating distributions of
long-term particulate matter and manganese exposures for residents of Toronto, Canada. Atmos Environ
33:2515-2526.
Clayton, C. A.; Pellizzari, E. D.; Whitmore, R. W.; Perritt, R. L., Quackenboss, J. J. (1999b) National human
exposure assessment survey (NHEXAS): distributions and associations of lead, arsenic, and volatile organic
compounds in EPA Region 5. J. Exposure Anal. Environ. Epidemiol. 9: 381-392.
Cohen, B. S.; Li, W.; Xiong, J. Q.; Lippmann, M. (2000) Detecting H+ in ultrafme ambient aerosol using iron
nano-film detectors and scanning probe microscopy. Appl. Occup. Environ. Hyg. 15: 80-89.
Conner, T. L.; Lonneman, W. A.; Seila, R. L. (1995) Transportation-related volatile hydrocarbon source profiles
measured in Atlanta. J. Air Waste Manage. Assoc. 45: 383-394.
Copeland, K. T.; Checkoway, H.; McMichael, A. J.; Holbrook, R. H. (1977) Bias due to misclassification in the
estimation of relative risk. Am. J. Epidemiol. 105: 488-495.
Crump, K. S. (2000) Manganese exposures in Toronto during use of the gasoline additive, methylcyclopentadienyl
manganese tricarbonyl. J. Exposure Anal. Environ. Epidemiol. 10: 227-239.
Daisey, J. M.; Mahanama, K. R. R.; Hodgson, A. T. (1998) Toxic volatile organic compounds in simulated
enviromental tobacco smoke: emission factors for exposure assessment. J. Exposure Anal. Environ
Epidemiol. 8: 313-334.
Dockery, D. W.; Spengler, J. D. (1981) Indoor-outdoor relationships of respirable sulfates and particles.
Atmps. Environ. 15: 335-343.
Dockery, D. W.; Pope, C. A., Ill; Xu, X.; Spengler, J. D.; Ware, J. H.; Fay, M. E.; Ferris, B. G., Jr.; Speizer, F. E.
(1993) An association between air pollution and mortality in six U.S. cities. N. Engl. J Med
329: 1753-1759.
Dominici, F.; Zeger, S. L.; Samet, J. (2000) A measurement error model for time-series studies of air pollution and
mortality. Biostatistics 1: 157-175.
Donham, K. J.; Cumro, D.; Reynolds, S. J.; Merchant, J. A. (2000) Dose-response relationships between
occupational aerosol exposures and cross-shift declines of lung function in poultry workers:
recommendations for exposure limits. J. Occup. Environ. Med. 42: 260-269.
Duan, N. (1982) Models for human exposure to air pollution. Environ. Int. 8: 305-309.
Duan, N. (1991) Stochastic microenvironment models for air pollution exposure. J. Exp. Anal. Environ. Epidemiol
1:235-257.
Dubowsky, S. D.; Wallace, L. A.; Buckley, T. J. (1999) The contribution of traffic to indoor concentrations of
polycyclic aromatic hydrocarbons. J. Exposure Anal. Environ. Epidemiol. 9: 312-321.
March 2001
5-103
DRAFT-DO NOT QUOTE OR CITE
-------
1 Ebelt, S. T.; Petkau, A. J.; Vedal, S.; Fisher, T. V.; Brauer, M. (2000) Exposure of chronic obstructive pulmonary
2 disease patients to particulate matter: relationships between personal and ambient air concentrations. J. Air
3 Waste Manage. Assoc. 50:1081-1094.
4 Esmen, N. A.; Hall, T. A. (2000) Theoretical investigation of the interrelationship between stationary and personal
5 sampling in exposure estimation. Appl. Occup. Environ. Hyg. 15: 114-119.
6 Evans, G. F.; Highsmith, R. V.; Sheldon, L. S.; Suggs, J. C.; Williams, R. W.; Zweidinger, R. B.; Creason, J. P.;
7 Walsh, D.; Rodes, C. E.; Lawless, P. A. (2000) The 1999 Fresno particulate matter exposure studies:
8 comparison of community, outdoor, and residential PM mass measurements. J. Air Waste Manage. Assoc.
9 50: 1887-1896.
10 Ferris, B. G., Jr.; Speizer, F. E.; Spengler, J. D.; Dockery, D.; Bishop, Y. M. M.; Wolfson, M.; Humble, C. (1979)
11 Effects of sulfur oxides and respirable particles on human health: methodology and demography of
12 populations in study. Am. Rev. Respir. Dis. 120: 767-779.
13 Fogh, C. L.; Byrne, M. A.; Roed, J.; Goddard, A. J. H. (1997) Size specific indoor aerosol deposition measurements
14 and derived I/O concentrations ratios. Atmos. Environ. 31: 2193-2203.
15 Freijer, J. I.; Bloemen, H. J. T. (2000) Modeling relationships between indoor and outdoor air quality. J. Air Waste
16 Manage. Assoc. 50:292-300.
17 Frey, H. C.; Rhodes, D. S. (1996) Characterizing, simulating and analyzing variability and uncertainty:
18 an illustration of methods using an air toxic emissions example. Hum. Ecol. Risk Assess. 2: 762-797.
19 Gladen, B.; Rogan, W. J. (1979) Misclassification and the design of environmental studies. Am. J. Epidemiol.
20 109:607-616.
21 Glen, G.; Lakkadi, Y.; Tippett, J. A.; del Valle-Torres, M. (1997) Development of NERL/CHAD: the National
22 Exposure Research Laboratory consolidated human activity database. Research Triangle Park, NC: U.S.
23 Environmental Protection Agency, Office of Research and Development; contract no. 68-D5-0049.
24 Goldstein, B. R.; Tardiff, G.; Hoffnagle, G.; Kester, R. (1992) Valdez health study: summary report. Ankorage, AK:
25 Alyeska Pipeline Service Company.
26 Hadwen, G. E.; McCarthy, J. F.; Womble, S. E.; Girman, J. R.; Brightman, H. S. (1997) Volatile organic compound
27 concentrations in 41 office buildings in the continental United States. In: Woods, J. E.; Grimsrud, D. T.;
28 Boschi, N., eds. Proceedings of healthy buildings/indoor air quality international conference; September;
29 Bethesda, MD. Washington, DC: Healthy Buildings/IAQ. [Healthy Buildings - International Conference;
30 v. 2, pp. 465-470.]
31 Hartwell, T. D.; Clayton, C. A.; Michie, R. M., Jr.; Whitmore, R. W.; Zelon, H. S.; Whitehurst, D. A.; Akland,
32 G. G. (1984) Study of carbon monoxide exposures of residents of Washington, D.C. Research Triangle Park,
33 NC: U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory; report no.
34 EPA-600/D-84-177. Available from: NTIS, Springfield, VA; PB84-230051.
35 Hayes, S. R.; Marshall, J. R. (1999) Designing optimal strategies to attain the new U.S. particulate matter standards:
36 some initial concepts. J. Air Waste Manage. Assoc. 49: PM192-PM198.
37 Henry, R. C.; Lewis, C. W.; Collins, J. F. (1994) Vehicle-related source compositions from ambient data:
38 the GRACE/SAFER method. Environ. Sci. Technol. 28: 823-832.
39 Houseman, E. A.; Ryan, L.; Levy; Richardson, D.; Spenger, J. J. (2001) Agri. Biol. Environ. Stat.: In Press.
40 Howard-Reed, C.; Rea, A. W.; Zufall, M. J.; Burke, J. M.; Williams, R. W.; Suggs, J. C.; Sheldon, L. S.; Walsh, D.;
41 Kwock, R. (2000) Use of a continuous nephelometer to measure personal exposure to particles during the
42 U.S. Environmental Protection Agency Baltimore and Fresno panel studies. J. Air Waste Manage. Assoc.
43 50:1125-1132.
44 Institute for Social Research. (1997) Child development supplement to the panel study of income dynamics.
45 Ann Arbor, MI: University of Michigan. Available: www.isr.umich.edu/src/child-development/home.html
46 [2001, January 31].
47 Isukapalli, S. S.; Georgopoulos, P. G. (2000) Modeling of outdoor/indoor relationships of gas phase pollutants and
48 particulate matter. Presented at: PM2000: particulate matter and health—the scientific basis for regulatory
49 decision-making, specialty conference & exhibition; January; Charleston, SC. Pittsburgh, PA: Air & Waste
50 Management Association.
51 Jang, M.; Kamens, R. M. (1999) Newly characterized products and composition of secondary aerosols from the
52 reaction of ec-pinene with ozone. Atmos. Environ. 33: 459-474.
53 Janssen, N. (1998) Personal exposure to airborne particles: validity of outdoor concentrations as a measure of
54 exposure in time series studies [dissertation]. Wageningen, Netherlands: Landbouwuniversiteit Wageningen.
March 2001
5-104
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Janssen, N. A. H.; Hoek, G.; Harssema, H.; Brunekreef, B. (1995) A relation between personal and ambient.
Epidemiology 6(suppl.): S45.
Janssen, N. A. H.; Hoek, G.; Harssema, H.; Brunekreef, B. (1997) Childhood exposure to PM10: relation between
personal, classroom, and outdoor concentrations. Occup. Environ. Med. 54: 888-894.
Janssen, N. A. H.; Hoek, G.; Brunekreef, B.; Harssema, H.; Mensink, I.; Zuidhof, A. (1998a) Personal sampling of
particles in adults: relation among personal, indoor, and outdoor air concentrations. Am. J. Epidemiol
147: 537-547.
Janssen, N. A. H.; Hoek, G.; Harssema, H.; Brunekreef, B. (1998b) Personal sampling of airborne particles:
method performance and data quality. J. Exposure Anal. Environ. Epidemiol. 8: 37-49.
Janssen, N. A. H.; Hoek, G.; Brunekreef, B.; Harssema, H. (1999a) Mass concentration and elemental composition
of PM10 in classrooms. Occup. Environ. Med. 56: 482-487.
Janssen, L. H. J. M.; Buringh, E.; van der Meulen, A.; van den Hout, K. D. (1999b) A method to estimate the
distribution of various fractions of PMIO in ambient air in the Netherlands. Atmos. Environ. 33: 3325-3334.
Janssen, N. A. H.; De Hartog, J. J.; Hoek, G.; Brunekreef, B.; Lanki, T.; Timonen, K. L.; Pekkanen, J. (2000)
Personal exposure to fine paniculate matter in elderly subjects: relation between personal, indoor, and
outdoor concentrations. J. Air Waste Manage. Assoc. 50: 1133-1143.
Jantunen, M. J.; Hanninen, O.; Katsouyanni, K.; Knoppel, H.; Kuenzli, N.; Lebret, E.; Maroni, M.; Saarela, K.;
Sram, R.; Zmirou, D. (1998) Air pollution exposure in European cities: the EXPOLIS study. J. Exposure
Anal. Environ. Epidemiol. 8: 495-518.
Jantunen, M. J.; Kousa, A.; Hanninen, O.; Koistinen, K.; Sram, R.; Lanki, T.; Lioy, P. J. (2000) Long term
population PM exposures correlate well with outdoor air levels - short term individual exposures do not.
Presented at: PM2000: particulate matter and health—the scientific basis for regulatory decision-making,
specialty conference & exhibition; January; Charleston, SC. Pittsburgh, PA: Air & Waste Management
Association.
Jenkins, R. A.; Palausky, M. A.; Counts, R. W.; Guerin, M. R.; Dindal, A. B.; Bayne, C. K. (1996a) Determination
of personal exposure of non-smokers to environmental tobacco smoke in the United States Lung Cancer
14(suppl. 1): S195-S213.
Jenkins, R. A.; Palausky, A.; Counts, R. W.; Bayne, C. K.; Dindal, A. B.; Guerin, M. R. (1996) Exposure to
environmental tobacco smoke in sixteen cities in the United States as determined by personal breathing zone
air sampling. J. Exposure Anal. Environ. Epidemiol. 6: 473-502.
Jensen, S. S. (1999) A geographic approach to modeling human exposure to traffic air pollution using GIS.
Roskilde, Denmark: National Environmental Research Institute.
Johnson, T. (1984) A study of personal exposure to carbon monoxide in Denver, Colorado. Research Triangle Park,
NC: U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory; report no
EPA-600/4-84-014.
Johnson, T. (1989) Human activity patterns in Cincinnati, Ohio [final report]. Palo Alto, CA: Electric Power
Research Institute; project no. RP940-06; report no. EN-6204.
Johnson, T. R. (1995) Recent advances in the estimation of population exposure to mobile source pollutants.
J. Exposure Anal. Environ. Epidemiol. 5: 551-571.
Johnson, T.; Long, T.; Ollison, W. (2000) Prediction of hourly microenvironmental concentrations of fine particles
based on measurements obtained from the Baltimore scripted activity study. J. Exposure Anal Environ
Epidemiol. 10:403-411.
Jones, N. C.; Thornton, C. A.; Mark, D.; Harrison, R. M. (2000) Indoor/outdoor relationships of particulate matter
in domestic homes with roadside, urban and rural locations. Atmos. Environ. 34: 2603-2612.
Keywood, M. D.; Ayers, G. P.; Gras, J. L.; Gillett, R. W.; Cohen, D. D. (1999) Relationships between size
segregated mass concentration data and ultrafine particle number concentrations in urban areas.
Atmos. Environ. 33: 2907-2913.
Kleeman, M. J.; Cass, G. R. (1998) Source contributions to the size and composition distribution of urban
particulate air pollution. Atmos. Environ. 32: 2803-2816.
Klepeis, N. E. (1999) An introduction to the indirect exposure assessment approach: modeling human exposure
using microenvironmental measurements and the recent national human activity pattern survey. Environ
Health Perspect. Suppl. 107(2): 365-374.
March 2001
5-105
DRAFT-DO NOT QUOTE OR CITE
-------
1 Klepeis, N. E.; Ott, W. R.; Switzer, P. (1994) A Total Human Exposure Model (THEM) for Respirable Suspended
2 Particles (RSP). Presented at: 87th annual meeting & exhibition of the Air & Waste Management
3 Association; June; Cincinnati, OH. Pittsburgh, PA: Air & Waste Management Association; paper no. 94-
4 WA75A.03. Available from: NTIS, Springfield, VA; PB94-197415.
5 Klepeis, N. E.; Ott, W. R.; Switzer, P. (1995) Modeling the time series of respirable suspended particles and carbon
6 monoxide from multiple smokers: validation in two public smoking lounges. Presented at: 88th annual
7 meeting and exhibition of the Air & Waste Management Association; June; San Antonio, TX. Pittsburgh,
8 PA: Air & Waste Management Association; paper no. 95-MPH.03.
9 Klepeis, N. E.; Ott, W. R.; Switzer, P. (1996) A multiple-smoker model for predicting indoor air quality in public
10 lounges. Environ. Sci. Technol. 30: 2813-2820.
11 Klepeis, N. E.; Switzer, P.; Ott, W. R. (2000) Estimating exposure to environmental tobacco smoke particles using
12 24-hour time-activity recall diaries. Presented at: PM2000: particulate matter and health—the scientific basis
13 for regulatory decision-making, specialty conference & exhibition; January; Charleston, SC. Pittsburgh, PA:
14 Air & Waste Management Association.
15 Koontz, M. D.; Rector, H. E. (1995) Estimation of distributions for residential air exchange rates: final report.
16 Washington, DC: U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxics;
17 contractnos. 68-D9-0166 and 68-D3-0013.
18 Koontz, M. D.; Niang, L. L. (1998) California population indoor exposure model (CPIEM): version 1.4F. Final
19 report A933-157: user's guide. Sacramento, CA: California Air Resources Board; GEOMET report no.
20 IE-2632.
21 Koutrakis, P.; Briggs, S. L. K.; Leaderer, B. P. (1992) Source apportionment of indoor aerosols in Suffolk and
22 Onondaga Counties, New York. Environ. Sci. Technol. 26: 521 -527.
23 Krewski, D.; Burnett, R. T.; Goldberg, M. S.; Hoover, K.; Siemiatycki, J.; Jerrett, M.; Abrahamowicz, M.;
24 White, W. H. (2000) Reanalysis of the Harvard Six Cities study and the American Cancer Society study of
25 particulate air pollution and mortality: a special report of the Institute's Particle Epidemiology Reanalysis
26 Project. Cambridge, MA: Health Effects Institute.
27 Lachenmyer, C.; Hidy, G. M. (2000) Urban measurements of outdoor-indoor PM2 5 concentrations and personal
28 exposure in the deep south. Part I. Pilot study of mass concentrations for nonsmoking subjects. Aerosol Sci.
29 Technol. 32:34-51.
30 Laden, F.; Neas, L. M.; Dockery, D. W.; Schwartz, J. (2000) Association of fine particulate matter from different
31 sources with daily mortality in six U.S. cities. Environ. Health Perspect. 108: 941-947.
32 ' Lai, A. C. K.; Nazaroff, W. W. (2000) Modeling indoor particle deposition from turbulent flow onto smooth
33 surfaces. J. Aerosol Sci. 31: 463-476.
34 Leaderer, B. P.; Naeher, L.; Jankun, T.; Balenger, K.; Holford, T. R.; Toth, C.; Sullivan, J.; Wolfson, J. M.;
35 Koutrakis, P. (1999) Indoor, outdoor, and regional summer and winter concentrations of PM10, PM25, SO42",
36 H+, NH/, NO3", NH3, and nitrous acid in homes with and without kerosene space heaters. Environ. Health
37 Perspect. 107:223-231.
38 Lee, S. C.; Chang, M. (1999) Indoor air quality investigations at five classrooms. Indoor Air 9: 134-138.
39 Lessler, J. T.; Kalsbeek, W. D. (1992) Nonsampling error in surveys. New York, NY: John Wiley & Sons, Inc.;
40 p. 7.
41 Liao, D.; Creason, J.; Shy, C.; Williams, R.; Watts, R.; Zweidinger, R. (1999) Daily variation of particulate air
42 pollution and poor cardiac autonomic control in the elderly. Environ. Health Perspect. 107: 521-525.
43 Ligocki, M. P.; Liu, H. I. H.; Cass, G. R.; John, W. (1990) Measurements of particle deposition rates inside
44 Southern California museums. Aerosol Sci. Technol. 13: 85-101.
45 Lillquist, D. R.; Lee, J. S.; Ramsay, J. R.; Boucher, K. M.; Walton, Z. L.; Lyon, J. L. (1998) A comparison of
46 indoor/outdoor PM10 concentrations measured at three hospitals and a centrally located monitor in Utah.
47 Appl. Occup. Environ. Hyg. 13:409-415.
48 Linn, W. S.; Gong, H., Jr.; Clark, K. W.; Anderson, K. R. (1999) Day-to-day particulate exposures and health
49 changes in Los Angeles area residents with severe lung disease. J. Air Waste Manage. Assoc.
50 49:PM108-PM115.
51 Lioy, P. J.; Waldman, J. M.; Buckley, T.; Butler, J.; Pietarinen, C. (1990) The personal, indoor and outdoor
52 concentrations of PM-10 measured in an industrial community during the winter. Atmos. Environ. Part B
53 24:57-66.
54 Lioy, P. J.; Wainman, T.; Zhang, J.; Goldsmith, S. (1999) Typical household vacuum cleaners: the collection
55 efficiency and emissions characteristics for fine particles. J. Air Waste Manage. Assoc. 49: 200-206.
March 2001
5-106
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Lippmann, M.; Ito, K.; Nadas, A.; Burnett, R. T. (2000) Association of paniculate matter components with daily
mortality and morbidity in urban populations. Cambridge, MA: Health Effects Institute; research report no.
Long, C. M.; Suh, H. H.; Koutrakis, P. (2000a) Characterization of indoor particle sources using continuous mass
and size monitors. J. Air Waste Manage. Assoc. 50: 1236-1250.
Long, C. M.; Suh, H. H.; Koutrakis, P. (2000b) Using time- and size-resolved particulate data to explore infiltration
and deposition behavior. Environ. Sci. Technol.: accepted.
Lurmann, F. W.; Korc, M. E. (1994) User's guide to the regional human exposure (REHEX) model. San Francisco,
CA: Bay Area Air Quality Management District; draft report no. STI-93150-1414-DR.
Macintosh, D. L.; Xue, J.; Ozkaynak, H.; Spengler, J. D.; Ryan, P. B. (1995) A population-based exposure model
for benzene. J. Expos. Anal. Environ. Epidemiol. 5: 375-403.
Mage, D. T. (1985) Concepts of human exposure assessment for airborne particulate matter Environ Int
11:407-412. ' '
Mage, D. T. (1998) How much PM of ambient origin are people exposed to? In: Measurement of toxic and related
air pollutants. Volume II: proceedings of a specialty conference; September; Gary, NC. Pittsburgh, PA: Air &
Waste Management Association; pp. 666-681. (A&WMA publication VIP-85).
Mage, D. T.; Ott, W. R. (1996) Accounting for nonuniform mixing and human exposure in indoor environments.
In: Tichenor, B. A., ed. Characterizing sources of indoor air pollution and related sink effects. Philadelphia,
PA: American Society for Testing and Materials; pp. 263-278.
Mage, D.; Wilson, W.; Hasselblad, V.; Grant, L. (1999) Assessment of human exposure to ambient particulate
matter. J. Air Waste Manage. Assoc. 49: 174-185.
Mar, T. F.; Norris, G. A.; Koenig, J. Q.; Larson, T. V. (2000) Associations between air pollution and mortality in
Phoenix, 1995-1997. Environ. Health Perspect. 108: 347-353.
McBride, S. J.; Ferro, A. R.; Ott, W. R; Switzer, P.; Hildemann, L. M. (1999) Investigations of the proximity effect
for pollutants in the indoor environment. J. Exposure Anal. Environ. Epidemiol. 9: 602-621.
McCurdy, T. (1995) Estimating human exposure to selected motor vehicle pollutants using the NEM series of
models: lessons to be learned. J. Exposure Anal. Environ. Epidemiol. 5: 533-550.
McCurdy, T. (1997) Modeling the dose profile in human exposure assessments: ozone as an example Rev Toxicol
1:3-23.
McCurdy, T. (2000) Conceptual basis for multi-route intake dose modeling using an energy expenditure approach
J. Exposure Anal. Environ. Epidemiol. 10: 86-97.
Mihlan, G. J.; Todd, L. A.; Truong, K. N. (2000) Assessment of occupational exposure patterns by
frequency-domain analysis of time series data. Appl. Occup. Environ. Hyg. 15: 120-130.
Miller, S. L.; Branoff, S.; Nazaroff, W. W. (1998) Exposure to toxic air contaminants in environmental tobacco
smoke: an assessment for California based on personal monitoring data. J. Exposure Anal Environ
Epidemiol. 8:287-311.
Monn, Ch.; Carabias, V.; Junker, M.; Waeber, R.; Karrer, M.; Wanner, H. U. (1997) Small-scale spatial variability
of particulate matter <10 urn (PM10) and nitrogen dioxide. Atmos. Environ. 31: 2243-2247.
Murray, D. M.; Burmaster, D. E. (1995) Residential air exchange rates in the United States: empirical and estimated
parametric distributions by season and climatic region. Risk Anal. 15: 459-465.
National Air Pollution Control Administration. (1969) Air quality criteria for particulate matter. Washington, DC:
U.S. Department of Health, Education, and Welfare, Public Health Service; NAPCA publication no AP-49
Available from: NTIS, Springfield, VA; PB-190251 IE A.
National Research Council. (1991) Human exposure assessment for airborne pollutants: advances and opportunities
Washington, DC: National Academy of Sciences.
National Research Council. (1998) Research priorities for airborne particulate matter. I. Immediate priorities and a
long-range research portfolio. Washington, DC: National Academy Press.
National Research Council. (1999) Research priorities for airborne particulate matter. II. Evaluating research
progress and updating the portfolio. Washington, DC: National Academy Press.
Navidi, W.; Thomas, D.; Langholz, B.; Stram, D. (1999) Statistical methods for epidemic-logic studies of the health
effects of air pollution. Cambridge, MA: Health Effects Institute; research report no. 86.
Nelson, P. R.; Conrad, F. W.; Kelly, S. P.; Maiolo, K. C.; Richardson, J. D.; Ogden, M. W. (1998) Composition of
environmental tobacco smoke (ETS) from international cigarettes part II: nine country follow-up Environ
Int. 24: 251-257.
March 2001
5-107
DRAFT-DO NOT QUOTE OR CITE
-------
1 Nieuwenhuijsen, M. J.; Noderer, K. S.; Schenker, M. B.; Vallyathan, V.; Olenchock, S. (1999) Personal exposure to
2 dust, endotoxin and crystalline silica in California agriculture. Ann. Occup. Hyg. 43: 35-42.
3 Oglesby, L.; Kiinzli, N.; Roosli, M.; Braun-Fahrlander, C.; Mathys, P.; Stem, W.; Jantunen, M.; Kousa, A. (2000)
4 Validity of ambient levels of fine particles as surrogate for personal exposure to outdoor air
5 pollution—results of the European EXPOLIS-EAS study (Swiss Center Basel). J. Air Waste Manage. Assoc.
6 50: 1251-1261.
7 Ott, W. R. (1984) Exposure estimates based on computer generated activity patterns. J. Toxicol. Clin. Toxicol.
8 21:97-128.
9 Ott, W. R. (1999) Mathematical models for predicting indoor air quality from smoking activity. Environ. Health
10 ' Perspect. 107(suppl. 2): 375-381.
11 Ott, W.; Thomas, J.; Mage, D.; Wallace, L. (1988) Validation of the simulation of human activity and pollutant
12 ' exposure (SHAPE) model using paired days from the Denver, CO, carbon monoxide field study. Atmos.
13 Environ. 22:2101-2113.
14 Ott, W. R.; Mage, D. T.; Thomas, J. (1992) Comparison of microenvironmental CO concentrations in two cities for
15 human exposure modeling. J. Exposure Anal. Environ. Epidemiol. 2: 249-267.
16 Ott, W. R.; Klepeis, N. E.; Switzer, P. (1995) Modeling environmental tobacco smoke in the home using transfer
17 ' functions. Presented at: 88th annual meeting and exhibition of the Air & Waste Management Association;
1 g June; San Antonio, TX; paper no. 95-WP84B.03. Research Triangle Park, NC: U.S. Environmental
19 Protection Agency, Atmospheric Research And Exposure Assessment Laboratory; report no.
20 EPA/600/A-95/079. Available from: NTIS, Springfield, VA; PB95-225488.
21 Ott, W.; Switzer, P.; Robinson, J. (1996) Particle concentrations inside a tavern before and after prohibition of
22 smoking: evaluating the performance of an indoor air quality model. J. Air Waste Manage. Assoc.
23 46:1120-1134.
24 Ott, W.; Wallace, L.; Mage, D. (2000) Predicting particulate (PM,0) personal exposure distributions using a random
25 component superposition (RCS) statistical model. J. Air Waste Manage. Assoc. 50: 1390-1406.
26 Owen, M. K.; Ensor, D. S.; Sparks, L. E. (1992) Airborne particle sizes and sources found in indoor air.
27 Atmos. Environ. Part A 26: 2149-2162.
28 Ozkaynak, H.; Spengler, J. (1996) The role of outdoor particulate matter in assessing total human exposure.
29 In: Wilson, R.; Spengler, J., eds. Particles in our air: concentrations and health effects. Cambridge, MA:
30 Harvard University Press.
31 Ozkaynak, H.; Thurston, G. D. (1987) Associations between 1980 U.S. mortality rates and alternative measures of
32 airborne particle concentration. Risk Anal. 7: 449-461.
33 Ozkaynak, H.; Ryan, P. B.; Spengler, J. D.; Laird, N. M. (1986) Bias due to misclassification of personal exposures
34 in epidemiologic studies of indoor and outdoor air pollution. In: Berglund, B.; Berglund, U.; Lindvall, T.;
35 Spengler, J.; Sundell, J., eds. Indoor air quality: papers from the third international conference on indoor air
36 quality and climate; August 1984; Stockholm, Sweden. Environ. Int. 12: 389-393.
37 Ozkaynak, H.; Spengler, J. D.; Ludwig, J. F.; Butler, D. A.; Clayton, C. A.; Pellizzari, E.; Wiener, R. W. (1990)
38 Personal exposure to particulate matter: findings from the Particle Total Exposure Assessment Methodology
39 (PTEAM) prepilot study. In: Indoor air '90: precedings of the 5th international conference on indoor air
40 quality and climate, volume 2, characteristics of indoor air; July-August; Toronto, ON, Canada. Ottawa, ON,
41 Canada: International Conference on Indoor Air Quality and Climate, Inc.; pp. 571-576.
42 Ozkaynak, H.; Spengler, J. D.; Xue, J.; Koutrakis, P.; Pellizzari, E. D.; Wallace, L. (1993) Sources and factors
43 influencing personal and indoor exposures to particles, elements and nicotine: findings from the particle
44 TEAM pilot study. In: Jantunen, M.; Kalliokoski, P.; Kukkonen, E.; Saarela, K.; Seppanen, O.; Vuorelma,
45 H., eds. Indoor air '93: proceedings of the 6th international conference on indoor air quality and climate, v. 3,
46 combustion products, risk assessment, policies; July; Helsinki, Finland. Helsinki, Finland: Indoor A.ir '93;
47 pp. 457-462.
48 Ozkaynak, H.; Xue, J.; Spengler, J.; Wallace, L.; Pellizzari, E.; Jenkins, P. (1996) Personal exposures to airborne
49 particles and metals: results from the particle TEAM study in Riverside, California. J. Exp. Anal. Environ.
50 Epidemiol. 6: 57-78.
51 Ozkaynak, H.; Xue, J.; Weker, R.; Butler, D.; Koutrakis, P.; Spengler, J. (1996b) The particle TEAM (PTEAM)
52 study: analysis of the data: final report, volume III. Research Triangle Park, NC: U.S. Environmental
53 Protection Agency, Atmospheric Research and Exposure Assessment Laboratory; report no.
54 EPA/600/R-95/098. Available from: NTIS, Springfield, VA; PB97-102495.
March 2001
5-108
DRAFT-DO NOT QUOTE OR CITE
-------
1 Patterson, E.; Eatough, D. J. (2000) Indoor/outdoor relationships for ambient PM2 5 and associated pollutants:
2 epidemiological implications in Lindon, Utah. J. Air Waste Manage. Assoc. 50: 103-110.
3 Pellizzari, E. D., Mason, R. E.; Clayton, C. A.; Thomas, K. W.; Cooper, S.; Piper. L.; Rodes, C.; Goldberg, M.;
4 Roberds, J.; Michael, L. (1998) Manganese exposure study (Toronto). Final report. Research Triangle Park,
5 NC: Research Triangle Institute, Analytical and Chemical Sciences; report no. RTI/6312/02-01 F.
6 Pellizzari, E. D.; Clayton, C. A.; Rodes, C. E.; Mason, R. E.; Piper, L. L.; Fort, B.; Pfeifer, G.; Lynam, D. (1999)
7 Participate matter and manganese exposures in Toronto, Canada. Atmos. Environ. 33: 721-734.
8 Phillips, K.; Howard, D. A.; Browne, D.; Lewsley, J. M. (1994) Assessment of personal exposures to environmental
9 tobacco smoke in British nonsmokers. Environ. Int. 20: 693-712.
10 Phillips, K.; Bentley, M. C.; Howard, D. A.;.Alvan, G. (1996) Assessment of air quality in Stockholm by personal
11 monitoring of nonsmokers for respirable suspended particles and environmental tobacco smoke. Scand. J.
12 Work Environ. Health 22(suppl. 1): 1 -24.
13 Phillips, K.; Bentley, M. C.; Howard, D. A.; Alvan, G.; Huici, A. (1997a) Assessment of air quality in Barcelona by
14 personal monitoring of nonsmokers for respirable suspended particles and environmental tobacco smoke.
15 Environ. Int. 23: 173-196.
16 Phillips, K.; Howard, D. A.; Bentley, M. C.; Alvan, G. (1997b) Assessment of air quality in Turin by personal
17 monitoring of nonsmokers for respirable suspended particles and environmental tobacco smoke. Environ. Int.
18 23:851-871.
19 Phillips, K.; Bentley, M. C.; Howard, D. A.; Alvan, G. (1998a) Assessment of air quality in Paris by personal
20 monitoring of nonsmokers for respirable suspended particles and environmental tobacco smoke. Environ. Int.
21 24:405-425.
22 Phillips, K.; Howard, D. A.; Bentley, M. C.; Alvan, G. (1998b) Assessment of environmental tobacco smoke and
23 respirable suspended particle exposures for nonsmokers in Lisbon by personal monitoring. Environ. Int.
24 24:301-324.
25 Phillips, K.; Howard, D. A.; Bentley, M. C.; Alvan, G. (1999) Assessment of environmental tobacco smoke and
26 respirable suspended particle exposures for nonsmokers in Basel by personal monitoring. Atmos. Environ.
27 33: 1889-1904.
28 Pope, C. A., Ill; Thun, M. J.; Namboodiri, M. M.; Dockery, D. W.; Evans, J. S.; Speizer, F. E.; Heath, C. W., Jr.
29 (1995) Particulate air pollution as a predictor of mortality in a prospective study of U.S. adults. Am. J.
30 Respir. Crit. Care Med. 151: 669-674.
31 Pope, C. A., Ill; Hill, R. W.; Villegas, G. M. (1999) Particulate air pollution and daily mortality on Utah's Wasatch
32 Front. Environ. Health Perspect.: 107: 567-573.
33 Praml, G.; Schierl, R. (2000) Dust exposure in Munich public transportation: a comprehensive 4-year survey in
34 buses and trams. Int. Arch. Occup. Environ. Health 73: 209-214.
35 Quintana, P. J. E.; Samimi, B. S.; Kleinman, M. T.; Liu, L.-J.; Soto, K.; Warner, G. Y.; Bufalino, C.; Valencia, J.;
36 Francis, D.; Hovell, M. H.; Delfino, R. J. (2000) Evaluation of a real-time passive personal particle monitor
37 in fixed site residential indoor and ambient measurements. J. Exposure Anal. Environ. Epidemiol.
38 10:437-445.
39 Ramachandran, G.; Vincent, J. H. (1999) A Bayesian approach to retrospective exposure assessment. Appl. Occup.
40 Environ. Hyg. 14: 547-557.
41 Rea, A. W.; Zufall, M. J.; Williams, R. W.; Howard-Reed, C.; Sheldon, L. S. (2001) The influence of human
42 activity patterns on personal PM exposure: a comparative analysis of filter-based and continuous particle
43 measurements. J. Air Waste Manage. Assoc.: In Press.
44 Ren, P.; Jankun, T. M.; Leaderer, B. P. (1999) Comparisons of seasonal fungal prevalence in indoor and outdoor air
45 and in house dusts of dwellings in one northeast American county. J. Exposure Anal. Environ. Epidemiol.
46 9:560-568.
47 Robinson, J. P.; Wiley, J. A.; Piazza, T.; Garrett, K.; Cirksena, K. (1989) Activity patterns of California residents
48 and their implications for potential exposure to pollution. Sacramento, CA: California Air Resources Board
49 report no. CARB-A6-177-33.
50 Robinson, J. P.; Switzer, P.; Ott, W. (1994) Microenvironmental factors related to Californians' potential exposure
51 to environmental tobacco smoke (ETS). Research Triangle Park, NC: U.S. Environmental Protection Agency,
52 Atmospheric Research and Exposure Assessment Laboratory; report no. EPA/600/R-94/116. Available from:
53 NTIS, Springfield, VA; PB94-214814.
54
March 2001
5-109
DRAFT-DO NOT QUOTE OR CITE
-------
1 Rodes, C.; Sheldon, L.; Whitaker, D.; Clayton, A.; Fitzgerald, K.; Flanagan, J.; DiGenova, F.; Hering, S.;
2 Frazier, C. (1998) Measuring concentrations of selected air pollutants inside California vehicles [final
3 report]. Sacramento, CA: California Environmental Protection Agency, Air Resources Board; contract no.
4 95-339.
5 Rojas-Bracho, L.; Suh, H. H.; Koutrakis, P. (2000) Relationships among personal, indoor, and outdoor fine and
6 coarse particle concentrations for individuals with COPD. J. Exposure Anal. Environ. Epidemiol.
7 10:294-306.
8 Roorda-Knape, M. C.; Janssen, N. A. H.; De Hartog, J. J.; Van Vliet, P. H. N.; Harssema, H.; Brunekreef, B. (1998)
9 Air pollution from traffic in city districts near major motorways. Atmos. Environ. 32: 1921-1930.
10 Ryan, P. B. (1991) An overview of human exposure modeling. J. Exposure Anal. Environ. Epidemiol. 1: 453-474.
11 Santos-Burgoa, C.; Rojas-Bracho, L.; Rosas-Perez, I.; Ramirez-Sanchez, A.; Sanchez-Rico, G.;
12 Mejia-Hernandez, S. (1998) Modelaje de exposition a particulas en poblacion general y riesgo de
13 enfermedad respiratoria [Particle exposure modeling in the general population and risk of respiratory
14 disease]. Gac. Med. Mex. 134:407-417.
15 Sarnat, J. A.; Koutrakis, P.; Suh, H. H. (2000) Assessing the relationship between personal particulate and gaseous
16 exposures of senior citizens living in Baltimore, MD. J. Air Waste Manage. Assoc. 50: 1184-1198.
17 Schwartz, J.; Levin, R. (1999) Drinking water turbidity and health. Epidemiology 10: 86-90.
18 Sexton, K.; Ryan, P. B. (1988) Assessment of human exposure to air pollution: methods, measurements, and
19 models. In: Watson, A. Y.; Bates, R. R.; Kennedy, D., eds. Air pollution, the automobile, and public health.
20 Washington, DC: National Academy Press; pp. 207-238.
21 Sexton, K.; Spengler, J. D.; Treitman, R. D. (1984) Personal exposure to respirable particles: a case study in
22 Waterbury, Vermont. Atmos. Environ. 18: 1385-1398.
23 Sheldon, L.; Clayton, A.; Jones, B.; Keever, J.; Perritt, R.; Smith, D.; Whitaker, D.; Whitmore, R. (1992a) Indoor
24 pollutant concentrations and exposures. Sacramento, CA: California Air Resources Board; report no.
25 ARB/R-92/486.
26 Sheldon, L.; Clayton, A.; Keever, J. Perritt, R.; Whitaker, D. (1992b) PTEAM: monitoring of phthalates and PAHs
27 in indoor and outdoor air samples in Riverside, California, volume II [final report]. Sacramento, CA:
28 California Environmental Protection Agency, State Air Resources Board, Research Division; contract no.
29 A933-144. Available from: NTIS, Springfield, VA; PB93-205649/XAB.
30 Sheldon, L.; Clayton, A.; Perritt, R.; Whitaker, D. A.; Keever, J. (1993a) Indoor concentrations of polycyclic
31 aromatic hydrocarbons in California residences and their relationship to combustion source use. In: Jantunen,
32 M.; Kalliokoski, P.; Kukkonen, E.; Saarela, K.; Seppanen, O.; Vuorelma, H., eds. Indoor air '93: proceedings
33 of the 6th international conference on indoor air quality and climate. Volume 3: combustion products, risk
34 assessment, policies; July; Helsinki, Finland. Helsinki, Finland: Indoor Air '93; pp. 29-34.
35 Sheldon, L.; Whitaker, D.; Keever, J.; Clayton, A.; Perritt, R. (1993b) Phthalates and PAHs in indoor and outdoor
36 air in a southern California community. In: Jantunen, M.; Kalliokoski, P.; Kukkonen, E.; Saarela, K.;
37 Seppanen, O.; Vuorelma, H., eds. Indoor air '93: proceedings of the 6th international conference on indoor
38 air quality and climate. Volume 3: combustion products, risk assessment, policies; July; Helsinki, Finland.
39 Helsinki, Finland: Indoor Air '93; pp. 109-114.
40 Sheldon, L.; Clayton, A.; Keever, J.; Perritt, R.; Whitaker, D. (1993c) Indoor concentrations of polycyclic aromatic
41 hydrocarbons in California residences. Sacramento, CA: Air Resources Board; ARB contract no. A033-132.
42 Shy, C. M.; Kleinbaum, D. G.; Morgenstern, H. (1978) The effect of misclassification of exposure status in
43 epidemiological studies of air pollution health effects. Bull. N. Y. Acad. Med. 54: 1155-1165.
44 Siegmann, K.; Scherrer, L.; Siegmann, H. C. (1999) Physical and chemical properties of airborne nanoscale
45 particles and how to measure the impact on human health. J. Mol. Struct. (Theochem) 458: 191-201.
46 Sinclair, J. D.; Psota-Kelty, L. A.; Weschler, C. J. (1988) Indoor/outdoor ratios and indoor surface accumulations of
47 ionic substances at Newark, New Jersey. Atmos. Environ. 22:461 -469.
48 Sinclair, J. D.; Psota-Kelty, L. A.; Weschler, C. J.; Shields, H. C. (1990) Measurement and modeling of airborne
49 concentrations and indoor surface accumulation rates of ionic substances at Neenah, Wisconsin. Atmos.
50 Environ. Part A 24:627-638.
51 Spengler, J. D.; Treitman, R. D.; Tosteson, T. D.; Mage, D. T.; Soczek, M. L. (1985) Personal exposures to
52 respirable particulates and implications for air pollution epidemiology. Environ. Sci. Technol. 19: 700-707.
53
54
March 2001
5-110
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Spengler, J. D.; Ozkaynak, H.; Ludwig, J.; Allen, G.; Pellizzari, E. D.; Wiener, R. (1989) Personal exposures to
participate matter: instruments and methodologies P-TEAM. In: Measurement of toxic and related air
pollutants: proceedings of the 1989 U.S. EPA/A&WMA international symposium; May; Raleigh, NC.
Pittsburgh, PA: Air & Waste Management Association; pp. 449-463. (A&WMA publication VIP-13).
Spier, C. E.; Little, D. E.; Trim, S. C.; Johnson, T. R.; Linn, W. S.; Hackney, J. D. (1992) Activity patterns in
elementary and high school students exposed to oxidant pollution. J. Exposure Anal. Environ. Epidemiol
2:277-293. .
Suggs, J. C.; Burton, R. M. (1983) Spatial characteristics of inhalable particles in the Philadelphia metropolitan
area. J. Air Pollut. Control Assoc. 33: 688-691.
Suh, H. H.; Spengler, J. D.; Koutrakis, P. (1992) Personal exposures to acid aerosols and ammonia Environ Sci
Technol. 26: 2507-2517.
Suh, H. H.; Koutrakis, P.; Spengler, J. D. (1993a) Indoor and outdoor acid aerosol and gas concentrations.
In: Jantunen, M.; Kalliokoski, P.; Kukkonen, E.; Saarela, K.; Seppanen, O.; Vuorelma, H., eds. Indoor air
'93: proceedings of the 6th international conference on indoor air quality and climate. Volume 3: combustion
products, risk assessment, policies; July; Helsinki, Finland. Helsinki, Finland: Indoor Air '93; pp. 257-262.
Suh, H. H.; Koutrakis, P.; Spengler, J. D. (1993b) Validation of personal exposure models for sulfate and aerosol
strong acidity. J. Air Waste Manage. Assoc. 43: 845-850.
Suh, H. H.; Koutrakis, P.; Spengler, J. D. (1994) The relationship between airborne acidity and ammonia in indoor
environments. J. Exposure Anal. Environ. Epidemiol. 4: 1-23.
Tamura, K.; Ando, M.; Sagai, M.; Matsumoto, Y. (1996a) Estimation of levels of personal exposure to suspended
particulate matter and nitrogen dioxide in Tokyo. Environ. Sci. (Tokyo) 4: 37-51.
Tamura, K.; Ando, M.; Miyazaki, T.; Funasaka, K.; Kuroda, K. (1996b) Indoor concentrations and personal
exposure to particulate matter at roadside houses. In: Indoor air '96: the 7th international conference on
indoor air quality, volume I; July; Nagoya, Japan. Tokyo, Japan: Institute of Public Health; pp. 599-602
Teschke, K.; Marion, S. A.; Vaughan, T. L.; Morgan, M. S.; Camp, J. (1999) Exposures to wood dust in U.S.
industries and occupations, 1979 to 1997. Am. J. Ind. Med. 35: 581-589.
Thatcher, T. L.; Layton, D. W. (1995) Deposition, resuspension, and penetration of particles within a residence
Atmos. Environ. 29: 1487-1497.
Thomas, K. W.; Pellizzari, E. D.; Clayton, C. A.; Whitaker, D. A.; Shores, R. C.; Spengler, J.; Ozkaynak, H.;
Froehlich, S. E.; Wallace, L. A. (1993) Particle total exposure assessment methodology (PTEAM) 1990
study: method performance and data quality for personal, indoor, and outdoor monitoring. J. Exposure Anal.
Environ. Epidemiol. 3:203-226.
Tsang, A. M.; Klepeis, N. E. (1996) Descriptive statistics tables from a detailed analysis of the national human
activity pattern survey (NHAPS) data. Las Vegas, NV: U.S. Environmental Protection Agency, Office of
Research and Development; report no. EPA/600/R-96/148.
U.S. Environmental Protection Agency. (1982) Air quality criteria for particulate matter and sulfur oxides. Research
Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
Office; EPA report no. EPA-600/8-82-029aF-cF. 3v. Available from: NTIS, Springfield, VA; PB84-156777.
U.S. Environmental Protection Agency. (1996) Air quality criteria for particulate matter. Research Triangle Park,
NC: National Center for Environmental Assessment-RTP Office; report nos. EPA/600/P-95/001aF-cF. 3v
U.S. Environmental Protection Agency. (1998) Particulate matter research needs for human health risk assessment
to support future reviews of the national ambient air quality standards for particulate matter. Research
Triangle Park, NC: National Center for Environmental Assessment; report no. EPA/600/R-97/132F.
U.S. Environmental Protection Agency. (2000) Aerometric Information Retrieval System (AIRS). Research
Triangle Park, NC: Office of Air Quality Planning and Standards. Available: www.epa.gov/ttn/airs/
[2000, October 6].
Vakeva, M.; Hameri, K.; Kulmala, M.; Lahdes, R.; Ruuskanen, J.; Laitinen, T. (1999) Street level versus rooftop
concentrations of submicron aerosol particles and gaseous pollutants in an urban street canyon
Atmos. Environ. 33: 1385-1397.
Van Vliet, P.; Knape, M.; De Hartog, J.; Janssen, N.; Harssema, H.; Brunekreef, B. (1997) Motor vehicle exhaust
and chronic respiratory symptoms in children living near freeways. Environ. Res. 74: 122-132.
Vette, A. F.; Rea, A. W.; Lawless, P. A.; Rodes, C. E.; Evans, G.; Highsmith, V. R.; Sheldon, L. (2001)
Characterization of indoor-outdoor aerosol concentration relationships during the Fresno PM exposure
studies. Aerosol Sci. Technol.: accepted.
March 2001
5-111
DRAFT-DO NOT QUOTE OR CITE
-------
1 Wainman, T.; Zhang, J.; Weschler, C. J.; Lioy, P. J. (2000) Ozone and limonene in indoor air: a source of
2 submicron particle exposure. Environ. Health Perspect. 108: 1139-1145.
3 Waldman, J. M.; Liang, C. S. K. (1993) Penetration and neutralization of acid aerosols at institutional settings.
4 ImJantunen, M.; Kalliokoski, P.; Kukkonen, E.; Saarela, K.; Seppanen, O.; Vuorelma, H., eds. Indoor air
5 '93: proceedings of the 6th international conference on indoor air quality and climate. Volume 3: combustion
6 products, risk assessment, policies; July; Helsinki, Finland. Helsinki, Finland: Indoor Air '93; pp. 263-268.
7 Waldman, J. M.; Liang, C. S. K.; Menon, P. (1990) Indoor exposures to acidic aerosols at child and elderly care
8 facilities. In: Indoor air '90: precedings of the 5th international conference on indoor air quality and climate,
9 volume 2, characteristics of indoor air; July-August; Toronto, ON, Canada. Ottawa, ON, Canada:
10 International Conference on Indoor Air Quality and Climate, Inc.; pp. 607-612.
11 Wallace, L. (1996) Indoor particles: a review. J. Air Waste Manage. Assoc. 46: 98-126.
12 Wallace, L. (2000a) Real-time monitoring of particles, PAH, and CO in an occupied townhouse. Appl. Occup.
13 Environ. Hyg. 15: 39-47.
14 Wallace, L. (2000b) Correlations of personal exposure to particles with outdoor air measurements: a review of
15 recent studies. Aerosol Sci. Technol. 32: 15-25.
16 Wallace, L.; Quackenboss, J.; Rodes, C. (1997) Continuous measurements of particles, PAH, and CO in an
17 occupied townhouse in Reston, VA. In: Measurement of toxic and related air pollutants. Volume II:
1 g proceedings of a specialty conference; April-May; Research Triangle Park, NC. Pittsburgh, PA: Air & Waste
19 Management Association, pp. 860-871. (A&WMA publication VIP-74).
20 Watson, J. G.; Cooper, J. A.; Huntzicker, J. J. (1984) The effective variance weighting for least squares calculations
21 applied to the mass balance receptor model. Atmos. Environ. 18: 1347-1355.
22 Watson, J. G.; Robinson, N. F.; Chow, J. C.; Henry, R. C.; Kim, B. M.; Pace, T. G.; Meyer, E. L.; Nguyen, Q.
23 (1990) The USEPA/DRI chemical mass balance receptor model, CMB 7.0. Environ. Software 5: 38-49.
24 Watson, J. G.; Chow, J. C.; Pace, T. G. (1991) Chemical mass balance. In: Hopke, P. K., ed. Receptor modeling for
25 a'ir quality management. New York, NY: Elsevier; pp. 83-116. (Data handling in science and technology—
26 volume 7).
27 Weschler, C. J.; Shields, H. C. (1999) Indoor ozone/terpene reactions as a source of indoor particles. Atmos.
28 Environ. 33:2301-2312.
29 White, W. H. (1998) Statistical considerations in the interpretation of size-resolved particulate mass data. J. Air
30 'waste Manage. Assoc. 48:454-458.
31 Wiener, R. W. (1988) Measurement and evaluation of personal exposure to aerosols. In: Measurement of toxic and
32 related air pollutants: proceedings of the 1988 EPA/APCA international symposium; May; Research Triangle
33 Park, NC. Pittsburgh, PA: Air Pollution Control Association; pp. 84-88. (APCA publication VIP-10).
34 Wiener, R. W. (1989) Particle total exposure assessment methodology-an overview of planning and
35 accomplishments. In: Measurement of toxic and related air pollutants: proceedings of the 1989 U.S.
36 EPA/A&WMA international symposium; May; Raleigh, NC. Pittsburgh, PA: Air & Waste Management
37 Association; pp. 442-448. (A&WMA publication VIP-13).
38 Wiener, R. W.; Wallace, L.; Pahl, D.; Pellizzari, E.; Whittaker, D.; Spengler, J.; Ozkaynak, H. (1990) Review of the
39 particle TEAM 9 home field study. In: Measurement of toxic and related air pollutants: proceedings of the
40 1990 EPA/A&WMA international symposium, v. 2; May; Raleigh, NC. Pittsburgh, PA: Air & Waste
41 Management Association; pp. 452-460. (A&WMA publication VIP-17).
42 Wiley, J. A.; Robinson, J. P.; Cheng, Y.-T.; Piazza, T.; Stork, L.; Pladsen, K. (1991a) Study of children's activity
43 patterns: final report. Sacramento, CA: California Air Resources Board; report no. ARB-R-93/489.
44 Williams, R.; Suggs, J.; Zweidinger, R.; Evans, G.; Creason, J.; Kwok, R.; Rodes, C.; Lawless, P.; Sheldon, L.
45 (2000a) The 1998 Baltimore Particulate Matter Epidemiology-Exposure Study: part 1. Comparison of
46 ambient, residential outdoor, indoor and apartment particulate matter monitoring. J. Exposure Anal. Environ.
47 Epidemiol. 10:518-532.
48 Williams, R.; Suggs, J.; Creason, J.; Rodes, C.; Lawless, P.; Kwok, R.; Zweidinger, R.; Sheldon, L. (2000b) The
49 1998 Baltimore particulate matter epidemiology-exposure study: part 2. Personal exposure assessment
50 associated with an elderly study population. J. Exposure Anal. Environ. Epidemiol.: 10: 533-543.
51 Williams, R.; Creason, J.; Zweidinger, R.; Watts, R.; Sheldon, L.; Shy, C. (2000c) Indoor, outdoor, and personal
52 exposure monitoring of particulate air pollution: the Baltimore elderly epidemiology-exposure pilot study.
53 Atmos. Environ. 34:4193-4204.
54 Wilson, W. E.; Suh, H. H. (1997) Fine particles and coarse particles: concentration relationships relevant to
55 epidemiologic studies. J. Air Waste Manage. Assoc. 47: 1238-1249.
March 2001
5-112
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Wilson, W. E.; Mage, D. T.; Grant, L. D. (2000) Estimating separately personal exposure to ambient and
nonambient particulate matter for epidemiology and risk assessment: why and how. J. Air Waste Manage.
Assoc. 50: 1167-1183.
Yakovleva, E.; Hopke, P. K.; Wallace, L. (1999) Receptor modeling assessment of particle total exposure
assessment methodology data. Environ. Sci. Technol. 33: 3645-3652.
Zeger, S. L.; Thomas, D.; Dominici, F.; Samet, J. M.; Schwartz, J.; Dockery, D.; Cohen, A. (2000) Exposure
measurement error in time-series studies of air pollution: concepts and consequences. Environ. Health
Perspect. 108: 419-426.
Zmirou, D.; Masclet, P.; Boudet, C.; Dor, F.; Dechenaux, J. (2000) Personal exposure to atmospheric polycyclic
aromatic hydrocarbons in a general adult population and lung cancer risk assessment. J. Occup. Environ.
Med. 42: 121-126.
March 2001
5-113
DRAFT-DO NOT QUOTE OR CITE
-------
-------
-------
United States
Environmental Protection Agency/ORD
National Center for
Environmental Assessment
Washington, D.C. 20460
Please make all necessary changes on the below label,
detach or copy, and return to the address in the upper
left-hand corner.
If you do not wish to receive these reports CHECK HEREtH ;
detach, or copy this cover, and return to the address in the
upper left-hand corner.
Official Business
Penally for Private Use
S300
EP/V600/P-99/002aB
-------