-------
  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
•§
!| 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 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   •§
                       
       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
-------