Finkelstein: Deposition Velocities of S02 and 03
1
Deposition Velocities of S02 and 03 over Agricultural and Forest Ecosystems
Peter L. Finkelstein1
NOAA, Atmospheric Sciences Modeling Division
Research Triangle Park, N.C., 27711
finkelstein.peter@epa.gov
1 On assignment to the National Exposure Research Lab., US EPA
Abstract The results of field studies that measured the flux and deposition velocity of
S02 and 03 are reported. Three of the studies were over agricultural crops (pasture, corn,
and soybean), and two were over forest (a deciduous forest and a mixed coniferous -
deciduous forest). In all cases the deposition velocity for S02 was higher than that for
03. Diurnal cycles of S02 deposition velocity were similar in shape, but not magnitude
for all surfaces; however those for 03 showed some difference between forest sites where
die peak was in die morning, and die agricultural sites where the peak occurred at mid-
day. Seasonal cycles of S02 were affected by deposition to surfaces when leaves weren't
active, yet surface conductance is significant, but not for 03 where stomatal uptake is the
primary pathway for deposition.
Keywords: deposition velocity, dry deposition, flux, ozone, sulfur dioxide
1. Introduction
Because of the high cost and complexity of direct flux measurements, national
dry deposition networks measure concentrations of air pollutants and infer the flux of
pollutants to the surface using a modeled deposition velocity. In the U. S., the 70+ site
Castnet operational dry deposition network operated by the U. S. EPA, and the 10+ site
Airmon research network operated by NOAA both use the multi-layer deposition velocity
model (MLM) (Meyers et al., 1998). In light of its importance, an evaluation of MLM
was desirable.
There have been numerous field studies that measured deposition velocity,
including work by Matt et al. (1987), Meyers and Baldocchi (1988), Erisman and
Duyzer (1991), Padro et al. (1991,1994), Fuentes et al. (1992), and Munger et al.
(1996). While these studies collected valuable information, they did not necessarily
collect the data needed to exercise the MLM, nor do they have a consistent set of
measurements over varying ecosystems. To overcome these problems, a study was
conducted that measured the deposition velocities of key pollutants and the associated
meteorological and biological variables over a variety of agricultural and natural
ecosystems, distributed about the eastern half of the United States. Studies were done

-------
2
over pastures in North Carolina and Alabama, soybeans in Kentucky and North Carolina,
com in Illinois, a salt-water estuary in New Jersey, and forests in North Carolina,
Pennsylvania, and New York. Study duration ranged from six weeks to full growing
seasons. Complete descriptions of the measurement systems and key sites may be found
in Meyers et al. (1998), and Finkelstein et al. (2000). This paper is a synthesis of the
observations from that program.
2. Data
Instruments at the various sites were mounted approximately 5 m above the top
of the crop, or, in the case of the forest sites, on a 36 m guyed, walk-up, scaffold tower.
Chemical analyzers were housed in an air conditioned box near the base of the tower, or
in shelters built on the tower at the forest sites. Wind velocity and turbulence were
measured with an ATI sonic anemometer. 03 and S02 were sampled from a draft tube
with the inlet immediately adjacent to the sonic. Fast response measurements of 03 were
made with a specially constructed analyzer that uses the chemiluminescent reaction of 03
with eosin-y dye. Fast response S02 measurements were taken with a modified Meloy
SA285-E total sulfur analyzer. Soil temperature and soil heat flux were measured near
the base of the tower. Leaf area index (LAI) and stomatal resistance measurements were
made at several locations throughout the canopy. Details on the instrumentation, data
handling, quality assurance, and data acceptance criteria are in Meyers et al. (1998) and
Finkelstein et al. (2000).
Data from five of the sites are discussed in this paper. The sites are a pasture in
Sand Mountain, Alabama (SM), a corn field in Bondville, Illinois (BV), a soybean field
near Nashville, Tennessee (NV), a deciduous forest near Kane, Pennsylvania (K), and a
mixed deciduous-coniferous forest in the Sand-Flats Forest in the Adirondack Mountains
of New York State (SF). The number of valid half-hour observations of deposition
velocity varied considerably among sites and between S02 and 03. The smaller number
of S02 observations is a result of the lower ambient concentration and high minimum
detectable limit (2 ppb) of our fast response S02 instrument. The number of half-hour
observations for S02 and 03 respectively are: BV-191, 704; SM-304, 1347; NV-374,
1852; K-1625,2861; SF-191,2494.
The Sand Mountain site was located about 1 km south of Crossville, AL, (34.29
N, 85.97 W). The measurement site was on a slight knoll with a downward slope of
about 1% in all directions. Vegetation was pasture: 52% fescue, 20% blue grass, and
20% white clover. Small trees were located about 500 m to the SW and to the NW. The
data collection period began April 15 and ended on June 13,1995. Surface moisture
was adequate for good growth throughout the period. Leaf area index (LAI) increased
from about 1.0 on April 15 to about 2.3 by June 13.
The Bondville study took place at the University of Illinois Soil and Crop
Experimental Station (40.05N, 88.37W). The fetch was uniform for several km. Data
collection ran from August 18 until October 1, 1994. At the start of the experiment the
corn was 1.8 to 2.4 m high. By September 2, the corn was starting to mature and by
October 3 essentially all the leaves were brown. LAI was 3.0 on August 18, increasing

-------
3
to 3.3 by the end of August and then slowly decreasing to 2.5 by October 1. Soil
moisture was adequate for the crop throughout the experiment.
The Nashville site was in a soybean field (36.65 N, 87.03 W), 60 km NNW of
Nashville and 2 km west ofKeysburg, KY. Gently rolling uniform fetch over the
soybean crop extended to at least 1500 m through the SE and southern quadrants. A com
field was adjacent to the soybeans, 140 m to the west at the closest point. The soybeans
were planted on June 13. Dry deposition sampling was initiated June 22 and ended
October 11, 1995. The beans went through a rapid growth period from July 10 through
August 5 with the LAI increasing from 1 to about 6. Precipitation was light after late
July and by mid August the beans were drought stressed. The LAI gradually decreased
to about 3 by the end of September.
The Kane Experimental Forest is located in northwestern Pennsylvania,
(41.60°N, 78.77°W). The area has gently rolling topography and elevation changes of
15 to 30 m within 1500 m of the site. The tree canopy is nearly uniform in height, and
tends to damp out variations in ground elevation. In the vicinity of the experiment, there
is a mix of tree species: 38% Black Cherry (Prumis serotina), 34% Red Maple (Acer
rubrum), 23% Sugar Maple (Acer saccharinum), and 5% others. The canopy is 22 to 23
meters high. Hie soil was moist throughout the year. Valid data were collected from
April 29 to October 23, 1997. Leaf bud occurred approximately the second week of
May. Leaves were senescent by the middle of October.
The Sand Flats State Forest is about 7 miles NE of Boonville, NY, (43.57° N,
75.24° W). Observations were taken from May 12 to October 20, 1998. Composition of
the forest included 20% White Pine (Pinus strbbus), 20% Black Cherry (Prunus
serotina), 17% Sugar Maple (Acer saccharinum), 15% Hemlock (Tsuga canadensis),
10% White Spruce (Picea glauca), 8% open space, and a scattering of others. The soil
throughout the area is sandy and well drained. The topography of the area is quite flat
within 1 km of the tower. Because of the mix of species, the vertical structure of the
forest was quite different from that at Kane, with a much higher density of branches and
needles at lower levels on the conifers mixing with the higher leaves of the deciduous
trees.
3. Data Summary and Synthesis
Mean daytime and nighttime deposition velocities, ± one standard error, for S02
and Oj at the five sites are shown in Figure 1(A and B). One of the most noteworthy
things to observe is that the S02 deposition velocity is consistently higher, by 0.2 to 0.4
cm/s than that for 03 Nighttime deposition velocity values are lower than they are during
the day, but they are not insignificant, especially for S02.
A plot of daytime, growing season, mean deposition velocity vs. LAI (Figure 2)
for all of the sites we studied shows how clearly deposition velocity depends on
vegetation mass. The relationship is quite linear for 03 and, with one curious exception,
for S02 as well. The LAI for the forest sites are underestimated, due to the leaf clumping
factor (Chason et at, 1991). An appropriate increase in forest LAI would make the
relationship even more linear. From the data displayed, except for the anomalous S02

-------
4
value at BV, the best linear fits for daytime average deposition velocity during the
growing season are: SOfiV = .2*LAI + .42 (R=995); and OfiV = ,15*LAI + .17
(R=.96).
Figure 3 (A and B) shows the average diumal curves for S02 and 03 at the five
sites. Because the number of observations is small, there is a fair degree of uncertainty in
the S02 curves. Nonetheless, one can see the expected minimum shortly after sunset,
with a slight increase in the predawn hours, probably caused by increased humidity and
surface moisture, followed by a rapid increase after sunrise, He peak occurs between
morning and mid-day. The values decrease quite rapidly in the late afternoon and
evening. The 03 curves are more regular because of the larger number of samples. The
values are quite low at night, except for the Sand Flats forest. Deposition velocities at all
sites rise rapidly in the early morning, and decrease rapidly in the late afternoon. The
forest sites have a peak in deposition velocity in the morning, while the crop sites peak at
mid-day. We speculate that this reduction at the forest sites after the early morning peak
is caused by sunlight heating the leaf surface, and a significant vapor pressure deficit,
which causes the stomata to partially close. The same effect does not seem to be present
in crops. The reason for the higher nighttime deposition velocities observed at Sand
Flats, which may also be seen in Figure IB, is not obvious, but we speculate that it may
be due to nocturnal stomata! conductance by some one or more species present at that
site. Musselman and Minnick (2000) have listed over one hundred species of plants for
which nocturnal stomatal conductance has been observed. And, while none of the
dominant species at the site were in their list, they note that the list is far from inclusive,
as tests have not been made on most types of plants.
Seasonal cycles of daytime deposition velocity, composed of weekly average
values, are shown in Figure 4 (A and B). Again, because of the smaller number of
observations, the S02 figures are less certain, and have more missing values. One point
of note in comparing these cycles is that the 03 cycles show a more pronounced
minimum in the early spring and late fall than do the S02 cycles. This difference is
especially evident in the Kane Forest data. We suspect that this difference is caused by
the deposition of S02 to the surfaces of the branches and other structures in the forest,
causing a significant deposition velocity of S02 even in the absence of leaves. Because
03 deposition is more strictly a stomatal effect, the deposition of 03 in the spring and fall
is greatly diminished. The comparison of the deposition velocity for S02 between the
Kane, Sand Mountain, and Nashville sites is instructive. At Kane, with significant
surfaces, the deposition velocity is rather high even without leaves. At Sand Mountain,
the grass covered pasture had some stomatal activity in the early spring, and some
surfaces for deposition. At Nashville there were no plants above ground at the start of
the experiment (week 24) and the deposition velocity was the smallest, only half the
value of the deposition velocity at Kane. On the other hand, comparing these sites for 03
shows quite similar values at the start of the respective growing seasons, again indicating
that surfaces are much more important for S02 deposition.
4. Conclusion

-------
5
In the field studies reported here, vegetation mass, as measured by LAI, is the
dominant variable controlling deposition velocity. We noted significant differences
between S02 and 03 deposition velocity. S02 deposition velocity is larger at all sites.
S02 deposition velocity has two important pathways, to the interior of leaves through the
stomata, and to the surfaces of the plants and ground. 03 deposition velocity is
principally controlled by plant stomata. Differences are noted in the diurnal cycle of
stomatal opening between forests and agricultural crops, with the peak time of day for
deposition velocity occurring in the morning over forests, and at mid-day over crops.
Measurable amounts of deposition occur at night for both pollutants. Nocturnal
deposition may be more significant than originally thought for plant damage (Musselman
and Massman, 1999) and deserves more careful study. Deposition to surfaces, the effect
of wetness, and the chemistry of that wetness, and deposition during the winter are also
areas that have received too little attention and are important in understanding total
loadings of pollutants to ecosystems and the relationship between dry deposition and
plant damage.
Acknowledgment
I wish to thank my colleagues, John F. Clarke, Thomas G. Ellestad, Eric O.
Hebert, and Jonathan E. Pleim. Funding was provided by the U.S. Environmental
Protection Agency. This paper has been reviewed and approved for publication by both
the Environmental Protection Agency and the National Oceanic and Atmospheric
Administration. Mention of specific products does not constitute endorsement by either
agency.
References
Chason, J. W., D. D. Baldocchi, and M. A. Huston: 1991, 'A comparison of direct and
indirect methods for estimating forest canopy leaf area', Ag. and Forest Meteor. 57,
107-128.
Erisman, J. W. and J. Duyzer: 1991, 'A micrometeorological investigation of surface
exchange parameters over heathland', Boundary-Layer Meteorology 57, 115-128.
Finkelstein, P. L., T. G. Ellestad, J. F. Clarke, T. P. Meyers, D. B. Schwede, E. O.
Hebert, and J. A. Neal: 2000, 'Ozone and sulfur dioxide dry deposition to forests:
Observations and model evaluation', J. Geophys. Res. 105 (D12) 15,365-15378. In Press.
Fuentes, J.D., T.J.Gillespie, G. den Hartog, and H.H. Neumann: 1992, 'Ozone deposition
onto a deciduous forest during dry and wet conditions', Ag. and Forest Meteor. 62, 1-
18.
Matt, D. R., R. T. McMillen, J. D. Womack, and B. B. Hicks: 1987, 'A comparison of
estimated and measured S02 deposition velocities', Water, Air and Soil Pollution, 36, 31-

-------
6
347.
Meyers, T. P., and D. D. Baldocchi: 1988, 'A comparison of models for deriving dry
deposition fluxes of 03 and S02 to a forest canopy', Tellus, 40b, 270-284.
Meyers, T. P., P. L. Finkelsteto, J. Clarke, T. G. Ellestad, and P. F. Sims: 1998, 'A
multi-layer model for inferring dry deposition using standard meteorological
measurements', J. Geophys. Res., 103,22645-22661.
Munger, J. W., S, C. Wofsy, P. S. Bakwin, S-M. Fan, M. L. Goulden, B. C. Daube, A. H.
Goldstein, K. E. Moore, and D. R. Fitzjarrald: 1996, 'Atmospheric deposition of reactive
nitrogen oxides and ozone in a temperate deciduous forest and subarctic woodland, 1.
Measurements and mechanisms', J.Geophys.Rts., 101, D7,12,639-12,657.
Musselman, R. C., and W. J. Massman: 1999, 'Ozone flux to vegetation and its
relationship to plant response and ambient air quality standards', Atmos. Env., 33,65-73.
Musselman, R, C., and T. J. Minniek: 2000, 'Nocturnal stomatal conductance and
ambient air quality standards for ozone', Atmos. Env., 34,719-733.
Padro, J., G. denHartog, and H. H. Neumann: 1991, 'An investigation of the ADOM dry
deposition module using summertime O, measurements above a deciduous forest',
Atmos. Env., 25,1689-1704.
Padro, J., W. J. Massman, G. Den Hartog, and H. H. Neumann: 1994, 'Dry deposition
velocity of Oj over a vineyard obtained from models and observations: The 1991
California ozone deposition experiment', Water, Air and Soil Pollution, 75,307-323

-------
7
Figure 1. Average deposition velocity, ± 1 standard error, for 03 and S02 during the day (A) and at night (B) at five
sites; Bondville (BV), Sand Mountain (SM), Nashville (NV), Kane (K) and Sand Flats (SF).

-------
O W
Figure 2. Average daytime deposition velocity (DV) for 03 and S02 vs. leaf area index (LAI) at soybean (S), Forest
(F), com (C), pasture (P) and estuary (W) sites.

-------
so.
1.5
1.0
«
E
o,
>
Q
0.5
0.0
1.5
1.0
«
E
>
Q
0.5
0.0
8 12 16
HR (LST)
t
\

(A)
	 SF
-o- K
-o- BV
SM
¦¦£>• NV
(B)
8 12 16
HR (LST)
20 24
	 SF
-o- K
-o BV
SM
NV
Figure 3. Average diurnal cycles of deposition velocity at five sites for S02 (A) and 03 (B).

-------
1.5
w
fj_
>
O
1.0
0.5
0.0
1.5
1.0
•52
I
§
0.5
0.0
SO„
,o
i
(A)
14 18 22
26 30
WEEK
34 38 42
SF
K
•O BV
SM
46 --D- NV
10
0,
* a tr\
-\
(B)
	 SF
k
•O BV
SM
14 18 22 26 30 34 38 42 46 ~a'
WEEK
Figure 4. Weekly average daytime values of deposition velocity, showing the seasonal cycles, for S02 (A) and 03
(B) at five sites.

-------
NERL—RTF—AMD—00-150	TECHNICAL REPORT DATA
1. REPORT NO.
EFA/600/A-01/008
2.
3. RE'
4. TITLE AND SUBTITLE
Deposition Velocities of S02 and 03 Over Agricultural
and Forest Ecosystems
5.REPORT DATE
6.PERFORMING ORGANIZATION CODE
7, AUTHOR(S)
Peter L. Finkelstein
8.PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
National Exposure Research Laboratory - RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
10.PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Same as 9.
13.TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The results of field studies that measured the flux and deposition velocity of SO, and 03 are reported. Three of the studies
were over agricultural crops (pasture, corn, and soybean), and two were over forest (a deciduous forest and a mixed
coniferous-deciduous forest). In all cases the deposition velocity for SO, was higher than that for 03. Diurnal cycles of
SO, deposition velocity were similar in shape, but not magnitude for all surfaces; however those for 0} showed some
difference between forest sites where the peak was in the morning, and the agricultural sites where the peak ocurred at
midday. Seasonal cycles of SOz were affected by deposition to surfaces when leaves weren't active, yet surface
conductance is significant, but not for 03 where stomatal uptake is the primary pathway for deposition.
17.	KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/ OPEN ENDED
TERMS
c. COS ATI



18. DISTRIBUTION STATEMENT
1
19. SECURITY CLASS (Tim Report)
21. NO. OF PAGES
20. SECURITY CLASS (This Page)
Unclassified
22. PRICE

-------