Examining the Temporal Variability of Ammonia and Nitric Oxide Emissions
from Agricultural Processes
Thomas E Pierce*
Atmospheric Modeling Division
National Exposure Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Email: pierce.tQm@epg.gov
Lucille W Bender
DynTel Corporation
Research Triangle Park, North Carolina 27709
Paper to be published in the Proceedings of the Air and Waste Management Association /
U.S. Environmental Protection Agency (AWMA/EPA) Emission Inventory
Conference, Raleigh, North Carolina, October 26-28,1999
*on assignment from the National Oceanic and Atmospheric Administration,
Air Resources Laboratory
-------
ABSTRACT
This paper examines the temporal variability of airborne emissions of ammonia from
livestock operations and fertilizer application and nitric oxide from soils. In the United
States, the livestock operations and fertilizer application categories comprise the majority
of the ammonia emissions inventory. Air quality modeling efforts for the most part have
assumed annual-average ammonia emission factors. Based on a literature review, we
have generated crude seasonal adjustments of ammonia emissions taking into account
climatic factors, manure spreading, and fertilizer application schedules. Nitric oxide
(NO) emissions from soils estimated with the Biogenic Emissions Inventory System
(BEIS2) comprise about 10% of total annual nitric oxide emissions across the United
States. BEIS2 distributes emissions by land use and modulates emissions based on
hourly soil temperature, with the highest emissions arising from fertilized soils during
warm conditions. A new algorithm has been developed that incorporates daily rainfall
patterns, fertilizer application schedules, and plant canopy growth. Simulations with this
new algorithm show more short-term variability and an overall reduction in soil NO
emissions compared to the BEIS2 algorithm, particularly in the midwestern United
States. The techniques introduced for estimating the temporal variability of ammonia and
nitric oxide emissions from agricultural operations may help improve the accuracy of fine
particulate and ozone models.
INTRODUCTION
The United States is one of the most productive agricultural nations in the world.
Unfortunately, intensive agricultural processes cause appreciable amounts of nitrogen
compounds to be emitted into the atmosphere. These emissions include ammonia (NHs)
from fertilizer application, NHa from animal husbandry, and nitric oxide (NO) from
fertilized agricultural soils. Both NHj and NO play an important role in tropospheric
chemistry and should be considered for simulating photochemical oxidants, acid
deposition, fine paniculate matter, and visibility. While annual average emission factors
exist for these agricultural categories, air quality models need better temporal resolution.
In this paper, we will introduce crude temporal adjustments for ammonia emissions and
will propose a more sophisticated approach for soil nitric oxide emissions.
AMMONIA EMISSIONS
According to the 1997 National Emissions Trends (NET) inventory (Nizich et al., 1998),
agricultural categories comprise 85% of the ammonia emissions in the United States. In
reviewing this inventory for seasonal adjustments, we noticed ~25% drop in total
ammonia emissions between 1992 and 1997, due almost entirely to a 40% decrease of
emissions in the livestock category. To examine the reasons for this decline, Table 1
compares an independent "back of the envelope" estimate with the NET estimate. The
independent estimate is based on recently published 1997 Census of Agriculture animal
population statistics and Battye et al. (1994) ammonia emission factors. The two 1992
estimates agree reasonably well, but the 1997 NET estimate is nearly 50% smaller than
the independent estimate. Conversations with EPA's Office of Air Quality Planning and
Standards revealed that economic data were used to "grow" the NET emissions from
-------
1992 to 1997, as agricultural census information was unavailable at the time that the NET
inventory was constructed. Following assembly of the NET inventory during 1999, the
U.S. Department of Agriculture published animal population numbers for 1997, showing
only small differences in animal populations between 1992 and 1997. The decrease in
ammonia emissions in the NET inventory can be attributed to a dip in agricultural
commodity prices between 1992 and 1997. For estimating year-to-year trends in
livestock emissions, it is therefore recommended that animal population estimates be
used rather than economic price estimates.
Table 1. The NET inventory versus an independent estimate of ammonia emissions from
livestock for 1992 and 1997. The independent estimate is based on agricultural statistics
from the U.S. Department of Agriculture (ht^://wjmv.n_ass.usda.gov/census/) and
emission factors from Battye et al. (1994).
1992
96.1 x 106 cows x 22,9 kg NH3/cow-year
57.6 x 106 x 9.2 kg NH3/hog-year
351.3 x 106 layers-pullets x 0.18 kg NH3/chicken-year
835.2 x 106 broilers x 0.18 kg NH3/chieken-year
87.6 x 106 turkeys x 0.86 kg NH3/turkey-year
10.8 x 106 sheep x 3.4 kg NH3/sheep-year
2201xl06kgNH3
530xl06kgNH3
63xl06kgNH3
150xl06kgNH3
75xl06kgNH3
37xl06kgNH3
Total independent calculation;
1992 NET inventory (livestock):
3056xl06kgNH3
2785xl06kgNH3
1997
99.0 x 106 cows x 22.9 kg NH3/cow-year
61.2 x 106 hogs x 9.2 kg NH3/hog-year
367.0 x 106 layers-pullets x 0.18 kg NH3/ehieken-year
1037.2 x 106 broilers x 0.18 kg NH3/chicken-year
104.3 x 106 turkeys x 0.86 kg NH3/turkey-year
7.8 x 106 sheep x 3.4 kg NH3/sheep-year
2267xlO°kgNH3
563xl06kgNH3
66 x 106kgNH3
187xl06kgNH3
90xl06kgNH3
27xl06kgNH3
itotal independent calculation:
£200x10
1997 NET inventory (livestock):
In Table 1, the 1992 NET estimate is -10% less than the independent estimate. This
relatively small underestimate is caused by omissions in the NET poultry sub-category.
Based on internal worksheets provided by OAQPS (T, Pace, personal conversation),
poultry contributes 28.6 x 106 kg NH3 to the 1996 NET inventory. Based on an emission
-------
factor of 0,18 kg NHa/chicken, this implies an annual chicken population of 159 million.
However, USDA animal statistics for 1997 indicate a layer/pullet population of 367
million and a commercial broiler production of 6742 million. Assuming that a new flock
of broilers is produced every 8 weeks, this implies that the annual average chicken
population to be used for the ammonia emission calculation should actually be 1404
million rather than 159 million, with resulting ammonia emissions from poultry estimated
at 253 x 106 kg. Turkeys also seem to have been overlooked, as none of the NET
breakdowns seems to include a turkey category. As shown in Table 1, turkeys would add
about 75 x 1 06 kg NHs to the national inventory.
After reviewing the livestock emissions reported in the NET inventory from 1992 to
1997, we offer the following recommendations:
(1) Inventories should use the latest available (or if necessary, projected) animal
population statistics.
(2) Inventories should include a category for turkeys.
(3) Inventories should account for the total chicken population (by including
layers/pullets — called "all chickens" by the USDA — and commercial broilers).
[Note: In preliminary versions of the 1998 NET inventory, these recommendations
appear to have been adopted.]
Now that some adjustments have been recommended for the year-to-year estimates of
ammonia emissions, we will attempt to address the seasonal distribution of fertilizer and
livestock emissions. For the fertilizer sub-category, seasonal allocations can be derived
from the seasonal distribution of fertilizer by crop type. Fertilizer amounts by crop type
from the U.S. Department of Agriculture (1996) and national agricultural land statistics
were used to construct Table 2.
Table 2. Distribution of nitrogen-based fertilizer usage by crop in the United States.
Crop
Corn
Wheat
Cotton
Soybeans
Miscellaneous
Fertilized area
(106ha)
28.26
23.81
5.34
2.68
1.54
Typical
fertilizer
application rate
(kg N/ha)
141
63
93
6
141
Total fertilizer
(106kgN)
3985
1500
497
16
217
Fertilizer
(percent of
total)
63%
24%
8%
2%
3%
-------
Corn and cotton are usually planted (and fertilized) from March to June (U.S. Department
of Agriculture, 1997). Therefore, it is assumed that ammonia emissions from fertilizer
applied to com and cotton (which comprise 71% of the total fertilizer) occur 75% during
the spring and 25% during the summer. Because wheat uses most of the other fertilizer
and is planted either as winter wheat or as spring wheat, and because some agricultural
crops (in the warmer climates) are planted year-round, it is assumed that the remaining
29% of the fertilizer is distributed equally among all four seasons. Based on this
rationale, seasonal allocations for ammonia emissions from fertilizer application are as
follows: winter -- 7%, spring — 60%, summer — 26%, and fall -- 7%. Rounded seasonal
allocations are proposed in Table 3.
Table 3, Proposed seasonal allocations of ammonia emissions from fertilizer application.
Season (months)
Winter (DJF)
Spring (MAM)
Summer (JJA)
Autumn (SON)
Allocation
10%
50%
30%
10%
Seasonal allocations for livestock are more difficult to pinpoint because of differences in
animal husbandry practices among the various animal species and the location of the
various emission "release" points. Most of the emissions in the annual inventory come
from cattle, swine, and poultry. These animals live either in controlled housing,
stockyards, or pasture. As a first approximation, housing emissions can be assumed to be
constant because of the controlled conditions and relatively stable rates of nutrition
intake, European measurements, however, raise questions regarding this assumption.
Hartung (1991) and Mannebeck and Oldenburg (1991) report a doubling of emissions
when ambient temperatures increased from 0 to 20°C, but Oosthoeck et al. (1991) report
that emissions from a dairy house remained nearly constant from January through June.
One way of examining temporal variability through the various stages of a livestock
operation is with a process model. Hutchings et al. (1996) developed and applied a
whole-farm model for a farm in the United Kingdom. Ammonia emissions for beef and
dairy cattle were distributed between grazing, housing, storage, spreading, and
application of manure. As shown in Table 4, modeled emissions from a beef cow and a
dairy cow differed by a factor of six (4.2 kg-N/year versus 25.5 kg-N/year respectively).
About half of this difference may be attributed to the increased N needs of the dairy cow,
and the other half may be attributed to the increased storage and application of the slurry
from a dairy cow that reportedly volatilizes more readily than a grazing cow's ammonia.
Although seasonal variations were not reported in this paper, process models like this
might be exercised with the appropriate animal husbandry regimes to estimate more
accurate emission factors and seasonal allocations.
Table 4. Distribution of ammonia losses from a cattle operation in the United Kingdom,
based on the model of Hutchings et al. (1996).
-------
Beef
Dairy
Grazing
25%
14%
House
9%
7%
Storage
28%
47%
Spreading
3%
3%
Applied
36%
29%
Lacking an appropriate process model for the United States, we have cursorily examined
mostly European literature to gain some guidance on seasonal allocations. First, we
assume that housing of livestock (which we are assuming to be uniform) contributes up to
50% of the emissions with the remainder distributed among lagoons/pits, slurry
application, and grazing, Bouwmeester and Vlek (1981) based on theoretical
considerations and wind tunnel measurements indicate that the rate of ammonia
volatilization from a wet solution increases by about a factor of two with a 10°C increase
in temperature. Yamulki et al. (1996) also report higher losses of ammonia from
agricultural fields during warm, dry conditions than during cool, damp periods. Harper et
al. (1983) measured ~50% increase in emissions from grazing cattle when temperatures
increased by ~10°C. Amberger (1991) measured emissions from cattle and pig slurries
and found that emissions increased by 50-150% when temperatures increased from 5°C
to 15°C. This study also found that the presence of vegetation (which becomes more
prevalent during the late spring) decreases volatilization by up to 50%. The literature
suggests that ammonia emissions vary seasonally when animal excrement is stored
outside in lagoons or pits (where temperature and moisture changes affect volatilization
rates), when slurry and poultry manure is applied to fields (where emissions increase due
to increased application rates and volatilization rates during warm/dry conditions), and
when grazing occurs (where activity patterns increase when animals are outside during
warm/dry conditions). Therefore, seasonal allocations for animal husbandry are
suggested in Table 5.
Table 5. Proposed seasonal allocations of ammonia for animal husbandry operations.
Season (months)
Winter (DIP)
Spring (MAM)
Summer (JJA)
Autumn (SON)
Allocation
15%
25%
40%
20%
NITRIC OXIDE EMISSIONS
The soil NO emissions algorithm in the second version of the Biogenic Emissions
Inventory System (BEIS2) considered only the influence of soil temperature (Williams et
al., 1992). In support of global tropospheric modeling studies, Yienger and Levy (1995)
introduced an empirical algorithm that considers soil moisture, fertilizer application
schedules and amount, and plant canopy interception of NO emitted by the soil. In
testing their approach for the next version of BEIS (BEIS3), we have made the following
assumptions:
(1) Most soil NO emitted in the United States comes from fertilized soils, and these
areas should be the focus for further algorithm development.
-------
(2) The emission of NO from fertilized soils is proportional to the N application rate,
and this loss rate is assumed to be 2.5% N as NO integrated through the growing
season.
(3) The pulse of emissions due to fertilizer application is assumed to be uniform for
the first 30 days of the growing season, and the loss rate is decreased linearly to
the grassland emission flux by the end of the growing season,
The BEIS2 algorithm used the following simple equation (Williams et al.51992) to
account for soil temperature:
ET= Ex exp[0.071x (T- T,)]
where ET is the emission rate at soil temperature T, E is the normalized emission rate at
30°C, and Ts is the normalized soil temperature (30°C). The new algorithm uses the
following equation;
E = R2.S X Tadj X Padj X Fadj X Cadj
where E is the emission rate corrected for environmental conditions, Ra.s is the
normalized emission rate assuming that 2.5% of the nitrogen is emitted as NO across the
growing season, TaCjj is the temperature adjustment factor, Padj is the precipitation
adjustment factor, Fadj is the fertilizer adjustment factor, and Cadj is the canopy adjustment
factor. To obtain Rj.5, fertilizer application rates were obtained for the U.S. Department
of Agriculture. Based on a loss rate of 2.5% and assuming a growing season of 214 days
(April 1 - October 31), normalized emission factors are shown in Table 6.
Table 6, Normalized soil NO emission rates assumed for BEIS3 as compared to BEIS2.
BEIS3 rates are based on a loss rate of 2.5% from nitrogen fertilizer and a growing
season of 214 days. Rate for miscellaneous crops is based on average fertilizer
application of all crops.
Crop
Potatoes
Corn
Sorghum
Barley
Cotton
Oats
Tobacco
Miscellaneous
Wheat
Alfalfa
Hay
Pasture
BEIS3 factor
(ugNOnf2!!*1)
258
145
145
97
97
97
9?
85
65
58
58
58
factor
(ugNOm'2h-J)
193
578
578
257
257
257
257
13
193
13
13
58
-------
Peanuts 58 13
Rye 58 13
Soybeans 58 13
Rice 1 1
The temperature adjustment (Tadj) is handled via the following set of expressions:
7W=0 forr<0
Tadj = exp(0.103 x T) 121.97 for 0 < T < 30
Tad/= I forJ>30
where T is given in °C. For the pulsing effect of emissions due to precipitation (Padj), an
empirical enhancement factor is applied if rain has fallen during the previous 13 days.
The enhancement factor depends on the amount of rain and the number of days after rain
as shown in Table 7.
Table 7. Precipitation enhancement factors (Padj) applied after a rainfall event.
Day
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Sprinkle
(1-5 mm)
5.0
2.2
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1,0
1.0
1.0
Shower
(5-15 mm)
10.0
6.8
4.6
3.2
2.2
1.5
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
Heavy rain
(>15 mm)
15.0
12.2
9.9
8.0
6.5
5.3
4.3
3.5
2.8
2.3
1.9
1.5
1.2
1.0
Fertilizer also produces a pulsing effect. Because this algorithm will be used across
regional modeling domains where fertilizer application will occur for several weeks, the
initial "burst" of NO from fertilizer is assumed to be elevated for the first 30 days of the
growing season (when most fertilizer is applied). Afterwards, emissions are assumed to
decrease linearly until the NO emission rate for grasslands is reached. The fertilizer
adjustment factor (F2dj) is calculated with the following sets of equations:
j =1 for d < 30
~
dg- 30
for 30< d < dg
-------
where d is the day of the growing season, and dg is the length of the growing season
(assumed here to be 214 days, April 1 - October 31).
The final adjustment factor is the reduction of emissions that occurs as the plant canopy
grows and intercepts the nitrogen emitted from the soil. We assume that emissions are
unattenuated during the first 30 days of the growing season. During the second 30 days,
emissions are reduced linearly down to a value of 50%. The set of equations used for this
adjustment factor (Cadj) is as follows:
Cadj = I ford<3Q
C*ij = 1.5 -(d/ 60) for30 60
This new algorithm has been tested for emissions from corn with meteorological data
from Des Moines, Iowa and compared to the BEIS2 algorithm as shown in Figure 1 .
Emissions with the new algorithm tend to be higher than the BEIS2 algorithm during the
first part of the growing season, when rain is more frequent, fertilizer effects are highest,
and canopy cover is small. By the middle of the growing season, the new algorithm is
consistently lower than the BEIS2 algorithm, particularly during dry, hot periods.
Overall, the new algorithm for this test example was a factor of 3 lower than the BEIS2
algorithm (91 p.g m"2 h"1 versus 27 ng m"2 h"1).
Next, the algorithm was implemented in an Urban Airshed Model (UAM) version of
BEIS2, with precipitation amounts provided from a processed MM5 meteorological
model data file. For a July 1995 simulation for the eastern United States, the new
algorithm gave 15% lower NO emissions than the BEIS2 algorithm (1517 metric tons
N/day versus 1815 metric tons N/day) as shown in Figure 2, After periods of elevated
rainfall (such as July 3-5), the emissions estimated with the proposed BEIS3 algorithm
rose above those estimated with BEIS2. Figures 3 and 4 compare the spatial patterns of
soil NO estimated for July 15, 1995. The new algorithm produces "hot" spots in the
midwestern United States, which correspond to those areas having recent rainfall. For
most areas, however, emissions with the new algorithm are lower than those estimated
with BEIS2.
Although observational data are unavailable for testing, simulations indicate that the new
algorithm reliably follows the approach proposed by Yienger and Levy (1995) thus
yielding soil NO emission patterns that are more representative of environmental and
temporal influences than the existing BEIS2 algorithm.
CONCLUSION
The temporal variability factors introduced here for ammonia emissions from agricultural
operations represent a crude step towards improved temporal characterization. Better
activity data should be gathered for animal livestock practices and field measurements
should be taken from animal husbandry operations typical to the United States. The
-------
proposed algorithm for soil nitric oxide emissions is more sophisticated than the existing
BEIS2 algorithm, which considers only soil temperature. However, the temporal
variability modeled for soil NO needs to be verified with field data. In addition, the
sensitivity of these temporal estimates versus annual average estimates for ammonia and
NO should be tested in air quality modeling studies,
ACKNOWLEDGEMENTS AND DISCLAIMER
The authors appreciate the help and guidance of Lara Joyce (DynTel), Bill Cure (North
Carolina Department of Natural Resources), Tom Pace and Sharon Nizich (U.S.
EPA/Office of Air Quality Planning and Standards),
This paper has been reviewed in accordance with the U.S. Environmental Protection
Agency's peer and administrative review policies and approved for presentation and
publication. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
REFERENCES
Amberger, A. (1991) Ammonia emissions during and after land spreading of slurry, in
Odour and Ammonia Emissions from Livestock Farming, V. Nielsen, J. Voorburg, and P.
L'Hermite (eds.), Eisevier, New York, NY, pp. 126-131.
Battye, R,, W. Battye, C, Overcash, and S. Fudge (1994) Development and Selection of
Ammonia Emission Factors, Report prepared for the U.S. Environmental Protection
Agency, Office of Research and Development, Washington DC, 99 pp.
Bouwmeester, R. and P. Vlek (1981) Rate control of ammonia volatilization from rice
paddies, Atmos, Environ., 15, pp. 131-140.
Harper, L., V. Catchpoole, and I. Vallis (1983) Gaseous ammonia transport in a cattle-
pasture system, in Nutrient Cycling in Agricultural Ecosystems, R. Lowrance, R. Todd,
L, Asmussen, and R, Leonard (eds.), University of Georgia College of Agricultural
Experimental Stations, Special Publication 23, pp. 353-372.
Hartung, J. (1991) Influence of housing and livestock on ammonia release from
buildings, in Odour and Ammonia Emissions from Livestock Farming, V. Nielsen, J.
Voorburg, and P, L'Hermite (eds.), Eisevier, New York, NY, pp. 22-30,
Hutehings, N., S. Sommer, and S. Jarvis (1996) A model of ammonia volatilization from
a grazing livestock farm, Atmos, Environ,, 30.589-599.
Mannebeck, H. and J, Oldenburg (1991) Comparison of the effects of different systems
on ammonia systems, in Odour and Ammonia Emissions from Livestock Farming, V.
Nielsen, J. Voorburg, and P. L'Hermite (eds.), Eisevier, New York, NY, pp. 42-49.
-------
Oosthoek, J., W. Kroadsma, and P. Hoelesma (1991) Ammonia emission from dairy and
pig housing systems, in Odour and Ammonia Emissions from Livestock Farming, V.
Nielsen, J, Voorburg, and P. L'Hermite (eds.), Elsevier, New York, NY, pp. 31-41.
U.S. Department of Agriculture (1996) Agricultural Chemical Usage, 1995 Field Crops
Survey, National Agricultural Statistics Service, Washington, DC
(http://usda.mannlib.comell.edu).
Williams, E., A. Guenther, and F. Fehsenfeld (1992) An inventory of nitric oxide
emissions from soils in the United States, J, Geophys. Res.. 97, pp. 7511-7519.
Yamulki, S,, R. Harrison, and K. Goulding (1996) Ammonia surface-exchange above an
agricultural field in southeast England, Atmos. Environ., 30, pp. 109-118.
Yienger, J. and H. Levy (1995) Empirical model of global soil-biogenic NOx emissions,
/. Geophys. Res,. 100, pp. 11447-11464.
FIGURES
Figure 1. Daily NO emission fluxes computed using the BEIS2 and BEIS3 algorithms for
corn based on meteorological data from Des Moines, Iowa for 1995.
Figure 2. Daily soil NO emissions summed over the UAM's modeling domain for the
eastern United States for July 1995. Precipitation data for the BEIS3 algorithm were
obtained from a MM5 simulation.
Figure 3. Daily soil NO emissions from 13 July 1995 calculated using the BEIS2
algorithm.
Figure 4. Daily soil NO emissions from 13 July 1995 calculated using the proposed
BEIS3 algorithm.
-------
Des Moines, Iowa
700
600
500
O
4/1
4/29
5/27
6/24
7/22
8/19
9/16
10/14
Soil NO-July 1995
2500
2000
1500
O
-------
Layer 1 no
Beis2
20000.0(89
15000.000
10000.000
5000.000
0.000 i
moles per day] 192
Pf;VE July 13.1995 0:00:00
MCNC Min= 0.000 at (13,1). Max=44547.973 at (7,135)
Beis3
20000.0(88
15000.000
10000.000
5000.000
0.000
moles perdayi
MCNC
July 13,1 995 0:00:00
Min= 0.000 at (13,1), Max=3921 4.766 at (8,88)
-------
100670
1. REPORT NO,
EPA/600/A-00/031
4. TITLE AND SUBTITLE
Examining the temporal vs
nitric oxide emissions fa
TECHNICAL REPORT DATA
2.
iriability of ammonia and
"om agricultural processes
7. AUTHOR (S)
Thomas E. Pierce
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Atmospheric Modeling Division
National Exposure Research Laboratory
U.S. Environmental Protection Agency
79 T.W. Alexander Drive, MD-80
Research Triangle park, NC 27711
12. SPONSORING AGENCY NAME AND
NATIONAL EXPOSURE RESEARC
OFFICE OF RESEARCH AND DI
U.S. ENVIRONMENTAL PROTEC
RESEARCH TRIANGLE PARK, 1\
ADDRESS
H LABORATORY
"VBLOPMENT
TION AGENCY
C 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTR&CT/GRAHT NO.
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This paper examines the temporal variability of airborne emissions of ammonia from
livestock operations and fertilizer application and nitric oxide from soils. In the
United States, the livestock operations and fertilizer categories comprise the
majority of the ammonia emissions inventory. Air quality modeling efforts for the
most part have assumed annual -average ammonia emission factors. Based on a
literature review, we have generated crude seasonal adjustments of ammonia emissions
taking into account climatic factors, manure spreading, and fertilizer application
schedules. Nitric oxide (NO) emissions from soils estimated with the Biogenic
Emissions Inventory System (BEIS2) comprise about 10% of total annual nitric oxide
emissions across the United States. BEIS2 distributes emissions by land use and
modulates emissions based on hourly soil temperature, with the highest emissions
arising from fertilized soils during warm conditions. A new algorithm has been
developed that incorporates daily rainfall patterns, fertilizer application
schedules, and plant canopy growth. Simulations with this new algorithm show more
short-term variability and an overall reduction in soil NO emissions compared to the
BEIS2 algorithm, particularly in the Midwestern United States. The techniques
introduced for estimating the temporal variability of ammonia and nitric oxide
emissions from agricultural operations may help improve the accuracy of fine
particulate and ozone models .
17.
a. DESCRIPTORS
KEY WORDS AMD DOCUMENT AHALYSIS
b. IDENTIFIERS/ OPEM ENDED TERMS
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This Page)
Unclassified
c . COSATI
21. NO. OF PAGES
22. PRICE
------- |