Measurements of ammonia emissions from oak and pine forests and development of a non-industrial ammonia emissions inventory in texas

ARTICLE IN PRESS

Atmospheric Environment 39 (2005) 7137–7153 www.elsevier.com/locate/atmosenv

Measurements of ammonia emissions from oak and pine forests and development of a non-industrial ammonia emissions inventory in texas Golam Sarwara,1, Richard L. Corsia,, Kerry A. Kinneya, Joel A. Banksa, Vince M. Torresa, Chuck Schmidtb a

Center for Energy and Environmental Resources (R7100), The University of Texas at Austin, 10100 Burnet Road, Austin, TX 78758, USA b Independent Environmental Consultant, 19200 Live Oak Road, Red Bluff, CA 96080, USA Received 23 March 2005; accepted 14 August 2005

Abstract Estimates of non-industrial source ammonia emissions in Texas were developed through the use of published emission factors and activity data for those sources. A total of 64 non-industrial source emission sub-categories were addressed, each falling into one of seven major source categories: animal husbandry, fertilizer applications, on-road vehicles, nonroad sources, municipal wastewater disposal, domestic sources, and natural soil and vegetation. Annual statewide ammonia emissions were initially estimated to be 921,000 metric tons, with greater than 50% originating from natural soil and vegetation. However, estimates for pine and oak forests were characterized as having a great deal of uncertainty. A series of field sampling events were conducted to determine ammonia fluxes from pine and oak forest floors in east Texas. Both dynamic and static chamber methods were used. The ammonia flux averaged 0.09 kg km
2 month
1 for pine forests and 0.13 kg km
2 month
1 for oak forests. These values are significantly lower than those previously measured and cited in the published literature. However, the ammonia fluxes measured in east Texas forests are reasonably consistent with those predicted using mechanistic models for evergreen pine and deciduous broadleaf forests in Alabama, California, Colorado, and Tennessee. Statewide annual ammonia emissions estimates, revised using the newly developed ammonia fluxes for oak and pine forests in Texas, dropped from 921,000 to 467,000 metric tons. The relative contribution of ammonia emissions from pine and oak forests dropped from 49% to less than 1%. Animal husbandry was predicted to be the dominant nonindustrial source, accounting for approximately 77% of non-industrial source ammonia emissions. r 2005 Elsevier Ltd. All rights reserved. Keywords: Ammonia emissions; Soil; Forests; Inventory

1. Introduction Corresponding author. Tel.: +1 512 475 8617;

fax: +1 512 471 1720. E-mail address: [email protected] (R.L. Corsi). 1 Current address: USEPA, Mail Drop E243-03, 109 T.W. Alexander Drive, RTP, NC 27711, USA. 1352-2310/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2005.08.016

Atmospheric chemical reactions are believed to be a major source of fine particulate matter (PM2.5). An important contributor toward many of these reactions is ammonia (NH3), which is emitted from a wide range of anthropogenic and natural sources.

ARTICLE IN PRESS 7138

G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

Tropospheric concentrations of NH3 are highly variable and dependent on proximity to sources, source strengths, meteorological conditions, and removal mechanisms. Typical atmospheric residence times for NH3 are on the order of 10 days, and ammonia mixing ratios over continents generally range over two orders of magnitude, from 0.1 to 10 parts per billion by volume (ppb) (Seinfeld and Pandis, 1997). Dentener and Crutzen (1994) estimated global NH3 emissions to be 45 million metric tons per year, with approximately two-thirds of this total being attributed to anthropogenic activities. Nearly one-half of global ammonia emissions were attributed to animal husbandry. Chemical reactions involving NH3 to produce secondary PM2.5 depend on the presence and relative concentrations of atmospheric nitrates and sulfates. In areas characterized by high ammonia and nitric acid concentrations and low sulfate concentrations, gaseous ammonia can react to form ammonium nitrate. In the presence of sulfuric acid, increasing concentrations of gaseous ammonia can react to form ammonium sulfate. Whether reacted with nitrate or sulfate, the ammonium ion (NH+ 4 ) is often observed to be an important component of tropospheric aerosols. The conversion of ammonia to ammonium (NH+ 4 ) is also significant with respect to transport of NHx, since the dry deposition of ammonia gas is generally 5–10 times faster than dry or wet deposition of ammoniumcontaining particles (Bouwman et al., 1997). There is significant evidence that natural soil is an important contributor to global ammonia emissions (Dawson, 1977). For example, ammonium is found at relatively high concentrations in rainwater. Gaseous concentrations of ammonia are also greater over soils with high pH, a condition that shifts the acid-base equilibrium in soil from ammonium ion to ammonia. Atmospheric ammonia concentrations are greater over land than over oceans, and increase with increasing soil temperature. However, measurements of ammonia emissions from natural soils are sparse and corresponding emission factors are characterized by significant uncertainties. These facts are particularly true for ammonia emissions from forested areas, e.g., pine and oak forests that cover large areas of east Texas. The primary source of nitrogen that is converted to ammonia is organic nitrogen associated with foliar litter. Thus, greater amounts of fresh litter deposition should lead to increased ammonia emissions. The intent of the study described herein was to develop a first estimate of non-point source

ammonia emissions in Texas. A total of 64 nonpoint sources of ammonia were considered in this study. Each source required significant reviews of existing literature and relevant databases prior to the selection of appropriate emission factors and source activity data. Given the extensive nature of these tasks, it is impossible to describe all aspects of the study in this paper. Instead, we have described the project methodology in general terms, and have listed several important references and databases. The reader is referred to the complete project report for a more extensive description of methodologies and results (Corsi et al., 2000a). We do provide details related to actual ammonia flux measurements from forest floors in east Texas and use the results to facilitate the overall ammonia inventory. 2. Methodology An extensive literature review was completed in order to identify potential non-industrial sources of ammonia emissions; As well as to identify and assess relevant emission factors. Ten bibliographic databases were searched using ‘‘ammonia’’ and ‘‘emissions’’ (inclusive) as keywords. Forty web sites were also found to contain information related to ammonia emissions, 14 of which were identified as relevant to this project. Personal contacts were also made with individuals known to be, or who are known to have been, involved with ammonia emissions estimates. In total, 655 publications were identified as containing information relevant to this project. Approximately 120 of these publications were selected for thorough review. Through this process, it quickly became evident that a small number of previous publications are frequently referenced and used by others to estimate ammonia emissions (Asman, 1992; Battye et al., 1994; Bouwman et al., 1997; Buijsman et al., 1987; Gharib and Cass, 1984; Klaassen, 1991; and Lee and Dollard, 1994). 2.1. Selection of source categories Based on a review of existing literature, a total of seven non-industrial source emission categories were selected for this study. Sixty-four sub-categories that fall within the major source categories are listed in Table 1. While it was obvious at the beginning of this study that some sources would be relatively insignificant, e.g., rabbits and untreated human waste, for completeness emission factors and

ARTICLE IN PRESS G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

7139

Table 1 Source categories considered in this study Primary source category

Sub-categories associated with each primary source

Animal husbandry

Cattle & calves (beef cows [3], milk cows [17], generic cows [21]); goats [4]; hogs & pigs [27]; horses [20]; mules, burros, & donkeys [0]; poultry (broilers [9], laying hens [13]; pullets under 13 weeks [1], turkeys [12], ducks [4], geese [0]); rabbits [2]; sheep (sheep & lambs—composite [23])

Fertilizer application

Ammonia (liquid) [1]; ammonium nitrate [12]; ammonium phosphate [2]; ammonium sulfate [2]; anhydrous ammonia [6]; mono-ammonium phosphate [1]; di-ammonium phosphate [1]; calcium ammonium nitrate [2]; NP-K [1]; other nitrogen solutions [1]; other N-P [1]; urea [25]

On-road vehicles

Diesel engines (heavy-duty [8], light-duty [7]); gasoline engines (heavy-duty [4], light-duty without catalyst [14], light-duty with three-way catalyst [18])

Non-road sources

Agricultural vehicles [0]; aircraft [2]; commercial equipment [0]; commercial marine vehicles [4]; construction & mining equipment [0]; industrial equipment [0]; lawn & garden equipment (commercial [0], residential [0]); logging equipment [0]; pleasure craft [0]; railroad (locomotive engines [2]); recreational equipment [0]

Municipal wastewater

Publicly owned treatment works (POTWs) [1]; residential septic tanks [0]

Domestic sources

Cats [9]; cigarettes [2]; cleaning products [1]; diapers [2]; dogs [9]; humans (perspiration & respiration) [18]; untreated human waste (homeless [1], other than homeless [1])

Natural soil/vegetation

Coniferous forest (pine [1], other [2]); desert [11]; grassland [5]; pasture [4]; rangeland [3]; scrubland [4]; temperate forest (dense oak [1], other [6]); urban land [1]

Note: Numbers in brackets correspond to the number of reported emission factors for each source.

activity data were sought for all source subcategories. This paper will focus entirely on those source sub-categories that were estimated to emit greater than 1000 metric tons yr
1 (mtpy) of ammonia on a statewide basis. Relevant sources are italicized in Table 1. 2.2. Emissions estimation Emissions for individual source sub-categories were estimated as the product of an emission factor and relevant activity data for that source: E i ¼ ðEFÞi Ai ,

(1)

where Ei is the emission rate for source sub-category i (kg yr
1), EFi the emission factor for source subcategory i (kg unit activity
1, e.g., kg vehicle mile traveled
1), and Ai the activity level (measure of activity) for source sub-category i (e.g., miles traveled yr
1). Emissions estimates for each of the 64 source sub-categories listed in Table 1 were determined for each of the 254 counties in Texas. County-specific emissions were estimated by allocating activity data to each county. Although temporal variations in ammonia emissions are likely to be significant for some source categories, e.g.,

fertilizer applications and natural soil and vegetation, attempts to estimate temporal variations in ammonia emissions were beyond the scope of this study. All emissions were estimated on an annual basis. County-level emissions for each of the seven primary source categories were estimated by summing over relevant source sub-categories for each county. Statewide ammonia emissions for each of the source sub-categories and primary source categories were estimated by summing over all counties. Total ammonia emissions for Texas were estimated by summing over all 64 source subcategories for all 254 counties in Texas. At the request of the sponsoring agency, emissions were standardized to a 1996 base year by using activity data specific to 1996. 2.3. Selection of emission factors A total of 348 emission factors were found for the 64 source sub-categories described above. The number of emission factors for each source is listed in brackets in Table 1. A tiered approach was used for selecting source-specific emission factors. In cases where two or more emission factors were available, an attempt was made to carefully assess

ARTICLE IN PRESS 7140

G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

the origin of each factor. Emission factors were excluded from further analysis for any one or more of the following: (1) incomplete description of factor development; (2) exclusion of all processes that are generally associated with a source; (3) development under conditions that are inconsistent with Texas; and (4) statistical inconsistency. In most cases, the remaining list of emission factors was then used to determine an arithmetic average emission factor for use in this study. Care was taken to avoid ‘‘extra weighting’’ of emission factors that were already included in those emission factors reported as averages of others. Emission factors by Harvey et al. (1983) were used for all on-road vehicles based on the large number of vehicles that were used in developing the factors as well as adherence to Federal Testing Procedures. An emission factor developed by Asman (1992) was used for ammonium nitrate fertilizer. For those sources in which only one published emission factor was found, that emission factor was used by default in this study. For example, values reported by either Asman (1992) or Bouwman et al. (1997) were used for all fertilizers that had only one reported emission factor. Finally, for those 12 source sub-categories without reported emission factors, values were assumed based on other sources, e.g., horses were used for mules, burros and donkeys, ducks for geese, diesel engines for agricultural vehicles. For septic tanks, it was assumed that the ammonia concentration in the headspace of the tank is always at equilibrium with the ammonia concentration in the underlying wastewater. Emissions were then estimated using Henry’s law and the assumption that gas displacement is equal to wastewater discharge to the tank. The contribution of these 12 source sub-categories to total statewide ammonia emissions was estimated to be only 0.05%. It is important to note that many of the emission factors, particularly for livestock, were developed based on studies completed in Europe. The authors acknowledge that ammonia emissions may differ between livestock raised in Europe and in Texas due to differences in diet. These differences may be significant for sources such as dairy cows, and may be underestimated in this study due to higher protein diets for livestock in the United States. However, because of a lack of actual data for livestock in the United States, nonetheless Texas, we have adopted several emission factors that were developed based on studies in Europe.

2.4. Acquisition of activity data A wide range of information sources was used to obtain necessary activity data. For example, the US Department of Commerce’s Census of Agriculture as well as the Texas Department of Agriculture’s report on Texas Agricultural Statistics was information sources for animal husbandry. The Association of American Plant Food Control Officials and the University of Kentucky were sources of information on commercial fertilizer sales and applications. The Texas Department of Transportation provided activity data related to vehicle miles traveled by various types of vehicles. The Texas Commission on Environmental Quality (TCEQ) provided data related to municipal wastewater flows on a county-by-county basis. Population data were obtained from the US Census for relevant domestic source sub-categories, as were data obtained from the American Veterinary Medical Association. Soil and vegetation coverage was estimated using the Biogenic Emission Inventory System (PC-BEIS). Two very important sources that were limited to a single emission factor were pine and oak forests. Use of the single emission factor led to an ammonia emission estimate of 257,000 mtpy for pine forest and 197,000 mtpy for oak forests in Texas, which accounted over 50% of the statewide annual emissions (Corsi et al., 2000b). However, estimates for pine and oak forests were characterized as having a great deal of uncertainty. As such, a series of field sampling events were completed to determine ammonia emissions from pine and oak forests in Texas.

2.5. Site selection process A set of criteria for selecting appropriate locations for field measurements was developed, which included the mix of tree species at a given location and requirements for monitoring equipment and activities. Forests were chosen that were representative of typical east Texas pine and oak forests. Virtually all the pine forests in east Texas consist of Loblolly Pine, and are relatively homogeneous. Therefore, selecting sites in pine forests for ammonia emissions monitoring was fairly straightforward. The situation was not as simple for oak forests. There are 28 species of oak found in Texas, and they occur in a wide range of terrains and vegetation type groupings.

ARTICLE IN PRESS G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

A screening analysis of potential sites was based on land use/land cover (LULC) and biomass density databases (Yarwood et al., 1997,1999; Wiedinmyer and Allen, 1999; Wiedinmyer and Strange, 1998). A detailed examination of vegetation data indicates that Post Oak (Quercus stellata) was the predominant oak species in Texas, with approximately 40% of the total oak biomass. This was followed by the two Live Oak species (Q. virginiana and Q. fusiformis) comprising approximately 17% of the oak biomass, and Shin Oak (Q. sinuata) with 10%. Thus, suitable Post Oak forests were identified for possible study sites. In addition, potential monitoring sites had to meet several requirements for the emissions measurement equipment and procedures. The site had to be within the interior of the forest, and far enough from urban, agricultural, or other developed areas to minimize edge effects and the influence of activities within these other areas. Electrical service had to be available near the monitoring sites to provide power for the test equipment. A minimum of two monitoring sites had to be available within each selected forest to allow for multiple samples, intended to compensate for spatial variation in ammonia emission characteristics. 2.6. Site descriptions Field sampling was completed in two pine forests (Davy Crockett National Forest and Sam Houston National Forest), and two oak forests (Purtis Creek State Park and Cooper Lake State Park located in east Texas). Each of these forests is hereafter referred to as a sampling ‘‘location’’. Each location was visited twice during the course of the study. Four sampling sites were selected at each location. A single sampling event was completed at each sampling site. A sampling event involved set-up of experimental instrumentation, at least one and possibly two sets of multi-hour flux chamber measurements, collection of soil samples, and ambient air sampling. Both Purtis Creek and Cooper Lake State Park were found to contain significant post oak stands and were used in this study. Purtis Creek State Park is located about 3.5 miles north of the city of Eustace, Texas and about 15 miles northwest of Athens, Texas. The monitoring site was located at 321210 N, 96100 W. The park (designated as Oak 1) surrounds a man-made lake. The trees were

7141

primarily Post Oak and Hickory, with some Mulberry, Eastern Red Cedar (Juniper), Cedar Elm, and occasional other Oaks. The soil was loose and sandy, with organic matter primarily in the top 5 cm. The monitoring site at Cooper Lake State Park was located at 331220 N, 951400 W. Cooper Lake State Park (designated as Oak 2) also surrounds a man-made lake. Based on site visits, the Davy Crockett and Sam Houston National Forests were determined to be suitable as pine forest sites. Both forests include a mixture of pine trees including Loblolly Pine, Shortleaf Pine and Longleaf Pine. However, Loblolly Pine is the dominant species. Sam Houston National Forest is in New Waverly, Texas. The monitoring site was located at 301330 N, 951390 W. The monitoring sites in the Sam Houston National Forest (designated as Pine 1) are in the Stubblefield Recreation Area, located adjacent to an oxbow lake of the San Jacinto River. The soil was loose and sandy, with organic matter primarily in the top 5 cm, due to decomposition of leaf litter. Davy Crockett National Forest is located near the city of Crockett, Texas. The monitoring site is located at 311240 N, 951100 W. The monitoring sites in the Davy Crockett National Forest (designated as Pine 2) are in the Ratcliff Lake Recreation Area. The vegetation and soil are similar to those found in the Sam Houston forest. 2.7. Ammonia emissions measurement The ammonia emission measurement system consisted of three components: an ammonia analyzer (Thermo Environmental Instruments (TEI), Model 17C), an emission isolation flux chamber, and a zero air generator (Advanced Pollution Instrumentation, Model 701). A schematic diagram of the ammonia emission collection and analysis system is shown in Fig. 1 (Adapted from Ecklund, 1992). The air generator supplied clean dry air to the flux chamber. The flux chamber had a cylindrical, stainless steel body and an acrylic dome cover. During use, the bottom of the body wall was inserted approximately 3 in into the soil, sealing the soil inside the chamber from the surrounding environment. The sweep zero air was introduced into the chamber by a perimeter distribution tube at ground level. The tube had evenly spaced holes which created jets of sweep air directed towards the center of the chamber. The flow rate of the sweep air was kept greater than the flow drawn by the

ARTICLE IN PRESS G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

7142

THERMOCOUPLE

TEMPERATURE READOUT

PRESSURE RELEASE

INLET

PLEXI GLASS TOP

FLOW METER

REAL TIME ANALYZER

ZERO AIR GENERATOR

7*

11*

OUTLET LINE

STAINLESS STEEL

16* CUT AWAY TO SHOW SWEEP AIR INLET LINE AND THE OUTLET LINE Fig. 1. Schematic diagram of ammonia emission collection and analysis system (Adapted from Ecklund, 1992).

analyzer sample pump; the excess air leaving the chamber through an open port at the top of the dome cover. The chamber was thus maintained under slight positive pressure with a continuous outward flow of air through the open port, isolating the chamber from the ambient atmosphere so that the sample feed to the analyzer contained only ammonia emitted from the soil. During monitoring events, the zero air flow rate to the chamber was either 5.0 or 3.0 liters per minute (lpm), depending on the specific test protocol; the sample pump drew 0.8 lpm through the analyzer. A portable canopy with removable sidewalls protected the instruments from rain and prevented overheating of the ammonia analyzer. Ambient temperature, relative humidity, and barometric pressure were measured with a compact weather station. The flux chamber was placed on the soil next to the analyzer system. Teflon tubing (1/4in diameter) was used to provide the zero air feed and connected the sample port to the analyzer. The open port at the top of the chamber dome allowed for air exhaust, pressure equalization, and access for soil temperature measurements. The sample line was heated to prevent condensation of moisture in the

tubing, and also included a filter to prevent particulate matter from entering the analyzer. A rotameter was used to control the flow rate of zero sweep air to the flux chamber. During field operations, the background signal of the TEI analyzer varied daily, and the instrument zero levels were adjusted accordingly to compensate. The system was assembled with the bottom of the flux chamber covered with a sheet of vinyl before it was set on the ground. Thus, only the zeroair sweep gas entered the flux chamber and was drawn into the sample line and analyzer. The resulting response of the entire sample collection and analysis system was used to verify that there was no systematic error or bias in the results, and that the system worked properly. This procedure was also used to verify and adjust the instrument background settings. A period of 60–90 min after system startup was required for the converters in the analyzer to reach operating temperature and for the system to stabilize. Concurrently, the zero air fed into the chamber purged residual ammonia and nitrogen oxides (NOx) from the chamber, sampling lines, and analyzer, ensuring that the system response was due

ARTICLE IN PRESS G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

to a true blank sample. The background test run continued until the indicated nitric oxide (NO) and NH3 levels stabilized and remained relatively steady for 5 min, at which time the system zero level was set. The run continued for a minimum of five additional minutes to ensure that stable zero-level concentration measurements of NO and NH3 were maintained. If necessary, the zero setting was repeated. Once a reliable background adjustment was made, the analyzer and flux chamber system were considered to be ready for actual emission measurements. Primary flux measurements refer to the initial dynamic measurements made when the flux chamber was first placed on the soil at a test site, after background adjustments had been made. Some ambient air, primarily entrained in the surface soil and leaf litter, was trapped in the flux chamber itself as it was placed on the soil; however, the constant, positive flow of sweep zero air through the chamber minimized this effect. The residual ambient air was fully swept from the chamber after the first 25–30 min of the primary flux test run; the emissions after this time were solely from soil and surface organic matter (forest leaf litter). The primary flux measurements continued until a steady reading was reached. To begin primary flux measurements, the flow rate of the sweep air was increased to 5.0 lpm, and the cover was temporarily kept on the bottom of the flux chamber to establish a baseline NH3 concentration reading. After approximately 5 min, the flux chamber was lifted off of the soil just enough to slide the cover out. The chamber was then pressed into the soil, with the outer wall of the chamber inserted to a depth of 2–3 in. During primary flux measurements, hourly readings were made of the ambient temperature, relative humidity, and barometric pressure. Temperature measurements were also made of the sweep air entering the flux chamber, and of the soil at the surface and depths of 5, 10, 15, 20, 25, and 30 cm. In general, the flow of the sweep air was significantly higher than the NH3 emission rate from the soil, and the progressive dilution caused the indicated NH3 concentration to decay to below detection limit (1 ppb) within 3–4 h for all but one event. In ten of sixteen cases, once the indicated concentration reached a steady level below 0.5 ppb, the sweep air feed was shut off, the analyzer sample line was disconnected, and the sample and vent

7143

ports were capped. The chamber was thus sealed, and the area within was isolated from surrounding ambient air. The chamber was sealed for a minimum of 2 h, during which time the NH3 concentration from soil emissions again increased to a measurable level. The peak ammonia concentration in the sealed (static) chamber was used to determine an ammonia flux during this recovery period. This technique is referred to herein as ‘‘static chamber measurement’’. After the flux chamber was sealed following the primary flux run, the analyzer sample line was connected to the zero air supply for purging. Measurements were then made of the ammonia concentration in the ambient air, with the inlet of the ammonia analyzer sample line placed approximately six feet above the ground. Following the ambient air measurements, the sample lines and analyzer were again purged with zero air. The sample line was then reconnected to the flux chamber, and the zero sweep air flow restarted in order to measure the ammonia that had accumulated while the flux chamber was sealed. The initial concentration measurement was taken as recuperation of ammonia levels due to emissions from the soil. As with the primary runs, the secondary flux measurements continued until a stable (typically zero) ammonia concentration measurement was reached. Several measures were taken to ensure the quality and accuracy of the emissions data. These covered the physical aspects of the collection system, calibration of the analyzer, and sampling operations. Because ammonia is a ‘‘sticky’’ compound that tends to adsorb to surfaces, the flux chamber was made of electro-polished stainless steel and the sample lines were lined with PFA Teflon (the least porous type) in order to minimize this effect. In addition, the sweep zero air was introduced into the flux chamber radially inward from the wall so that the main air flow direction was toward the sampling and exit ports at the top of the chamber near the centerline. The sweep air flow rate gave a complete change of air in the chamber every 6 min, yielding a chamber response time to a step change in concentration of 25–30 min. The sample line from the chamber to the analyzer was relatively short, approximately 2 m. At the 0.8 lpm sampling rate, the residence time of air in the line was slightly less than 5 s, giving a 20–25 s response time. A 65 W continuous electrical heating strip kept the sample line temperature elevated to prevent condensation

ARTICLE IN PRESS 7144

G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

of moisture and to minimize ammonia sorption on the tubing walls. The TEI ammonia analyzer was calibrated using a TEI Model 146 Dynamic Gas Calibrator, with supply gases of 20 parts per million by volume (ppm) NO and 10 ppm NH3, diluted with zero air. Part of the NO was converted in the calibrator to NO2 with internally generated ozone, controlled with gas-phase titration. The analyzer was calibrated using inlet concentrations of 100 ppb for NO, NO2, and NH3, matching the expected peak forest emission concentration. The analyzer works by passing the sample stream through various combinations of high-temperature converters and scrubbers to determine the concentrations of NO, NOx, and total Nx in the sample. The ammonia concentration is calculated from the difference between the total Nx and NOx measurements. The calibration of the analyzer was checked in the laboratory periodically between and after forest measurement activities; no significant changes were found. Extensive calibration tests of the sampling and analysis system (including the flux chamber) showed that the time to reach equilibrium for a high ammonia concentration, e.g., 100 ppb sample feed, was approximately 150–175 min. The decay time of the indicated ammonia level after switching back to zero air was shorter, typically 90–120 min. Actual peak ammonia concentrations measured in the field were less than 5 ppb, with 3–4 ppb peak levels typical. Based on preliminary calibration tests it is reasonable to assume an upper bound of 25–30 min for the system to reach equilibrium with a sample concentration of 10 ppb. The sample times for both primary and secondary flux periods were 3–5 h, well beyond the time necessary for any transient sorption/desorption effects to be neutralized. In addition, the emission rates were calculated using the integrated measurements over the entire sampling period, which compensates for transient effects. These characteristics, combined with the practice to continue a flux measurement period until the indicated ammonia concentration was at the limit of detection (0.5 ppb or less) ensure that any ammonia which may have adsorbed to the chamber or sampling lines was removed by the sweep zero air as the test progressed. This effective purging of the system further ensured that any ammonia that may have adsorbed to the flux chamber walls during the period when the chamber was sealed was then desorbed and collected by the analyzer during the secondary flux measurement periods.

2.8. Soil samples Following flux measurements, soil samples were collected from within the flux chamber, a second site immediately adjacent to the chamber, and at a third site approximately 20–30 feet away from the chamber. The samples from inside and outside of the chamber were compared to determine the effect of the flux measurement procedures on moisture, ammonia, and other organic compounds present in the soil. The third sample was used to determine local variability in soil properties. Samples were taken of the surface leaf litter and of the soil at depths of 5, 10, and 15 cm. Leaf litter and soil samples were collected in 475 and 120 ml borosilicate glass jars with Teflon seals, respectively. The sample jars were then enclosed in two layers of zipper-seal polyethylene bags and placed in a cooler with ice for transport back to the laboratory. The samples remained refrigerated at 4 1C until analysis. The soil and leaf litter samples were analyzed for ammonia and nitrogen using EPA Methods 350.1 and 350.2 (EPA, 1983). The pH and moisture content of the soil and leaf samples were determined based on EPA Methods 150.1 and 160.3, respectively. In all cases, representative sub-samples of the soil or leaf litter stored in the sampling jars were selected for ammonia, pH and moisture content analyses. 2.9. Calculation of emission factor The objective of flux chamber measurements was to determine gaseous ammonia emissions per unit area of forest floor. For primary flux chamber measurements, emissions flux was determined using Eq. (2): Ef ¼

QðC out
C in Þ , A

(2)

where Ef is the ammonia flux (mg m
2 min
1), Q is the volumetric flow rate of air through the chamber (m3 min
1), Cout is the concentration of ammonia exiting chamber (mg m
3), Cin is the concentration of ammonia entering with chamber inlet air (mg m
3), and A is surface area over which flux chamber is placed (0.13 m2). For this study, Q was 0.005 m3 min
1, the volume of the flux chamber was 0.025 m3. Since a zero air supply-conditioning unit was employed, Cin was always zero, a fact that was confirmed for each sampling event.

ARTICLE IN PRESS G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

The manufacturer-specified detection limit for the ammonia analyzer used in this study is 1 ppb, which corresponds to an ammonia concentration of approximately 0.7 mg m
3 at 20 1C and 1 atmosphere. At 0.7 mg m
3, the minimum quantifiable emission factor is approximately 14 kg km
2 yr
1. In all cases, measurements were made until stable concentrations were achieved. Any stable, non-zero concentration less than 0.7 mg m
3 was taken to be less than the manufacturer-specified detection limit, and corresponding emission factors were reported aso14 kg km
2 yr
1 (o1.2 kg km
2 mos
1). The ammonia emission rate per unit area of forest floor during static chamber measurement was estimated using Eq. (3): Ef ¼

m CV ¼ , DtA DtA

(3)

where, m is mass accumulated in head space over time Dt (mg), Dt is time from initial sealing of chamber to ammonia measurement (min), A is surface area of forest floor covered by chamber (m2), C is the ammonia concentration measured in chamber air after Dt (mg m
3), and V is chamber head space volume (m3). For most static chamber experiments, the value of Dt was on the order of 150 min or more. Using this value and a minimum quantifiable concentration of 0.7 mg m
3, Eq. (3) leads to an approximate lowest quantifiable static chamber flux rate of 0.46 kg km
2 yr
1. In fact, for static chamber measurements, Ef was greater than 0.46 kg km
2 yr
1, but less than the primary flux chamber measurement quantifiable limit of 14 kg km
2 yr
1.

3. Results and discussion 3.1. Emission factors A summary of emission factors resulting from summertime measurements in both pine and oak forests of east Texas is presented in Table 2. Results associated with both dynamic and static chamber experiments are presented in Table 2. Dynamic experiments were attempted at each site. Emission factors for 15 of the 16 dynamic experiments were less than 14 kg NH3 km
2 yr
1 (1.2 kg NH3 km
2 month
1). The emission factor for the remaining dynamic experiment was 14 kg NH3 km
2 yr
1 (1.2 kg NH3 km
2 month
1). It is important to recognize that all field experiments were completed during the summer months of 2001, and some

7145

Table 2 Summary of measured ammonia emission factors (Summer Months) Forest name

Forest type Emission factor (kg km
2 month
1) Dynamic

Static

o 1.2 1.2 o 1.2 o 1.2 0.3–1.2

— — 0.07 0.11 0.09

o o o o o

1.2 1.2 1.2 1.2 1.2

0.05 0.10 0.12 0.07 0.08

Oak

o o o o o

1.2 1.2 1.2 1.2 1.2

0.12 0.12 0.17 — 0.14

Cooper Lake Oak

o o o o o

1.2 1.2 1.2 1.2 1.2

— — 0.08 — 0.08

Sam Houston Pine

Average Davy crockett Pine

Average Purtis creek

Average

Average

temporal variations in ammonia emissions are possible. Although static chamber monitoring was not employed at every site, it did lead to consistent results among those sites where it was employed. Emission factors derived from static chamber measurements for pine forests ranged between 0.05–0.12 kg NH3 km
2 month
1 and in oak forests between 0.08–0.17 kg NH3 km
2 month
1. Arithmetic mean emission factors based on the use of the static chamber method at Sam Houston and Davy Crockett National Forests were 0.09 kg NH3 km
2 month
1 and 0.08 kg NH3 km
2 month
1, respectively. The arithmetic mean emission factor over four static chamber experiments in oak forests was 0.13 kg NH3 km
2 month
1. The reader is cautioned that static flux measurements are not generally employed or described in the published literature. Nevertheless, due to the limitations associated with dynamic chamber experiments, we opted to use static chamber results as being representative of summertime emission factors for pine and oak forests in east Texas.

ARTICLE IN PRESS 7146

G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

Before comparing the emission factors derived from this study with previous measurements or estimates, it is instructive to review the state of knowledge related to ammonia emissions from undisturbed soil, particularly as related to forest litter and soil. Of particular significance are the conflicting arguments made about ammonia emissions from undisturbed soil and from forests, and the sparseness of measured emissions. Kim (1973) was the first to report measurements of ammonia emissions from forest floors. Our work is the second study focusing on the measurements of ammonia emissions from forest floors. Most other investigators either measured or modeled ammonia concentrations above the forest canopy. Using the results of a simple ammonia emissions model for undisturbed non-fertilized soil, Dawson (1977) argued that undisturbed land is likely the primary source of global ammonia emissions. However, Dawson (1977) acknowledged that, as of 1977, emission from uncultivated and unfertilized vegetated land had not been measured. Fifteen years later, Langford et al. (1992) noted that gaseous ammonia fluxes in unmodified forests were virtually non-existent. Schlesinger and Hartley (1992) indicated that little is known about the volatile loss of ammonia from non-agricultural soils in which ammonium (NH+ 4 ) is derived from mineralization of organic nitrogen. Asman et al. (1998) observed that emissions from B-Napus canopies following the deposition of leaf litter appear to be ‘‘significant’’. They recommended that decomposition of leaf litter as a ground source of ammonia emissions needs further investigation. However, Pryor et al. (2001) suggested that soil conditions, particularly surface soil pHo6.5, precludes a large efflux of ammonia emissions from forest floors. The emission factors determined in this study are significantly lower than any measured previously for pine and oak forests. However, it is important to note that the forests used in this study were not artificially fertilized, while several of those that formed the basis for previous studies had been amended with urea-nitrogen (Camire and Bernier, 1981; Marshall and DeBell, 1980; Overrein, 1968). The remaining two emission factors (Kim, 1973; Langford and Fehsenfeld, 1992) for pine/coniferous forests differ by a factor of 500, suggesting the difficulties and potential errors associated with selecting a single emission factor for estimating ammonia emissions from forests.

The emission factors reported by Kim (1973) for pine and oak forests in South Korea are approximately four orders of magnitude greater than the emission factors determined in this study. The emission factor reported by Langford and Fehsenfeld (1992) for coniferous forests is 32 times greater than the summertime pine forest emission factor determined for this study. A small number of researchers have predicted ammonia fluxes from forest soils based on mathematical models that differ significantly in complexity. Bouwman et al. (1997) developed a simple model to estimate global emissions of ammonia and estimated 0.03 g m
2 yr
1 (2.5 kg km
2 month
1) ammonia emissions from temperate forests, a value approximately 20–30 times greater than those measured in this study for east Texas. However, the Bouwman et al. model did not account for soil pH effects and may therefore underestimate emissions from alkaline soils, while potentially over-estimating emissions from acidic soils, e.g., soils in east Texas. Dawson (1977) developed a more sophisticated model than Bouwman et al. (1997) to estimate global emissions of ammonia from un-disturbed land. His model accounted for pH effects on ammonia/ammonium equilibrium partitioning in soil and considered the degree of exchangeable ammonium resulting from a balance of litter decomposition and nitrification. Resulting emissions from un-disturbed soil within 30–401N latitude were predicted to be 9.2 billion kg
1 yr
1 over an area of 15.57 million km2. This leads to an emission factor of 49 kg km
2 month
1, approximately 500 times greater than forest floor emission factors derived in this study for east Texas. Dawson (1977) did not separate forest emissions from other land types. Langford et al. (1992) used the Dawson (1977) model to estimate ammonia emissions from forest floors using soil-specific properties (surface pH and ammonium concentration) from three forests in the United States. For a montane coniferous forest in
1 Colorado (pH ¼ 5.22; NH+ 4 ¼ 17.3 g kg ) they estimated an emission factor (normalized here to a monthly average) of 0.25 kg km
2 month
1. For a temperate coniferous forest in Alabama (pH ¼ 5.3;
1 NH+ 4 ¼ 5.6 g kg ) they estimated an emission factor of 0.10 kg km
2 month
1. For a temperate deciduous forest in Tennessee (pH ¼ 4.7;
1 NH+ 4 ¼ 17.9 g kg ) they estimated an emission factor of 0.05 kg km
2 month
1. These emission factors in US forests bound the values that were

ARTICLE IN PRESS G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

derived based on our measurements in east Texas. While the assumed soil surface temperature of 20 1C was less than those observed in this study, the soil pH and ammonium concentration levels were reasonably consistent with those measured in east Texas. Potter et al. (2001) employed the NASA–CASA model within a GIS framework and completed a rigorous modeling evaluation of ammonia emissions from native soils in California. The model accounts for area water balances, soil pH and moisture content, litter fall, nutrient allocation, soil nitrogen mineralization, seasonal carbon fixation, surface temperature, and soil ammonia emissions. While recent ammonia flux estimates for California are greater than those measured in east Texas, Potter et al. (2001) predicted significant seasonal variations in ammonia emissions, with 20-fold or greater differences in predicted emissions between different months. Soil moisture was predicted to have a significant influence on ammonia emissions, with much lower emissions from moist soils owing to lower gas diffusivities. Soil pH was also predicted to have a significant effect on ammonia emissions, with the highest emissions predicted to occur for conditions of high pH and low moisture content. In several California counties that are dominated by evergreen forests the soil pH ranged from 5.5 to 6.19. For comparison, six of the eight pine forest locations that we studied in east Texas had soil surface (5 cm depth) pH of less than or equal to 5.5, and 7 of 8 oak forest locations had soil surface pH less than 5.3. Using the same algorithm that was employed in the NASA–CASA model, we estimated a 20-fold reduction in ammonia flux for a pH drop from 6 to 5 at a surface temperature of 30 1C, everything else being equal. Based on a version A (moderate pH effects) model application, Potter et al. (2001) estimated annual ammonia emissions from evergreen needleleaf forests in California to be 810 mtpy over a total area of 12.4 million hectares. This translates to an average emissions flux of 0.54 kg km
2 month
1, six times greater than the summertime pine forest emission factor determined in this study for east Texas. Similarly, the average emission flux for deciduous broadleaf forests was estimated to be 2.6 kg km
2 month
1, 20 times greater than the summertime oak forest emission factor determined in this study for east Texas. The relatively small differences in emission factors between sites in this study preclude a

7147

rigorous evaluation of the effects of soil/liter properties on ammonia emissions from forests in east Texas. However, the relatively low emissions factors are consistent with low values of soil pH.

3.2. Soil properties and their effects The mean soil temperature, pH, moisture content and ammonia concentration observed at each of the forest monitoring sites are summarized in Table 3. The highest moisture levels were found within the leaf litter and lower, but relatively constant, moisture contents were observed within 0–15 cm soil depths. The soil at Davy Crockett National Forest had the highest moisture content. In contrast, the soils at the remaining sites were relatively dry, with mean soil moisture contents ranging from 2.8% to 7.5%. In general, soils with moderate moisture content are expected to release the greatest fraction of ammonia. If the soil is too dry, microbial activity may be inhibited and the soil can adsorb ammonia directly; if the soil is too wet, it inhibits the diffusion of ammonia to the soil surface. The NH3–N concentration measured in the soil and leaf litter samples varied from 0.4 to 47.3 mg kg
1. As expected, the average ammonia concentration in the leaf litter across all the forests (i.e., 18 mg kg
1) was greater than the concentration in the underlying soil (e.g., 5.3 mg kg
1 for the 0–5 cm soil horizon). Results from these analyses indicate that, in general, the ammonia concentrations in the soil continued to decline with depth from the surface. These variations with depth support the common assumption that the surface litter and near-surface soils are the most important regions to consider when quantifying ammonia emissions. As expected, soil temperatures were greatest at the surface and declined with depth. Since flux experiments were conducted over different periods of the day and night, the associated soil temperatures observed during the experiments follow diurnal variations. The minimum soil temperature observed was 22.7 1C and the maximum observed on a hot afternoon when the ambient temperature was nearly 36 1C was 30.9 1C. In general, high temperatures are expected to increase ammonia volatilization; however, the low pH of the soils present in Texas forests seems to be an overriding factor that ultimately suppresses ammonia emissions.

ARTICLE IN PRESS G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

7148

Table 3 Mean soil properties at forest monitoring sites Site

Sam Houston

Temperature (1C) Soil: surface Soil: 5 cm down Soil: 10 cm down Soil: 15 cm down

Davy Crockett

Purtis Creek

Cooper lake

Range observed Min

Max

26.2 25.1 24.8 24.6

25.3 23.8 24.4 24.1

27.7 25.5 25.1 24.4

28.9 26.2 25.9 25.7

23.7 22.7 23.3 23.1

30.9 27.1 26.6 26.1

litter 5 cm down 10 cm down 15 cm down

5.3 5.5 5.7 5.6

5.3 5.2 5.4 5.4

5.3 5.1 5.6 5.6

5.9 4.8

4.2 4.3

6.3 6.1

Moisture Content (%)c Leaf litter Soil: 5 cm down Soil: 10 cm down Soil: 15 cm down

5.0 6.8 6.7 6.5

27.9 12.5 10.9 11.4

20.7 10.3 5.5 2.8

pHa Leaf Soil: Soil: Soil:

NH3–N (mg kg
1)a Leaf litter Soil: 5 cm down Soil: 10 cm down Soil: 15 cm down

b

b

b

b

b

b

22.8 2.8 2.8 3.1

14.2 7.5

11.6 1.8 1.4 2.2

36.3 20.0 17.3 14.8

11.5 3.8 5.4

27.7 1.8

10.5 5.3

b

b

b

b

b

0.7 0.4 5.4 2.8

47.3 23.1 5.5 2.8

b b

a

Soils inside flux chamber. Sample not collected. c Soils immediately adjacent to flux chamber. b

3.3. Ambient ammonia concentrations in forest canopy The mean of in-canopy ambient ammonia concentration measured at each site varied over a relatively narrow range of 1.5–2.9 ppb, reasonably consistent with a value of 1.7 ppb reported by Pryor et al. (2001) measured above deciduous forests in southern Indiana during the fall. However, the arithmetic mean value measured in east Texas is a factor of 4–6 greater than springtime measurements over the same Indiana forests and mean summertime (day) surface concentrations in forests as reported by Langford et al. (1992). 3.4. Omission of canopy effects Several recent studies have indicated that canopy vegetation (leaves and needles) can significantly affect ammonia emissions from forests due to two phenomena: re-absorption of ammonia and vegetative emissions of ammonia. The net effects of forest canopies on ammonia emissions appear to be

dynamic, with canopies sometimes serving as a net sink and sometimes serving as a net source of ammonia. Andersen et al. (1993) measured ammonia fluxes above spruce forests in Denmark. They observed the canopy to be a net source of ammonia emissions during 10 of 34 experiments, and a net sink during 24 experiments. Wyers and Erisman (1998) studied the exchange of ammonia over coniferous forests in the Netherlands for over two years. They observed vertical ammonia fluxes to vary in direction (to and from the top of the canopy). They discussed stomatal release, drying of leaf surfaces, and ammonium aerosol evaporation as potential ammonia sources within forests. Pryor et al. (2001) also measured ammonia fluxes over a deciduous broadleaf forest in southern Indiana. They observed that, on average, the forest canopy was a sink of ammonia. They did observe a reverse flux (canopy as net source of emissions) on some spring days, with a magnitude of up to 0.2 mg m
2 h
1. They attributed ammonia sources to stomatal release, leaf drying, and evaporation of ammonium nitrate particles.

ARTICLE IN PRESS G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

In none of the aforementioned studies did the researchers actually measure ammonia emissions from forest litter and soil. It is also important to recognize that fluxes to and from the forest canopies were all made above the canopy. Thus, canopy ‘‘sink effects’’ all relate to uptake from the overlying ambient atmosphere as opposed to the underlying terrestrial environment. Measurements of ammonia concentrations within the forest canopy, including above the soil surface, were not found in the published literature, nor was any discussion of the possibility of horizontal transport of ammonia in the air space located above forest soil but below the main portion of the vegetative canopy. Forest canopies may have a significant influence on net ammonia emissions from forests. However, the current knowledge base related to canopy emissions and uptake do not allow for accurate predictions of these phenomena. Given the relatively low emission fluxes measured for pine and oak forests in Texas, canopy uptake would not have a significant effect on overall non-industrial source ammonia emissions in Texas. However, the impact of canopy emissions cannot be ascertained based on the results of this study. 3.5. Statewide ammonia emissions estimates for nonindustrial sources Updated statewide non-industrial source emissions estimates for ammonia are listed in Table 4 for each primary source category and source subcategory that exceeds 1000 mtpy. The predicted statewide total ammonia emissions for the year 1996 were 467,000 mtpy. The dominant source was predicted to be animal husbandry, which was estimated to contribute 77% of total ammonia emissions. Within this category, generic cows, beef cows, and horses alone were estimated to contribute 65% of non-point source ammonia emissions in Texas. As described earlier in this paper, the reader is cautioned that ammonia emissions estimates for livestock, and especially cows, are subject to significant uncertainty due to their development in Europe where the diet of livestock can differ considerably from those in the United States. Fertilizer applications were estimated to account for almost 8% of non-point source ammonia emissions in Texas, and were distributed between six different types of fertilizer. Natural soil/vegetation was estimated to be responsible for almost 7% of total non-point ammonia emissions in Texas, a

7149

value that is close to that predicted by Dentener and Crutzen (1994) for the natural soil and vegetation contribution to global ammonia emissions. Within this category, scrublands and grasslands were estimated to contribute about 5% of non-point source ammonia emissions. Domestic sources were estimated to contribute about 6% of non-point source ammonia emissions in Texas, with emissions from dog and cat urine contributing nearly 2/3 of those emissions. On-road vehicles were estimated to contribute only 2% of non-point source ammonia emissions, with nearly all of these emissions originating from light-duty gasoline engines with three-way catalysts. Municipal wastewater was estimated to contribute less than 1% of non-point source ammonia emissions in Texas. We utilized emissions factors for pine and oak forests from a 1972 study in South Korea (Kim, 1973) in our previous estimates of annual ammonia emissions in Texas (Corsi et al., 2000b). The impact of replacing Kim’s emission factors with those measured during this study is enormous. Predicted statewide non-point source ammonia emissions are reduced by almost a factor of two, from 921,000 mtpy to 467,000 mtpy. Predicted statewide ammonia emissions from pine forests, located primarily in east Texas, are reduced from 257,000 mtpy to only 16 mtpy. Similarly, predicted statewide ammonia emissions from oak forests are reduced from 197,000 mtpy to 22 mtpy. Revised ammonia emissions estimates for oak and pine forests in Texas are small and not included in Table 4. Reductions in predicted ammonia emissions from pine and oak forests have a significant effect on the overall contribution of natural soil and vegetation and animal husbandry to statewide non-point source ammonia emissions. The natural soil and vegetation source category drops from 52% to just 3% of statewide non-point source emissions. The overall contribution of animal husbandry to nonpoint source ammonia emissions increases by a factor two, from 39% to 77%. There are several reasons why the previous estimates for natural soil/vegetation are suspect. First, only one emission factor was obtained for pine and oak forests from the published literature. These factors were based on research involving a limited amount of experimental data that were collected in 1972. Significant advances in analytical instrumentation have occurred in the past three decades. The emission factors were collected in pine and oak forests in South Korea, and it is not clear

ARTICLE IN PRESS 7150

G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

Table 4 Sources emitting greater than 1000 metric tons yr
1 Source category

Emission factor (NH3 unit
1)

References

Emissions (thousands of metric tons yr
1)

Animal husbandry Generic cows Beef cows Horses Broilers Milk cows Pullets (laying age) Goats Turkeys Sheep Hogs & pigs

kg head
1 yr
1 22 15 25 0.23 33 0.44 2.2 0.89 2.7 5.6

1–6, 9 7–8 1, 3, 4, 6–11, 16 4, 7, 11, 12 3–4,6–8,10–11,13–14 4, 7, 10–12 3, 4, 6, 15 4, 7, 8, 10, 11 1, 2–10, 13, 16 1, 4–10, 13, 17

361 193 81 31 18 13 9 5 4 4 3

Fertilizer application N-P-K Anhydrous ammonia Other nitrogen solutions Urea Ammonium sulfate Ammonium nitrate

kg metric ton
1 of N applied 49 49 30 121 140 24

4 6 6 3, 9, 13 4, 9 4

37 12 10 6 5 3 1

Natural Soil/Vegetation scrublands grasslands other temperate forests Urban land area

kg m
2 yr
1 0.0001 0.00004 0.0004 0.0004

6, 18 6, 18 6, 15, 18 15

31 12 10 6 2

Domestic sources Dogs Humans Cats

kg/head/yr 2.18 0.44 0.69

7,8,10,13,15 3,5,7,8,10,13,16,19 7,8,10,13,15

27 12 8 5

On-road vehicles Light-duty gasoline engines w/ 3-way catalysts

kg vehicle mile traveled
1

Municipal wastewater POTWs

10

0.0001

20

10

14

6 6


1

kg million gallons wastewater 8.6

Statewide total

467

Notes: For some source categories only one reference is provided; emission factor from this reference was directly used in this study for these categories. Where multiple references are provided, an average emission factor was calculated from the listed references and was used in this study. (1—ApSimon et al., 1987; 2—Kruse et al., 1986; 3—Dianwu and Anpu, 1994; 4—Asman, 1992; 5—Battye et al., 1994; 6— Bouwman et al., 1997; 7—Lee and Dollard, 1994; 8—Heisler et al., 1988; 9—Buijsman et al., 1987; 10—Dickson et al., 1991; 11—Sadeghi and Dickson, 1992; 12—Roe and Strait, 1998; 13—Sutton et al., 1995; 14—Warn et al., 1990; 15—Gharib and Cass, 1984; 16—Moller and Schieferdecker, 1989; 17—Kruse et al., 1989; 18—Schlesinger and Hartley, 1992; 19—Atkins and Lee, 1993; 20—Harvey et al., 1983).

from the original source as to whether the soil or canopy conditions in the forests that were sampled were consistent with those that are observed in the forests of Texas. 3.6. Rural versus urban counties As described above the dominant non-industrial source of ammonia emissions is predicted to be

associated with animal husbandry. This is particularly true in rural counties, many of which are characterized by an animal husbandry contribution that exceeds 85%. We also predicted the relative contributions of various source categories to nonindustrial ammonia emission in urbanized counties in Texas. Results are listed in Table 5. Bexar County contains the greater San Antonio area, Dallas County the City of Dallas, El Paso County the City

ARTICLE IN PRESS G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

7151

Table 5 Contribution to non-industrial ammonia emissions in urbanized counties County

NH3 Emissions (metric tons yr
1)

Domestic (%)

Animal husbandry (%)

On road vehicles (%)

Wastewater (%)

Fertilizer (%)

Bexar Dallas El Paso Harris Tarrant Travis

4800 7100 2800 10,900 4500 2600

39 41 35 44 45 34

37 27 41 26 29 38

14 18 9 15 17 15

9 14 7 10 7 10

2 1 9 5 3 3

of El Paso, Harris County the City of Houston, Tarrant County the City of Ft. Worth, and Travis County the City of Austin. As expected, the results for urbanized counties in Texas are considerably different than those for rural counties. Contributions due to animal husbandry are still relevant, but approximately 1/2 to 1/3 of the contribution in most rural counties. Contributions due to fertilizer usage are relatively small, except for El Paso County, which has a significant base of agricultural activity along the Rio Grande River. On-road vehicles are predicted to emit slightly more ammonia than publicly owned treatment works (wastewater). Domestic sources exceed those of animal husbandry in four of the six urban counties listed in Table 5, and exceed animal husbandry when emissions from these counties are summed. As listed in Table 1, domestic sources are individually small but extremely numerous in urban areas, e.g., domestic cats. Interestingly, we predict the major contributors to domestic sources to be humans (perspiration and respiration), and dogs and cats (primarily urine). In fact, dogs and cats are predicted to contribute nearly 2/3 of domestic non-industrial ammonia emissions in urban counties, on the order of, or greater than, emissions from on-road vehicles and wastewater combined. For example, in urban areas cats and dogs are born through uncontrolled reproduction at rates estimated as high as seven times the birth rate of Americans, but the numbers are characterized by a high degree of uncertainty and likely vary considerably between cities based on the effectiveness of spray/neuter programs. This surprising result indicates a need for additional studies to reduce uncertainties in emissions estimates for this source category.

4. Conclusions A summer field study was conducted to determine ammonia fluxes from pine and oak forest floors in east Texas. Both dynamic and static chamber methods were employed. Dynamic chamber experiments proved to be too insensitive to determine ammonia fluxes. Thus, static chamber results were used to estimate ammonia fluxes from forest litter and soil. The ammonia flux averaged 0.09 kg km
2 month
1 for pine forests and 0.13 kg km
2 month
1 for oak forests during the summertime monitoring period. These values are significantly lower than those previously measured and cited in the published literature, and are approximately four orders of magnitude less than emission factors that were employed for pine and oak forests in a previous non-industrial source ammonia emissions inventory for Texas. Low ammonia emissions from pine and oak forest floors in east Texas are likely influenced greatly by the acidic nature of forest litter and surface soils. An estimate of non-point source ammonia emissions in Texas was developed. Animal husbandry is predicted to be the largest non-point source of ammonia emissions in Texas, with cattle predicted to emit greater than 280,000 metric tons of ammonia in 1996. Fertilizer application is the second largest non-point source of ammonia emissions. Temporal variations were not considered in this study and emissions from fertilizer are likely to be much greater than annual average emissions during specific months. Acknowledgments This study was funded by the Texas Commission on Environmental Quality. The authors wish to

ARTICLE IN PRESS 7152

G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153

acknowledge Mr. Steve Anderson for his guidance. The authors acknowledge contributions from the following students of The University of Texas at Austin: Stacey Fredenberg, Marie Dondelle, Katherine Dombrowski, Satoshi Takahama, and Widianto. The authors also acknowledge the Texas Parks and Wildlife Department, and the United States Department of Agriculture for their approval of setting up the instruments and for taking measurements in the parks. References Andersen, H.V., Hovman, M.F., Hummelshoj, P., Jensen, N.O., 1993. Measurement of the ammonia flux to a spruce stand in Denmark. Atmospheric Environment 27A, 189–210. ApSimon,, H.M., Kruse, M., Bell, J.N., 1987. Ammonia emissions and their role in acid deposition. Atmospheric Environment 21, 1939–1946. Asman, W.A.H., 1992. Ammonia Emissions in Europe; Updated Emissions and Emission Variations, Report to the National Institute of Public Health and Environmental Protection, report No. 228471008, the Netherlands. Asman, W.H., Sutton, M.A., Schjorring, J.K., 1998. Ammonia: emission, atmospheric transport and deposition. New Phytologist 139, 27–48. Atkins, D.H.F., Lee, D.S., 1993. Indoor concentrations of ammonia and the potential contribution of humans to atmospheric budgets. Atmospheric Environment 27, 1–7. Battye, R., Battye, W., Overcash, C., Fudge, S., 1994. Development and Selection of Ammonia Emission Factors. Report to the United States Environmental Protection Agency, EPA600/R-94/190 (1994). Bouwman, A.F., Lee, D.S., Asman, W.A.H., Dentener, F.J., Van Der Hoek, K.W., Olivier, J.G.J., 1997. A global highresolution emission inventory for ammonia. Global Biochemical Cycles 11 (4), 561–587. Buijsman, E., Maas, H.F.M., Asman, W.A.H., 1987. Anthropogenic NH3 emission in Europe. Atmospheric Environment 21, 1009–1022. Camire, C., Bernier, B., 1981. Retention de l’azote et evolution des propprietes d’un humus brut de station de pin gris apres application d’engrais azotes. Canadian Journal of Forest Research 11, 51–61. Corsi, R.L., Torres, V.M., Carter, G., Dombowski, K., Dondelle, M., Fredenberg, S., Takahama, S., Taylor, T., 2000a. Nonpoint source ammonia emissions in Texas: a first estimate. Report to the Texas Natural Resource Conservation Commission. Corsi, R.L., Torres, V.M., Fredensberg, S., Dondelle, M., Dombrowski, K., Takahama, S., Taylor, T., Anderson, S., 2000b. A screening analysis of non-point source ammonia emissions in Texas. In: Proceedings of the 93rd Annual AWMA Conference, Salt Lake City, Utah. Dawson, G.A., 1977. Atmospheric ammonia from undisturbed land. Journal of Geophysical Research 82 (21), 3125–3133. Dentener, F.J., Crutzen, P.J., 1994. A three-dimensional model of the global ammonia cycle. Journal of Atmospheric Chemistry 19, 331–369.

Dianwu, Z., Anpu, W., 1994. Estimation of anthropogenic ammonia emissions in Asia. Atmospheric Environment 28, 689–694. Dickson, R.J., Cass, G.R., Markovich, L.J., Oliver, W.R., Wright, D.A., Cioni, C.A., Berard, L.A., Wilkinson, J.G., Riswold, J.R., 1991. Development of the Ammonia Emission Inventory for the Southern California Air Quality Study, for the Electric Power Research Institute (EPRI). Radian Corporation, Sacramento, CA. Ecklund, B., 1992. Practical guidance for flux chamber measurements of fugitive volatile organic emission rates. Journal of the Air and Waste Management Association 42 (12), 1583–1591. EPA, 1983. Methods for Chemical Analysis of Water and Wastes. Environmental Protection Agency, Office of Research and Development, Washington, DC EPA/600/4-79/ 020, March 1983. Gharib, S., Cass, G.R., 1984. Ammonia Emissions in the South Coast Air Basin—1982, Open file report 84-2, Environmental Quality Laboratory, California Institute of Technology. Harvey, C.A., Garbe, R.J., Baines, T.M., Somers, J.H., Hellman, K.H., Carey, P.M., 1983. A study of the potential impact of some unregulated motor vehicle emissions, Society of Automotive Engineers, Technical Paper No 830987, 1983. Heisler, S.L., Collins, H., Collins, J., Hayden, P., Lurmann, F., Yeransian, T., Young, J., 1988. Interim Emissions Inventory for Regional Air Quality Studies, EA-6070. Electric Power Research Institute, Palo Alto, CA. Kim, C.M., 1973. Influence of vegetation types on the intensity of ammonia and nitrogen dioxide liberation from soil. Soil Biology Biochemistry 5, 163–166. Klaassen, G., 1991. Past and Future Emissions of Ammonia in Europe, Report SR-91-01. IIASA, Austria. Kruse, M., ApSimon, H.M., Bell, J.N., 1986. An Emission Inventory for Ammonia Arising from Agriculture in Great Britain, Imperial College Centre for Environmental Technology, University of London. Kruse, M., ApSimon, H.M., Bell, J.N., 1989. Validity and uncertainty in the calculation of an emission inventory for ammonia arising from agriculture in Great Britain. Environmental Pollution 56, 1989. Langford, A.O., Fehsenfeld, F.C., 1992. Natural vegetation as a source or sink for atmospheric ammonia: a case study. Science 255, 581–583. Langford, A.O., Fehsenfeld, F.C., Zachariassen, J., Schimel, D.S., 1992. Gaseous ammonia fluxes and background concentrations in terrestrial ecosystems of the United States. Global Biochemical Cycles 6 (4), 459–483. Lee, D.S., Dollard, G.J., 1994. Uncertainties in current estimates of emissions of ammonia in the United Kingdom. Environmental Pollution 86, 267–277. Marshall, V.G., DeBell, D.S., 1980. Comparison of four methods of measuring volatilization losses of nitrogen following urea fertilization of forest soil. Canadian Journal of Soil Science 60, 549–563. Moller, D., Schieferdecker, H., 1989. Ammonia emission and deposition of NHx in the GDR. Atmospheric Environment 23, 1187–1193. Overrein, L.N., 1968. Lysimeter studies on tracer nitrogen in forest soil: I nitrogen losses from leaching and volatilization after addition of urea-N. Soil Science 106, 280–290.

ARTICLE IN PRESS G. Sarwar et al. / Atmospheric Environment 39 (2005) 7137–7153 Potter, C., Krauter, C., Klooster, S., 2001. Statewide inventory estimates of ammonia emissions from native soils and chemical fertilizers in California. Report to the Emissions Inventory Branch, California Air Resources Board, Contract Number 98-716. Pryor, S.C., Barthelmie, R.J., Sorenson, L.L., Jensen, B., 2001. Ammonia concentrations and fluxes over a forest in the midwestern USA. Atmospheric Environment 35, 5645–5656. Roe, S.M., Strait, R.P., 1998. Methods for Improving National Ammonia Emission Estimates. In:Proceedings of the Air and Waste Management Association Emissions Inventory Conference, San Diego, California. Sadeghi, V.M., Dickson, R.J., 1992. AUSPEX Ammonia Emission Inventory/Data Collection Effort. Radian Corporation. Schlesinger, W.H., Hartley, A.E., 1992. A global budget for atmospheric NH3. Biogeochemistry 15, 191–211. Seinfeld, J.H., Pandis, S.N., 1997. Atmospheric Chemistry and Physics. Wiley, New York. Sutton, M.A., Place, C.J., Eager, M., Fowler, D., Smith, R.I., 1995. Assessment of the magnitude of ammonia emissions in the UK. Atmospheric Environment 29, 1393–1411.

7153

Warn, T.E., Zelmanowitz, S., Saeger, M., 1990. Development and Selection of Ammonia Emission Factors for the 1985 NAPAP Emissions Inventory, EPA-600/7-90-014. US Environmental Protection Agency, Washington DC. Wiedinmyer, C., Allen, D., 1999. Biogenic Land Cover Database for Central Texas. Report to the Texas Natural Resource Conservation Commission, Austin, Texas. Wiedinmyer, C., Strange, W., 1998. Biogenic Land Cover Database for Victoria. Report to the Texas Natural Resource Conservation Commission, Austin, Texas. Wyers, G.P., Erisman, J.W., 1998. Ammonia exchange over coniferous forest. Atmospheric Environment 32 (3), 441–451. Yarwood, G., Lee, K., Allen, D., Quigley, C., Wiedinmyer, C., Strange, W., Guenther, A., Baugh, B., 1997. Leaf Biomass Density Data for North-Central Texas; Final Report. Report to the Texas Natural Resource Conservation Commission. Yarwood, G., Wilson, G., Allen, D., Quigley, C., Strange, W., Wiedinmyer, C., Guenther, A., 1999. Leaf Biomass Density Data for Southeast Texas; Final Report. Report to the Texas Natural Resource Conservation Commission.