Grantee Research Project Results
Final Report: A Modeling Investigation of NHx Cycling in the Troposphere and Its Impact on Particulate Matter and Acidic Substances Budgets
EPA Grant Number: R826773Title: A Modeling Investigation of NHx Cycling in the Troposphere and Its Impact on Particulate Matter and Acidic Substances Budgets
Investigators: Mathur, Rohit , Alapaty, Kiran , Shankar, Uma , Houyoux, Marc , Adelman, Zac
Institution: MCNC / North Carolina Supercomputing Center
EPA Project Officer: Hahn, Intaek
Project Period: October 1, 1998 through September 30, 2001 (Extended to March 31, 2003)
Project Amount: $488,744
RFA: Air Pollution Chemistry and Physics (1998) RFA Text | Recipients Lists
Research Category: Air
Objective:
The fate of nitrogen-containing species in the atmosphere is of considerable interest, given their role in the formation of acidic substances, particulate matter (PM), and tropospheric ozone, and potential eutrophication and nutrient loading effects resulting from their deposition. Although significant attention has been devoted towards studying oxidized nitrogen in the atmosphere, little effort has been devoted towards quantifying the budgets of reduced nitrogen species. The overall objective of this research project was to improve the current understanding of the cycling of reduced nitrogen compounds in the atmosphere, and to investigate the coupling of such compounds with atmospheric aerosols and other criteria pollutants responsible for the acidifying atmospheric load, through the development, enhancement, and continuous evaluation of comprehensive multipollutant regional models.
Summary/Accomplishments (Outputs/Outcomes):
To investigate the processes governing the fate of NHx in a consistent manner and to develop a preliminary assessment of the adequacy of existing NH3 emissions and the physical and chemical process pathways dictating the fate of airborne-reduced nitrogen species, modeling analyses were conducted with two regional models: (1) the Extended-Regional Acid Deposition Model (RADM); and (2) the comprehensive gas-aerosol version of the Multiscale Air Quality Simulation Platform (MAQSIP); MAQSIP also has served as a prototype for the U.S. Environmental Protection Agency's Community Multi-scale Air Quality/Models-3 system.
RADM was enhanced to include detailed treatment of the physical and chemical processes regulating the fate of ammonia emissions, and to model the interactions and chemical and thermodynamic coupling between NOx-SOx-NHx species; the model is termed the Extended-RADM. Extended-RADM model simulations for 30 representative cases, previously used in the National Acid Precipitation Assessment Program (NAPAP) assessment, were aggregated to construct climatological annual and seasonal averages. To account for uncertainties in both the magnitude and seasonal variability of NH3 emissions in the inventory used, regional seasonal correction factors for NH3 emissions were developed through successive model applications and detailed comparisons with measurements of ambient levels and deposition amounts. In these analyses, we included comparisons of model results against measurements of ambient levels of HNO3, SO42-, NO3-, and NH4+, and wet deposition amounts of SO42-, NO3-, and NH4+, but gave more weight towards satisfying three specific criteria:
The adjustment in NH3 emissions should not disturb the modeled ambient and wet deposition for sulfate through perturbations in cloud pH related oxidation pathways, such as through ozone oxidation
Because NH3 provides a pathway for formation of particulate NO3-, the predicted ambient NO3-/(HNO3 + NO3-) ratio should retain the spatial regional signature indicated by measurements.
Any bias structure in modeled NH4+ wet deposition should be similar to that for SO42- and NO3- wet deposition.
This approach provides a physically based, self-consistent constraint for potential growth in NH3 emissions, as opposed to a purely empirically based one, in which NH3 increases are solely dictated by discrepancies between model and observed ambient NHx levels without accounting for other modeled species.
The comparison of these model-predicted values with measurements from regional networks suggests that the model can capture the spatial and seasonal trends in ambient concentrations and wet deposition amounts of various species; the correlations for these model-observed comparisons were favorable for most species examined, with R2 values generally in the range of 0.4-0.7.
These analyses suggest that on an annual basis, the 1985 NAPAP NH3 emissions inventory is a factor of 2.75 low. The estimated emissions correction factors show significant seasonal variation, and yield maximum NH3 emissions during summer, followed by spring, fall, and winter, a trend that is consistent with recent NH3 flux measurements. These calculations also suggest that the range between summer and winter NH3 emissions is a factor of 3-4. For European conditions, Asman (1992) found a factor of 2.2 difference in NH3 emissions between the warm and cool months. The larger differences in seasonal variations reported here are due to larger variations in seasonal temperatures in the United States, and consequently, larger variations in NH3 emissions due to volatilization.
The relative amounts of reduced and oxidized nitrogen in the gas and particle phase also are captured well by the model, and are consistent with available measurements. The ratio of particle to total oxidized nitrogen shows a distinct seasonal cycle with a winter high (about 50 percent) and summer low (10-15 percent).
The spatial trends in the degree of neutralization, defined as the ratio of the equivalent of basic species to the equivalents of acidic species, show that on an annual basis, the eastern United States is, by and large, NH3 limited. Both the model calculations and the measurements show the highest degree of neutralization during winter, with large portions of the eastern United States being characterized by aerosols that are nearly fully neutralized (> 90 percent). However, for the rest of the year, much of the eastern United States is NH3 limited.
The Extended-RADM calculations, which are representative of emissions levels during the late 1980s-early 1990s period, also indicate that on an annual domain-wide basis, reduced nitrogen species contribute 47±8 percent of the total nitrogen wet deposition in the eastern United States. This is consistent with the measured value of 43±9 percent that is representative of this period.
MAQSIP was used to simulate the regional and local distributions of reduced and oxidized nitrogen and related species. Model simulations were performed at three grid resolutions using grid nesting: (1) the eastern United States using a grid resolution of 36 km to examine the regional characteristics; (2) North Carolina and surrounding states to examine local characteristics in vicinity of the high NH3 emissions in eastern North Carolina, using a finer resolution 12-km grid; and (3) a 4-km resolution grid over the high NH3 emissions source region in eastern North Carolina. Meteorological inputs were created using the PSU/NCAR Mesoscale Model (MM5). Emission inputs for these simulations were based on the 1996 National Emissions Inventory. Simulations were conducted for the June 19-30, 1996, period. These were characterized by two separate ozone episodes in the Southeast, and have been used for ozone attainment demonstration studies in NC Enhancements to the representation of various physical and chemical processes in the model. These include: alternate methods for parameterization and estimation of NH3 dry deposition velocities; use of alternate thermodynamic equilibrium modules for representing the gas/aerosol partitioning of airborne-reduced and oxidized nitrogen; representation of sea-salt emissions and chemistry, and its impact on modeled aerosol composition; and methods for representing the optical and radiative properties of aerosols, and estimation of the aerosol direct radiative forcing over the eastern United States.
Comparison of model predictions with measurements from the Interagency Monitoring of Protected Visual Environments (IMPROVE) and Clean Air Status and Trends Network (CASTNet) networks suggest that the model captures the spatial variability in the measured PM2.5 concentrations, as well as the compositional characteristics of for SO42-, NO3-, NH4+ in the inorganic PM mass (r2 = 0.5). Both the measurements and the model show that in the eastern United States, SO42- constitutes a relatively large fraction of the fine particulate mass. The regional mean aerosol NO3- fraction to the total airborne oxidized nitrogen was 15.5 percent, which is consistent with the observed regional mean value of 13.8 percent inferred from the CASTNet measurements over the eastern United States for June 1996.
Simulated spatial distributions of the surface level NH3/NHx ratio show sharp gradients in the vicinity of NH3 source regions, suggesting a dominant role of NH3 sink processes. Comparison of the simulated ratio with limited available measurements suggest that improving grid resolution (from 36 km to 4 km) enables better representation of the spatially heterogeneous trends in the distribution of the ratio, especially in vicinity of dense NH3 sources.
Modeled representation of the gas/particle partitioning of airborne nitrogen was examined through comparisons of the RPMARES and ISORROPIA thermodynamic equilibrium modules for both box-model and 3-dimensional (3-D) calculations. Comparisons of 3-D fields suggest that for the SO42--H2O-NH4+-NO3- system, the aerosol compositions predicted by the two models are similar for a wide range of conditions.
Uncertainties in model representation of NH3 dry deposition was investigated through comparisons of three alternate parameterizations of the stomatal resistance (Rs): (1) a radiation-based formulation in which short-wave radiation is the driver (Wesley, 1989) (hereafter referred to as WES); (2) a more detailed scheme that, besides radiation, also accounts for variation in other physical parameters such as air temperature, relative humidity, and soil moisture (Jarvis-type: Jarvis, 1976; Pleim and Xiu, 1995) (hereafter referred to as PX); and (3) a photosynthesis-based scheme suggested by Caolaltz/Ball Berry (Ball, et al., 1987; Baldocchi and Meyers, 1998) using the gas exchange model (GEM). Because the formulations of PX and GEM need to be used in a meteorological model, we have implemented the WES and GEM formulations into the MM5, even though it already has the option to use a PX formulation (Xiu and Pleim, 2001). The estimated Rs values for each scheme then were used to estimate dry deposition velocities in MAQSIP; this approach provides a more consistent treatment of processes between the meteorological and chemistry transport models.
Several interesting features were noted from these simulations: (1) the computed canopy conductance shows significant variability not only across different land-use categories, but also between the three formulations for a given land-use type; (2) both the PX and the GEM formulations estimate larger canopy conductance compared to the WES formulation for almost all canopy types; (3) the photosynthesis-based formulation (GEM) shows most heterogeneity across different land-use types. In contrast, the PX formulation shows much less variability for the relatively dry conditions simulated in these experiments (June, 1996); and (4) both the Jarvis type (PX) and the photosynthesis-based model (GEM) estimate VdNH3 values that are significantly higher than those estimated by the Wesley-type parameterization of Rs, and also are more consistent with the magnitude suggested by recent VdNH3 measurements in Europe.
Quantitative estimates of the role of dry deposition in shaping the local and regional distributions of airborne-reduced nitrogen during typical summer conditions were developed through analysis of modeled process budgets. Analysis of NH3 budgets for Sampson County, NC, indicate the dominant role of transport processes (87 percent) in depleting surface-level NH3 emissions. Of this, approximately 76.5 percent of the total can be attributed to turbulent mixing. Hence, of the total NH3 emitted at the surface in Sampson County, a large fraction is mixed upwards through the depth of the boundary layer. At the surface level, 2 percent of the NH3 emissions are converted to particulate NH4+, cloud process (mixing in this case as the simulation period is devoid of significant precipitation) accounts for 5.5 percent, while local dry deposition within the county accounts for the remaining 5.5 percent.
As the NH3 is mixed upward through the depth of the boundary layer, it is converted to particulate NH4+. Our modeled budgets integrated over the depth of the boundary layer (0-2 km) suggest that of the total NH3 emitted in Sampson County, 27 percent is converted to NH4+, 5.5 percent is locally deposited within the county due to dry deposition, 52 percent is transported out of the county due to advection, while cloud mixing accounts for the remaining 15.5 percent; the cloud mixing budgets suggested by our model calculations can be interpreted as venting of free NH3 to the free troposphere, and support the theory of cloud venting being a possible source of NH3 in the free troposphere.
Journal Articles on this Report : 3 Displayed | Download in RIS Format
Other project views: | All 14 publications | 4 publications in selected types | All 3 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Dennis RL, Mathur R. Airshed domains for modeling atmospheric deposition of oxidized and reduced nitrogen to the Neuse/Pamlico system of North Carolina.Hydrological Science & Technology 2001;17(1-4):107-117. |
R826773 (2002) R826773 (Final) |
not available |
|
Mathur R, Dennis RL. Seasonal and annual modeling of reduced nitrogen compounds over the eastern United States:emissions, ambient levels, and deposition amounts. Journal of Geophysical Research-Atmospheres 2003;108(D15):4481. doi: 10.1029/2002JD002794. |
R826773 (2002) R826773 (Final) |
Exit |
|
Mathur R, Shankar U, Hanna A, Odman MT, McHenry JN, Coats CJ, Alapaty K, Xiu A, Arunachalam S, Olerud DT, Byun DW, Schere KL, Binkowski FS, Ching JKS, Dennis RL, Pierce TE, Pleim JE, Roselle SJ, Young JO. Multiscale Air Quality Simulation Platform (MAQSIP): initial applications and performance for tropospheric ozone and particulate matter. Journal of Geophysical Research 110(D13308), 35 pp. doi: 10.1029/2004JD004918. |
R826773 (Final) |
Exit |
Supplemental Keywords:
air, atmospheric chemistry, chemical transport, regional models, model evaluation, acid deposition, CASTNET, nitrogen, tropospheric ozone, particle size, particulate matter, PM, particulates, PM2.5, aerosol particles, air modeling, air pollution models, ambient aerosol, ambient aerosol particles, ambient air, eutrophication, fate and transport, fine particles, human exposure, nitrogen removal, oxides, ozone, regional scale, sulfur., RFA, Scientific Discipline, Air, Ecology, Environmental Chemistry, tropospheric ozone, Engineering, Engineering, Chemistry, & Physics, ambient aerosol, fate and transport, particle size, particulates, eutrophication, particulate matter, aerosol particles, fine particles, air modeling, ozone, ambient air, sulfur, air pollution models, human exposure, regional scale, PM2.5, nitrogen removal, troposphere, oxidesProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.
Project Research Results
- 2002 Progress Report
- 2001 Progress Report
- 2000 Progress Report
- 1999 Progress Report
- Original Abstract
3 journal articles for this project