Modeling Study of Dry Deposition of Ammonia in North Carolina.
No Thumbnail Available
Files
Date
2008-05-09
Authors
Journal Title
Series/Report No.
Journal ISSN
Volume Title
Publisher
Abstract
Proper description of removal processes for ammonia is essential in an air quality modeling system to accurately represent the ammonia and nitrogen budget. This problem is significant in Southeast U.S., both politically and environmentally in North Carolina (NC) due to increased odor, ammonia (NH3), and other emissions from hog and livestock sources. The transport and fate of NH3 are affected by various atmospheric processes and parameters such as meteorological variables, chemical reactions, and dry and wet deposition. Meteorological parameters such as temperature, humidity, and wind fields are very important in any air quality study as they directly affect the transport and chemistry of species of concern such as ozone (O3) and NH3. Input of accurate meteorological variables is vital for accurate simulations of ambient air quality. Dry deposition refers to the downward transport of gaseous and particulate species from the atmosphere onto surfaces in the absence of precipitation/hydrometeors. Direct measurements of individual parameters such as friction velocity, leaf area index, and stomatal resistance that are used in dry deposition calculations are generally lacking. This limits our understanding of the important physical, chemical, and biological processes involved in dry deposition and the evaluation of the accuracy of the model in replicating them.
The objectives of the proposed research are as follows:
Evaluate dry deposition predictions from the MM5⁄CMAQ modeling system
Identify influential parameters for the calculation of dry deposition of ammonia
Improve dry deposition parameterizations in the MM5⁄CMAQ modeling system
The approach undertaken to achieve these objectives is divided into three steps. A box model is first used to conduct sensitivity tests for summer and winter conditions to identify the role played by several parameters in the calculation of dry deposition. Parameters considered are leaf area index (LAI), surface roughness length (SRL), surface relative humidity (RHs), cuticle resistance (rcutmax), and canopy wetness content (CWC). The number of test simulations performed are 3 (site locations) * 10 (RHs) * 11 (CWC) * 5 (LAI) * 5 (SRL) * 2 (rcutmax). Results indicate the importance of proper specification of RHs, LAI, and surface roughness length. Box model results justify the use of a categorization scheme to parameterize CWC, which is currently treated as 0 or 1.
Then, a 3-D meteorological model (i.e., Penn State University /National Center for Atmospheric Research (PSU⁄NCAR) mesoscale model, known as MM5, model (v. 3.7)) is then used to evaluate the impact of grid resolution and nudging on meteorological predictions. Overall, the differences between model results with and without nudging scheme are relatively small (up to 3% in terms of normalized mean biases (NMB)). When the domain was divided into specific regions such as rural, urban, and coastal based on scale of development and location, the differences between the simulations were found to be much higher. Using a finer grid resolution with nudging generally provides the best overall results during summer. Although differences in the overall domain are relatively small (e.g., up to 3% in terms of NMBs), the differences in area-specific analyses may be higher (e.g., up to 30% in NMB for WSP). Based on sensitivity evaluations in the box and meteorological models, the effects of grid resolution, nudging, and changes in dry deposition parameterizations are studied in a 3-D air quality model (i.e., CMAQ).
The chemistry transport model (i.e., U.S. EPA Models-3 Community Multiscale Air Quality (CMAQ) model) is used to analyze the impact of both meteorology and dry deposition parameterization changes (from the box model study) on a 3-D model. Impacts of meteorology and dry deposition on the air quality model are evident. For July, underpredictions in SO42- (NMBs of -17.6%, -31.3%, and -32.1% for the baseline, modified rcutmax, and modified CWC simulations, respectively) and NH4+ (NMBs of -29.7%, -46.3%, and -46.1% for the base, modified rcutmax, and modified CWC simulations, respectively) are greater in both the sensitivity simulations, while NO3- values do not vary significantly (NMBs of -87.5%, -87.4%, -87.6% for the base, modified rcutmax, and modified CWC simulations, respectively). Ammonia deposition velocities and fluxes are compared to observational data obtained from Phillips et al. (2004) for summer (July and August) and winter (December), 2002. The 12-km simulation predicts higher values than the 4-km simulation other than for nighttime winter. In the base case during both summer and winter, NH3 Vd was underestimated during daytime and overestimated at nighttime. With modified maximum cuticle resistance coefficient the deposition velocity increased for day and night. With modified canopy wetness contents, the increase in day and night Vd was slightly lower.
Improved understanding and representation of dry deposition process in the modeling system would aid in maintaining a total-N2 budget and establishing regulatory standards for NH3 emissions for North Carolina.
Description
Keywords
CMAQ, North Carolina, ammonia, MM5, dry deposition
Citation
Degree
MS
Discipline
Marine, Earth and Atmospheric Sciences
