Browsing by Author "Pal Arya, Committee Member"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • No Thumbnail Available
    Incorporation of the Model of Aerosol Dynamics, Reaction, Ionization, and Dissolution (MADRID) into the Weather Research and Forecasting Model with Chemistry (WRF/Chem): Model Development and Retrospective Applications
    (2009-07-11) Hu, Xiaoming; Pal Arya, Committee Member; Ken Schere, Committee Member; Yang Zhang, Committee Chair; Lian Xie, Committee Member; Sethu Raman, Committee Member
    Gas/particle mass transfer process plays an important role in determining aerosol mass concentrations and shaping aerosol size distribution. Its treatments in three dimensional (3-D) Air Quality Models (AQMs), however, are largely uncertain. In this thesis work, the gas/particle mass transfer approaches in an aerosol module are improved and evaluated to identify an accurate yet computationally efficient approach for use in 3-D AQMs. The aerosol module with the improved gas/particle mass transfer approaches has been incorporated into a state-of-science air quality forecasting (AQF) system and evaluated with two 3-D applications. Several stand alone condensation schemes used in AQMs are first evaluated with a hypothetical condensation-only case. The original formulation of the Bott scheme as implemented in several AQMs is found to be subject to upstream diffusion thus does not warrant continuous use without modifications. The analytical predictor of condensation with a moving center approach (APC_MC) is shown to be more accurate than the Bott and Trajectory-Grid (T-G) condensation schemes, thus has been incorporated into the Model of Aerosol Dynamics, Reaction, Ionization and Dissolution (MADRID) to solve the gas/particle mass transfer process explicitly. The improved hybrid (i.e., hybrid/APC_MC) and kinetic (i.e., kinetic/APC_MC) approaches and the pre-existing bulk equilibrium approach in MADRID are tested using observational data. The hybrid/APC_MC and kinetic/APC_MC are recommended for 3-D applications due to the best compromise between accuracy and computational efficiency. The improved MADRID has been incorporated into WRF/Chem (referred to as WRF/Chem-MADRID hereafter). WRF/Chem-MADRID with three gas/particle mass transfer approaches (i.e., bulk equilibrium (EQUI), hybrid/APC_MC (HYBR), and kinetic/APC_MC (KINE)) has been tested and evaluated with a 5-day episode from the TexAQS-2000. WRF/Chem-MADRID simulates meteorological parameters fairly well. Simulated hourly O3 shows a high correlation coefficient (0.83) with observations and the overall bias is about -1.8 ppb. Some daily peak O3 mixing ratios are underpredicted, which is possibly due to uncertainties in emissions, inaccurate predictions of small scale meteorological processes, and missing of an OH source and chlorine chemistry in the gas phase mechanism. WRF/Chem-MADRID (EQUI), (HYBR), and (KINE) overpredict PM2.5 by 37.1%, 35.8%, and 36.5%, respectively. Major differences in simulation results by three gas/particle mass transfer approaches occur over coastal areas, where WRF/Chem-MADRID (EQUI) predicts higher PM2.5 concentrations than those predicted by WRF/Chem-MADRID (HYBR) and (KINE) due to improperly redistributing condensed nitrate from the chloride depletion process to fine mode. In comparison, WRF/Chem-MADRID (KINE) correctly predicts chloride depletion process. WRF/Chem-MADRID (HYBR) predicts chloride depletion process correctly for the last two sections (sections 7 and 8), which are solved by the kinetic approach, while the predictions for section 6 may be still biased due to the use of bulk equilibrium approach. In addition to its surface concentration, the column abundance of aerosol is also evaluated. WRF/Chem-MADRID captures the regional-scale AOD distribution and its day-to-day variability while biases exist over certain areas. For the application to the 2004 NEAQS episode, WRF/Chem-MADRID gives comparable overall O3 performance as other AQMs and better O3 performance than some other AQM over certain areas possibly due to the more realistic convective mixing treatment in the model. WRF/Chem-MADRID (HYBR) and WRF/Chem-MADRID (KINE) show better skill than WRF/Chem-MADRID (EQUI) in terms of nitrate predictions over coastal areas. Model simulations confirmed that NEI99 v3 overestimates the actual emissions in 2004, particularly over urban areas.
  • No Thumbnail Available
    Measurement, Analysis, and Modeling of Fine Particulate Matter in Eastern North Carolina
    (2007-02-28) Goetz, Stephen; Viney Aneja, Committee Chair; Yang Zhang, Committee Member; Pal Arya, Committee Member
    An analysis of fine particulate mass concentrations in eastern North Carolina was conducted in order to investigate the impact of the hog industry and its emissions of ammonia into the atmosphere. This analysis included collecting acidic gas and inorganic fine particulate concentrations at a hog facility and at a site ˜10 miles away, while the regional impact of hog industry was studied with data, which was obtained from the North Carolina Division of Air Quality, for multiple regional sites (Fayetteville, Goldsboro, Jacksonville, Kenansville, Kinston, Raleigh, Wilmington). This regional fine particulate data was then simulated using ISORROPIA, a thermodynamic model that simulates the gas and aerosol equilibrium of inorganic atmospheric species. The local analysis showed the dominance of the ammonium sulfate aerosol, and the seasonal observations showed the impact of both urban areas and marine areas on this region. While no meteorological trends were seen in the local data, the time series plots showed an environment where regional sulfate plays a large part. While nitrate is present in this environment, it is present 1 order of magnitude less then the sulfate aerosol. The ammonium and sulfate values are highly correlated, and the molar ratio is consistent with the relative values of ammonium, sulfate, and nitrate present. The regional observational data analyses show that the major constituents of fine particulate matter are organic carbon, sulfate, nitrate, ammonium and elemental carbon. The observed PM2.5 concentration is positively correlated with temperature but negatively-correlated with wind speed. The correlation between PM2.5 mass and wind direction at some locations indicates the impact of the emissions from hog facilities on PM2.5 formation. The modeled results overpredict the observed results in each case, where the nitrate concentrations had the largest percentage overprediction. The predicted total inorganic PM concentrations are overpredicted by 40-45% of the observed values under conditions with median initial total PM species concentrations, median RH and median temperature. The ambient conditions with high PM precursor concentrations, low temperature and high relative humidity favor the formation of the secondary PM. The model runs of the individual days at the three speciated sites showed overprediction for all species, where some predicted values of ammonium were within a factor of 2 of the observed concentrations.