Yerramilli Anjaneyulu, Srinivas Challa Venkata, Dasari Hari Prasad, Tuluri Francis, White Loren D, Baham Julius M, Young John H, Hughes Robert, Patrick Chuck, Hardy Mark G, Swanier Shelton J
Trent Lott Geospatial Visualization Research Centre, Mississippi e-Center, Jackson State University, 1230 Raymond Road, Jackson MS 39204, USA.
Int J Environ Res Public Health. 2009 Mar;6(3):1055-74. doi: 10.3390/ijerph6031055. Epub 2009 Mar 11.
Atmospheric dispersion calculations are made using the HYSPLIT Particle Dispersion Model for studying the transport and dispersion of air-borne releases from point elevated sources in the Mississippi Gulf coastal region. Simulations are performed separately with three meteorological data sets having different spatial and temporal resolution for a typical summer period in 1-3 June 2006 representing a weak synoptic condition. The first two data are the NCEP global and regional analyses (FNL, EDAS) while the third is a meso-scale simulation generated using the Weather Research and Forecasting model with nested domains at a fine resolution of 4 km. The meso-scale model results show significant temporal and spatial variations in the meteorological fields as a result of the combined influences of the land-sea breeze circulation, the large scale flow field and diurnal alteration in the mixing depth across the coast. The model predicted SO(2) concentrations showed that the trajectory and the concentration distribution varied in the three cases of input data. While calculations with FNL data show an overall higher correlation, there is a significant positive bias during daytime and negative bias during night time. Calculations with EDAS fields are significantly below the observations during both daytime and night time though plume behavior follows the coastal circulation. The diurnal plume behavior and its distribution are better simulated using the mesoscale WRF meteorological fields in the coastal environment suggesting its suitability for pollution dispersion impact assessment in the local scale. Results of different cases of simulation, comparison with observations, correlation and bias in each case are presented.
利用HYSPLIT粒子扩散模型进行大气扩散计算，以研究密西西比湾沿海地区点源高架源排放的空气传播污染物的输送和扩散。在2006年6月1日至3日的典型夏季期间，使用具有不同空间和时间分辨率的三个气象数据集分别进行模拟，代表弱天气形势。前两个数据是美国国家环境预测中心（NCEP）的全球和区域分析数据（FNL、EDAS），而第三个是使用天气研究和预报模型生成的中尺度模拟数据，其嵌套区域的精细分辨率为4公里。中尺度模型结果显示，由于海陆风环流、大尺度流场以及沿海混合层深度的日变化的综合影响，气象场存在显著的时空变化。模型预测的SO₂浓度表明，在三种输入数据情况下，轨迹和浓度分布有所不同。虽然使用FNL数据进行的计算显示总体相关性较高，但白天存在显著正偏差，夜间存在负偏差。使用EDAS场进行的计算在白天和夜间均显著低于观测值，尽管羽流行为遵循沿海环流。在沿海环境中，使用中尺度WRF气象场能更好地模拟羽流的日行为及其分布，表明其适用于局部尺度的污染扩散影响评估。文中给出了不同模拟案例的结果、与观测值的比较、每种情况下的相关性和偏差。
