Analysis of the methods of predicting high concentrations of air pollutants

Abstract

The model used currently for predicting high levels of atmospheric particulate loads is shown to be inappropriate for the five-state area encompassing, Texas, New Mexico, Oklahoma, Arkansas, and Louisiana. It is shown that within this region high pollution can exist even when upper-air winds are double the values prescribed in the national model. The importance of subsidence temperature inversions as a primary factor in incidents of high pollution is challenged. It is shown that pollution episodes can occur in any month, but summer occurrences cover a small geographical area and the particulate counts are low during this season. Vector statistics are employed to ascertain the climatology of the direction and speed of the surface and upper winds. The assumption of a circular wind distribution is proven to be invalid over the majority of the region. It is shown that high counts of suspended particulates can occur even when significant amounts of precipitation fall during the pollution episode. The discontinuance of the national model for predicting high levels of air pollution is recommended. The need for future studies which may lead to the development of a regionally unique forecasting model is cited.