Development of a Short-term Prediction Model for Predicting Photochemical Oxidants in a Local Area

In this paper a short-term prediction model is developed to support the photochemical smog emergency measure. This model is to predict the daily maximum level of oxidants in a local area using the data collected in the morning of the day. The model is using a fuzzy model based on the revised model of Takagi-Sugeno method, and the weighted linear least squares are used to construct the multiple linear regression equations so that prediction of the high oxidants concentration level around the emergency level suits well. There exist six air pollution monitoring stations in this area. Two different models are developed and evaluated their applicability. Model 1 is to predict the daily maximum level of oxidants in this area directly, and model 2 is to predict the daily maximum levels of oxidants in this area from the predicted values at each monitoring station. It is shown that the model 2 is better than the model 1 to predict the daily maximum level of oxidants for the area