Numerical Model-Based Artificial Neural Network Model and Its Application for Quantifying Impact Factors of Urban Air Quality

摘要: Knowledge of the relationship between air quality and impact factors is very important for pollution control urban environment management. Relationships winter pollutant concentrations local meteorological parameters, synoptic-scale circulations precipitation were investigated based on observed concentrations, high-resolution data from Weather Research Forecast model gridded reanalysis data. Artificial neural network (ANN) was developed using a combination numerical derived variables indicating emission circulation type variations estimating daily SO2, NO2, PM10 over Lanzhou, Northwestern China. Results indicated that ANN can satisfactorily reproduce level their day-to-day variations, with correlation coefficients modeled ranging 0.71 to 0.83. The effect four factors, i.e., type, condition, variation, wet removal process, quantified winters 2002–2007. Overall, condition main factor causing followed by process. With limited data, this work provides simple effective method identify pollution, which could be widely used in other areas regions planning or management purposes.