Nonlinear estimation for PM_(2.5) transmission effects in Jefferson Co., Texas

摘要

Urban air pollution consists of emission and transmission of air pollutants. In this article, we discuss a complex valued nonlinear regression model to characterize the effect of pollutant transmission by estimating the direction and velocity of the spatial propagation at a fixed sequence of time intervals. In case the pollutant is transmitted from a distant source during a certain time period, the monitoring stations in the local network detect the signal with time delays, which are nonlinearly related to the velocity and angle of approach or azimuth. An observed vector time series in the monitoring network is modeled as the sum of a lagged or delayed stochastic signal and a zero mean stationary noise vector. The Fourier transformation is applied to the model, and the maximum likelihood estimation of the transmission parameters under such a model is considered. Those estimates can be used to identify the possible sources of local pollutants and for better understanding of spatial distribution of pollutants. Data used in this study are based on hourly PM_(2.5) records from three monitoring stations in the Beaumont-Port Arthur area in Texas for April to December 2002. The results indicate that during the months May, November and December, PM_(2.5) transmission were significant, with the dominant direction from the east. The velocity estimates during the period range from 4 to 20 km/h. During the summer season (June to October) the transmission effect was not very evident except for the first week of August and the first week of September. The model well detects both consistent and momentary transmissions. The parameter estimates and their standard errors are given for each week separately.