Spatially and seasonally non-stationary relationshipsbetween PMsub10/sub and related factors in Eastern China bygeographically weighted regression

摘要

The potential of satellite data used for particular matter monitoring is a crucial subject in air qualityresearch. PM10 is influenced by many meteorological factors and has a difference correlation withaerosol optical depth in different place. Geographically weighted regression (GWR) model have beenproved to be an effective methods for spatial variation analysis. This paper presented results from astudy of PM10 concentration from API in eastern China from 2005 to 2010. Wavelet analysis was usedfor analyzing the periodicity characteristics of PM10 and AOD. The correlations between PM10 andmeteorological factors were also analyzed without AOD and with AOD added, respectively. Obviousspatial and seasonal non-stationary distributions of PM10 concentration were found with spatialauto-correlation analysis. PM10 concentration and AOD have similar periods and discontinuitycharacteristics in 41 months scale and 70 months scale. Correlation between PM10 concentration andmeteorological factors were improved when AOD added as a factor, and the tempo-spatial distributionsof the correlations were non-stationary in eastern China because of differences of the regional weatherconditions and the pollution sources.© (2012) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.