Development of an Algorithm for Calculating PM10 from Geostationary Satellite Data

Abstract

In this study, an algorithm for estimating particulate matter up to 10 micrometers in size ( PM10) over Thailand using geostationary satellite images was developed. Ground-based measurements of PM10 at various stations of Pollution Control Department (PCD) were collected. For satellite data, MTSAT satellite images were used to estimate aerosol optical depth using a look-up table algorithm. Relationship between PM10 and satellite-derived aerosol optical depth was examined and a statistical-based model for estimating PM10 was obtained as a result. Validation of the model was accomplished against an independent PM10 measurement data collected at the PCD stations. The comparison yielded good agreement with mean bias error (MBE) of -4.6 g/m 3(-11.3 % ) and root mean square Error (RMSE) of 10.9 g/ m 3 (26.9 % ). The model was used to generate graphic maps of PM10 for the entire region of Thailand.