A Newly Computational Intelligence Algorithm for Air Quality Mapping Using SPOT Image over Penang, Malaysia

The effect of aerosols on the global climate system is one of the major uncertainties of present climate predictions. This study is dealing with obtaining particulates matter of less than 10-micro in diameter (PM10) using SPOT data. The objectives of this study is to measure air quality parameters and to develop a model for relating ground truth data to the remote sensing images over Penang, Malaysia. In situ data of the PM10 measurements were collected simultaneously with the digital images acquisition using a DustTrak meter. The sun radiation measurements at the ground surface were collected using a handheld spectro radiometer. The digital images were separated into three bands namely red, green and blue bands for multispectral algorithm calibration. The selected parameter was particulate matter less than 10 micron (PM10). The digital numbers of the corresponding in situ data were converted into irradiance and then reflectance. The atmospheric reflectance values were extracted from the satellite observation reflectance values subtrated by the amount given by the surface reflectance. The atmospheric reflectance values were later used for PM10 mapping using the calibrated algorithm. The relationship between the reflectance and the corresponding air quality data was determined using regression analysis. A new algorithm was developed for detecting air pollution from the digital images chosen based on the highest correlation coefficient, R and lowest root mean square error, RMS for PM10. The results show that the use of remotely sensed data produced better spatial resolution air quality map compared to the spacing between ground stations.