Home Articles Aerosol Remote Sensing Using Landsat TM Data Over Penang

Aerosol Remote Sensing Using Landsat TM Data Over Penang

09/01/2009 2 Minutes Read

Abstract

Aerosol Remote Sensing Using Landsat TM Data Over Penang

Lim Hwee San

School of Physics, Universiti Sains Malaysia,
Malaysia
Email: [email protected]

Accurate retrieval of air quality information from space is necessary to provide a reliable high spatial resolution map. Aerosol remote sensing is very much dependent on the quite accurate knowledge of the surface reflectance retrieved from satellite image. In this paper, we retrieved atmospheric reflectance from a Landsat TM image over Penang, Malaysia for air quality mapping using an algorithm. The algorithm was developed based on aerosol characteristics in the atmosphere to determine the concentration of particulate matter of size less than 10 micron (PM10). Ground measurements of PM10 were collected simultaneously with the satellite image acquisition using a DustTrak meter. The digital numbers (DN) for each band corresponding to the ground-truth locations were extracted and then the digital numbers for the two visible bands, (blue and red), were converted into radiance and reflectance values. A total of 7 dates of Landsat TM satellite images were analysed in this study. The relfectance values recorded by the satellite sensor at the top of atmosphere was the sum of the surface reflectance and atmospheric relfectance. So, the reflectance measured from the satellite sensor at the top of atmosphere was subtracted by the amount given by the surface reflectance to obtain the atmospheric reflectance. In this study, the surface reflectance values in the visible bands, (red and blue), were retrieved from dark targets using their relationship with the mid-infrared band data at 2.1 micro meter. The retrieved atmospheric reflectances were used for algorithm regression analysis. The proposed algorithm produced high correlation coefficient (R) and low root-mean-square error (RMS). PM10 concentration over Penang Island, Malaysia was mapped using the proposed algorithm. Finally, the PM10 map was smoothed using an average filter of window size 3 by 3 and colour-coded for visual interpretation.