Aerosol Optical Depth Indices Derived From Simulated MODIS Apparent Reflectance Using The6S Radiative Transfer Model

Nowadays, air pollution has become an increasing serious concern in the modern society. Atmospheric aerosol is one of the important factors. In addition, the effect of aerosol on climate and environment has been the hot issues of atmospheric science and related fields. Aerosol Optical Depth (AOD) is the key factor to calculating the magnitude of aerosol concentration, to evaluating atmospheric pollution and to researching the climatic and environmental effects of atmospheric aerosol. It can be a particularly important research direction for areas such as social and economic development, human’s life and production in terms of how to effectively and conveniently get the spatial distribution of AOD.As an efficient and convenient information exaction method, remote sensing index is widely used to extract different ground features. Typical indices include vegetation index, water index and urbanization index. This paper is based on basic principles for developing remote sensing index, and use6S (Second Simulation of the Satellite Signal in the Solar Spectrum) radiation transfer model to simulate the sensitivity of AOD concentration to MODIS channel1to channel7top of atmospheric reflectance. The result shows that channel3is the most sensible to AOD, while channel7is the most insensible to AOD concentration. Based on this, three different AOD indices were designed in this paper, that is, channel differencing, channel ratioing and normalized differencing. Simulation results show that the indices are significantly correlated with AOD concentration, with a maximum coefficient of relationship reaching0.9905.Values of AOD550were obtained from observations in AERONET site of Taihu Lake station. The values indicate the AOD value in this area is greater than o.5for all seasons. Also, the atmospheric turbidity is relatively high. By using the construction method on AOD indices and MODIS reflectance data, the differencing, ratioing and normalized differencing AOD indices were developed. Then, the paper analyzed the correlation of these three AOD indices and AOD measurements. The correlation analysis shows that channel differencing is more appropriate to indicate aerosol concentration. The correlation coefficient of the four seasons reaches0.5853in spring,0.8604in summer,0.6851in autumn and0.3327in winter, respectively. Consequently, channel differencing is convenient to indicate atmospheric aerosol pollution.