Investigating the impact of haze on cloud detection of passive satellite by comparing MODIS, CloudSat and CALIPSO

Abstract:

The cloud detection algorithm for passive sensors is usually based on a fuzzy logic system with thresholds determined from previous observations. In recent years, haze and high aerosol concentrations with high AOD occur frequently in China and may critically impact the accuracy of the MODIS cloud detection. Thus, we comprehensively explore this impact by comparing the results from MODIS/Aqua (passive sensor), CALIOP/CALIPSO (lidar sensor), and CPR/CloudSat (microwave sensor) of the A-Train suite of instruments using an averaged AOD as an index for an aerosol concentration value. Case studies concerning the comparison of the three sensors indicate that MODIS cloud detection is reduced during haze events. In addition, statistical studies show that an increase in AOD creates an increase in the percentage of uncertain flags and a decrease in hit rate, a consistency index between consecutive sets of cloud retrievals. Therefore, we can conclude that the ability of MODIS cloud detection is weakened by large concentrations of aerosols. This suggests that use of the MODIS cloud mask, and derived higher level products, in situations with haze requires caution. Further improvement of this retrieval algorithm, is desired as haze studies based on MODIS products are of great interest in a number of related fields.