Some issues of using CALIPSO data to study the backscattering properties of aerosol in the vicinity of clouds

CALIOP provides a range resolved large dataset of global clouds and aerosols. Unlike MODIS observations, CALIPSO data are not affected by cloud adjacent effect. We use CALIOP to analyze the properties of aerosols near clouds. However, due to the inherently large noise and the large dynamic range of cloud and aerosol optical properties, errors in cloud-aerosol classification can occur, especially in the transition region that surrounds clearly identified clouds. This presentation uses CALIOP data to revisit the topic of aerosol backscattering enhancement near clouds and its correlation with cloud coverage, but with particular attention to the effects of data selection based on CAD values characterizing the confidence level of cloud-aerosol discrimination. The results show that the closer the distance to clouds, the more data has lower confidence levels. The results also indicate that aerosol backscattering enhancements near clouds can be overestimated if low-confidence data are not removed. The results also show that near-cloud changes in aerosol properties estimated under different CAD values behave generally in the same manner, but the values and their slopes near clouds strongly depend on the CAD value selection.