Cloud amount uncertainty in merged CloudSat-CALIPSO radar-lidar observations

Three dimensional structure of cloud cover is one of the Essential Climate Variable required for accurate monitoring of the state and change of global climate. Joint CloudSat-CALIPSO space mission have provided the most reliable and comprehensive 3D information on cloud distribution worldwide to date. However, the data resulted from observations collected every 16 days – sampling interval which can be considered infrequent for most of climate-oriented applications. The reliability of the data also depends on cloud regime, and area (grid cell size) over which the data are aggregated, further complicating the uncertainty aspect of lidar-radar profiling missions. The important question related to the CloudSat-CALIPSO dataset is whether 16-day revisit period for CloudSat-CALIPSO mission is sufficient to provide a climate characteristics at high statistical significance? We address that problem evaluating the full CloudSat-CALIPSO record (2006-2011), available to the scientific community as 2B-GEOPROF-LIDAR product. The analysis focuses on two aspects. First, we perform a point estimation to determine the minimum significance level at which the lidar-radar data (mean value) is statistically significant. Second, using a bootstrap approach we calculate confidence intervals for the mean value at fixed .95 and .99 thresholds. Therefore we reveal how wide is the actual uncertainty range at 16-day revisit. The analysis accounts for grid box size over which individual lidar-laser profiles were aggregated. The study was founded by National Science of Poland under the contract no. UMO-2017/25/B/ST10/01787.