Bayesian Cloud Detection over Land for Climate Data Records

Cloud detection is a necessary step in the generation of land surface temperature (LST) climate data records (CDRs) and affects quality uncertainty. We present here sensor-independent Bayesian cloud algorithm show that it suitable for use production LST CDRs. evaluate performance with reference to two manually masked datasets Advanced Along-Track Scanning Radiometer (AATSR) find 7.9% increase hit rate 4.9% decrease false alarm when compared operational mask. then apply four instruments aboard polar-orbiting satellites, which together can produce global, 25-year CDR: second (ATSR-2), AATSR, Moderate Resolution Spectroradiometer (MODIS Terra) Sea Land Surface Temperature (SLSTR-A). The assessed respect situ ceilometer measurements periods overlap between sensors. consistency sensors, mean differences 4.5% ATSR-2 vs. AATSR MODIS, 2.5% MODIS SLSTR-A. This important because consistent needed observational stability CDR. application scheme CDRs thus shown be viable approach achieving over time.