Synchronized Declines in Global Landsat Cloud Occurrence Probability and Area.

Abstract Cloud cover impacts Earth's radiation budget as both a cause and an effect of climate change. While long-term satellite observations are vital for climate studies, existing products are limited by coarse resolutions that fail to capture fine-scale cloud heterogeneity. Despite the five-decade Landsat archive, global pixel-level analysis remains constrained by computational limitations. We investigated spatiotemporal patterns of cloud occurrence probability (COP) and cloud contaminated area (CCA) from continental to global scales using 1984–2023 data from Landsat-5, Landsat-7, and Landsat-8. Results reveal pronounced variations: global average COP is 34.39%, with Europe recording the highest rate (43.95%) and Oceania the lowest (21.23%). Tropical regions exhibit higher COP (48.10%) than extra-tropical zones (39.51%). Global mean CCA covers 47.05% of Earth’s surface (7.01 × 10 8 km 2 ), with Asia contributing the largest share (36.6%). Unlike the stable tropics, CCA shows significant seasonality at mid- and high-latitudes, peaking in July and reaching a minimum in December. A key finding is the synchronized decline in global cloud cover COP and CCA during the 2014–2023 OLI era ( p < 0.05), with South America exhibiting the most pronounced regional reduction. Beyond validating the Landsat archive, our findings demonstrate the unique capacity of 30-m observations to resolve sub-grid cloud heterogeneity, offering a potential benchmark for assessing partial coverage biases in coarser global products. This high-resolution record provides an essential foundation for refining cloud-climate feedbacks in future climate modeling.