A multi-model based dataset of global atmospheric moisture source-sink relationships and atmospheric basins.

A comprehensive understanding of the atmospheric hydrological cycle is crucial for studying the transboundary impacts of climate and anthropogenic changes on global water resources. This study presents AMSSRAB, a global dataset of atmospheric moisture source-sink relationships (AMSSRs) between evaporation and precipitation locations, and the derived atmospheric basins (ABs) that regionalize the global atmospheric hydrological cycle. By integrating three atmospheric moisture tracking models (WAM2layers, UTrack, and WaterSip), AMSSRAB provides a multi-model ensemble representation of atmospheric moisture flows. The dataset provides seasonal atmospheric moisture source-sink relationships at 1° resolution over 40 years (1979-2018). The derived atmospheric basins identify quasi-independent moisture circulation systems characterized by high internal recycling. AMSSRAB demonstrates strong agreement with bias-corrected reconstruction data and previously published datasets while offering a regional perspective that complements existing evaporationshed-precipitationshed frameworks. This dataset enables researchers to investigate moisture source-sink relationships at multiple scales, analyze regional moisture recycling patterns, and examine atmospheric moisture responses to climate variability within the Earth's atmospheric water system.