 This study uses remote sensing and model-based datasets with a machine learning approach to predict global land subsidence magnitude at high spatial resolution, two kilometers, estimate aquifer storage loss due to consolidation of 17 cubic kilometers per year globally, and quantify key drivers of subsidence. The results show that roughly 73% of the mapped subsidence occurs over crop land and urban areas, highlighting the need for sustainable groundwater management practices in these areas. This article was authored by M.D. Fahim Hassan, Mayan Smith, Sonos Vagidian, and others.