Global Groundwater Drought Assessment Revisited: A Holistic Re‐Evaluation of the GRACE‐Groundwater Drought Index Across Major Aquifers

Abstract The Gravity Recovery and Climate Experiment (GRACE) and GRACE Follow‐On missions have enriched global groundwater monitoring, forming the basis for tools that detect groundwater drought, including the GRACE‐Groundwater Drought Index (GGDI). The reliability of GGDI is fundamentally tied to the accurate isolation of a representative groundwater storage anomaly (GRACE‐GWA) signal from GRACE observations, a challenge heightened by the scarcity of direct water budget measurements and the diverse methodologies applied in GRACE data processing. In this global assessment, we integrate multi‐model GRACE‐GWA estimates into the GGDI framework to examine how variability among these estimates influences groundwater drought interpretation across 37 study aquifers. Results reveal substantial sensitivity of key drought indicators to input uncertainty, with maximum observed intra‐basin discrepancies reaching 11 events, 122 months in maximum duration, 63.33 months in average duration, 24.47 in severity, and 5.4 in intensity. Aquifer memory, inferred from GGDI autocorrelation, reveals pronounced variability, most notably in the Nubian Basin where memory estimates range from 3 to 61 months amongst multi‐model realizations. Aquifers with higher memory tended to experience fewer drought events, yet those droughts were typically longer and more intense. Our findings underscore that even modest discrepancies in GRACE‐GWA methodologies can translate into considerable uncertainties in both drought indicators and aquifer memory, thereby compromising the reliability of groundwater drought assessments.