Dynamics of model-based groundwater-land surface response times as a dryland flash drought diagnostic
摘要
The delayed response of groundwater to surface soil moisture anomalies (Topt), which reflects how quickly surface signals propagate to aquifers, varies across wetness regimes. Understanding its spatiotemporal variability may help diagnose flash droughts, yet it remains underexplored. Here we examine global relationships between modeled Topt and flash droughts using a dynamic exponential filter. We find that background atmospheric aridity generally controls this relationship, since longer Topt occurs with frequent flash droughts in drylands. However, seasonal variability is largely governed by background terrestrial wetness conditions, with stronger bidirectional Topt–flash drought sensitivities during wetter seasons, even in non-hotspots. This is because flash droughts driven by rapid atmospheric dryness propagation to the land are reflected in Topt memory, predominantly shaped by land-atmosphere interplays. Combined evapotranspiration-runoff deficit in dry conditions, especially the dominance of evapotranspiration, rapidly prolongs Topt and reduces its sensitivity to flash droughts. These insights highlight Topt as a promising flash drought diagnostic indicator, offering a practical pathway for improving prediction.