Leveraging global climate models to assess multi-year hydrologic drought

Abstract

Global climate models (GCMs) offer value for assessments of future water supply and multi-year hydrologic drought. Leveraging GCM data, we develop and analyze global scenarios of mean annual runoff over a span of 640 years. Runoff data from eighteen GCMs are evaluated for skill and bias-adjusted to reflect observations. Unprecedented projections of mean runoff, drought severity, and drought duration are found for 37%, 28%, and 23% of analyzed global land area, respectively, with regions on all continents presenting a risk of a drier future. Conversely, northern latitudes show evidence of increasing runoff, less severe, and shorter-duration droughts. Outside these regions, projections are either indistinguishable from internal climate variability or unreliable due to conflicting signal-to-noise ratios and ensemble agreement. Our analysis contributes to a global gap in understanding future multi-year hydrologic droughts, which can pose significant socio-economic risks.