Evaluation of drought indices at interannual to climate change timescales: a case study over the Amazon and Mississippi river basins

Abstract
The present study compares three meteorological drought indices (scPDSI, SPI and SPEI respectively) and their ability to account for the variations of annual mean river discharge on both interannual and climate change timescales. The Standardized Runoff Index (SRI) is used as a proxy of river discharge. The Mississippi and Amazon river basins provide two contrasted testbeds for this analysis. All meteorological drought indices are derived from monthly 2-meter temperature and/or precipitation, using either gridded observations or outputs of a global climate model. The SPI based solely on precipitation is not outperformed by the SPEI (accounting for potential evapotranspiration) and the scPDSI (based on a simplified water balance) at detecting interannual SRI variations. Under increasing concentrations of greenhouse gases, the simulated response of the areal fraction in drought is highly index-dependent, suggesting that more physical water balance models are needed to account for the impact of global warming on hydrological droughts