The use and abuse of drought indicators in (paleo)climate research

Abstract

Abstract Europe’s recent multi-year droughts have led to substantial socio-economic consequences, but understanding the severity and cause of these extreme events still poses a lot of challenges, Here we provide a detailed analysis of short-term and multi-year drought events in Europe, by employing three different long-term reconstructions/simulations for three different drought indicators (i.e., the Palmer Drought Severity Index – PDSI, the Standardized Precipitation Evapotranspiration Index -SPEI and the Soil Moisture Index – SMI). When analyzing the year-to-year variability we found that the driest years, at the European level, in terms of magnitude and spatial extent, are different for the three indicators (i.e., PDSI – 1659, SPEI – 1738, and SMI – 1868). The same holds true if we look at multi-year extreme drought events. While PDSI indicates the decade 1941–1950 as the driest one over the last 600 years, both in magnitude and spatial coverage, SPEI has identified the decade 2011–2018 as the driest one on record. When looking at soil moisture multi-year droughts, SMI indicates that the decade 1781–1790 was the driest one on record followed by 2011–2018. At a more regional scale (i.e., the Mediterranean region - MED, Central Europe – CEU, and the northern part of Europe - NEU) the results are also divergent between the three drought indicators and the driest years/decades vary depending on the region/indicator. Based on these findings, we argue that one should be extremely careful when using a single (paleo)drought indicator to infer the extremeness of recent long-lasting drought events, each indicator having its own advantages and disadvantages and responding differently to the currently observed changes in the climate system.