Capturing the Spatiotemporal Dynamics of Soil Moisture Droughts under a Changing Climate

Droughts are an increasingly disastrous threat to water supplies, ecosystems and food production; causing economic losses, famines and environmental degradation. Recent devastating droughts include the 2017-2018 drought over the Horn of Africa with more than 16.3 million people in need of humanitarian aid. In California, the prolonged 2011-2016 drought caused more than $5.5 billion economic losses in the agricultural sector. At this moment, Europe is experiencing the worst drought in recent memory.

To better predict and respond to current and future drought impacts it is important to comprehend how droughts originate, develop, and propagate. Yet, only few studies have analysed how droughts evolve simultaneously in space and time. So far, none of them have done this at the global scale and with a future perspective.

To address this issue, we analyse the spatiotemporal dynamics of soil moisture drought and the impacts of climate change on its drought dynamics. Within this contribution, we present the first insights of our study and discuss what the implications of putting a focus on spatiotemporal drought cluster characteristics may be for the identification of current and future drought hotspots: crucial information for the prioritization of disaster risk reduction and risk financing.

In our study, we used the latest CMIP5 simulations applied in four state-of-the-art global hydrological models (H08, LPJmL, WaterGAP, PCR-GLOBWB) under the ISIMIP2b framework to evaluate the occurrence of soil moisture droughts. Soil moisture drought conditions are estimated with a variable monthly percentile threshold while clustering techniques are applied to identify drought clusters and to track how these clusters evolve through space and time. Five spatiotemporal cluster characteristics are evaluated and compared across regions: area; duration; displacement; intensity; and severity. For each location, we estimate, moreover, the density of tracks, the lead time before a drought cluster arrives; and the distance from the location of onset. By comparing the results for the low (rcp2.6) and medium (rcp6.0) emission scenarios to current conditions, we evaluate how climate change impacts the spatiotemporal soil moisture drought dynamics and where drought hotspots may emerge towards the future.