ÌÇÐÄÊÓƵ

The average longest periods of drought could be ten days longer by the end of the century than previously predicted by climate models, according to research in Nature. The findings suggest that the hazards droughts pose to societies and ecosystems in the coming decades may be greater than expected.

Climate models project an increase of dry extremes in many world regions, but uncertainties make it difficult to implement effective adaptation strategies to minimize the environmental impact of droughts.

Irina Petrova, Diego Miralles, and colleagues investigated potential biases in drought projections made by a range of climate models under both moderate- and high-emissions scenarios (the IPCC's SSP2-4.5 and SSP5-8.5 scenarios). They then calibrated these projections with historical observations of the longest number of consecutive dry days each year—known as the longest annual dry spell—between 1998 and 2018.

The authors estimate that increases in the longest annual dry spell length predicted by the end of the century by the calibrated models could be 42–44% higher on average than those predicted by the non-calibrated models under both emissions scenarios. This indicates that the average global longest annual dry spell could be ten days longer than previously expected by 2080 to 2100.

The authors also found that the increase in the longest annual dry spell length predicted by the calibrated models in North America and southern Africa and Madagascar is approximately twice as large as that predicted by the non-calibrated models.

However, they note that future decreases in the longest annual dry spell length predicted by the calibrated models in Central–East Asia may be almost three times greater than those predicted by non-calibrated models under both emissions scenarios. This finding suggests an increased risk of more frequent rainfall and flooding in some regions.

The findings emphasize the need for a reassessment of drought risks around the world and highlight the importance of correcting existing biases in climate models to increase confidence in their projections.