New UGent Study published in Nature Reveals Overlooked Future Drought Hazards to Society and Ecosystems

Using an innovative approach to correct future climate models predictions with real-world observations, researchers have found that the expected increase in the duration of annual dry periods due to climate change will be significantly higher than climate models project. As a result, strategic planning for global water resources, agriculture, and ecosystems may be fundamentally flawed in many regions of the world.

Key Findings

• By the end of this century, the duration of longest annual dry periods could be, on average, 10 days longer than expected.
• Future increases in the dry periods are most underrated in southwestern North America, Southern Africa and parts of the Mediterranean, raising concerns about water scarcity, food security, and human safety in these vulnerable regions.
• The same systematic errors remain in the latest generation of climate models that preclude effective drought risk assessment for human and natural systems. 

Low confidence in magnitude and timing of future drought predictions

Understanding the severity and duration of droughts in the near future strongly depends on the ability of climate models to accurately simulate the complex interactions between land, atmosphere, and oceans. Enhancing the reliability of the future predictions for drought extremes remains a critical challenge for the scientific community today. At the same time, the media and public discussions increasingly highlight the question: 'How likely is it that my family and I will face longer, hotter dry periods in the next 20 years, and how severe might these conditions become?'

“At present, confidence in the magnitude and timing of climate model-projected increases in drought and dryness remains low, leaving societies largely unprepared” - says Dr. Petrova. To reduce this uncertainty, Dr. Petrova and Prof. Diego Miralles from Ghent University, together with an international team of researchers, conducted a study aiming to improve future predictions for prolonged dry periods on global and regional scales. By analysing future projections from 55 state-of-the-art climate models, researchers could pinpoint global regions where future drought risks to societies and ecosystems are likely underestimated in current climate model simulations.

Using present-day observations to correct future climate model predictions

The study employs a so-called emergent constraint (EC) approach that seeks for physically plausible relationships between past and future model uncertainties. Applying this method to an ensemble of climate models, the researchers identified global regions where past model errors in prolonged dry periods influence the magnitude of their future simulated increase in a meaningful way. “According to this discovered past-future dependency, we found that in 20-30% of global land areas, the same climate models that currently simulate longer dry periods would consistently project larger increases in their duration by the end of the century.” - explains Dr. Petrova. By demonstrating that models accurately reflecting the current climate are likely to be better at predicting future changes, and by incorporating multiple real-world observations into the revealed past–future relationship, the researchers were able to calibrate future model predictions of prolonged dry periods. These new, calibrated model projections did not only reduce the uncertainty of future predictions by up to 26% globally for the most severe climate change scenario, but further revealed that the future increase in the duration of prolonged dry periods will likely be 42-44% larger on average by year 2100 than models suggest. “These findings refine our understanding of future drought hazards, highlighting that we may face harsher droughts than previously anticipated,” says Prof. Miralles.

Future drought risks and implications for policy and adaptation 

The most severe underestimation of future projected increase in prolonged annual dry periods was identified in southern Africa, southwestern regions of the US and parts of the Mediterranean. Most of these regions already now experience water shortages and increased drought hazards on an annual basis. The study reveals that some regions of Southern Africa could experience a projected 12-day increase in dry period duration as early as 2050 — nearly 50 years sooner than previously anticipated. “These results are worrisome.” - Dr. Petrova says - “They emphasise the need for policymakers to reassess climate resilience strategies and adaptation plans, especially in regions projected to experience more severe-than-anticipated drought and dryness conditions.”

"Early and effective measures are crucial to mitigating the impacts of drought on agriculture, water supply, ecosystems, and population. Our study represents another step toward more reliable climate projections that can better guide the development of these adaptation strategies," concludes Dr. Petrova.

About the Authors

The study was conducted by researchers from the Hydro-Climate Extremes Laboratory of Ghent University, Dr. Irina Yu. Petrova, Prof. Dr. Diego G. Miralles in collaboration with scientists from Sorbonne University, Barcelona Supercomputing Centre, Pohang University of Science and Technology, Université de Toulouse, and others, demonstrating the importance and high impact of scientific collaborations between countries and institutions in addressing one of our grand challenges in climate research.

Contact

Contact Dr. Irina Petrova at [irina.petrova@ugent.be] or Prof. Dr. Diego Miralles at [diego.miralles@ugent.be] for further details.

Publication Details

The full study, "Observation-constrained projections reveal longer-than-expected dry spells," is published in Nature .