Anthropogenic and atmospheric variability intensifies flash drought episodes in South Asia

Droughts, among the most complex extreme weather events, are influenced by both natural variability and human activities. While traditional droughts develop gradually, flash droughts (FDs) are characterized by their rapid onset and intensification over short periods (weeks to months), posing significant threats to agriculture, water resources, and ecosystems. Recent severe FDs in various regions globally have highlighted the urgent need for improved understanding and prediction capabilities. The multivariate nature of FDs makes their impacts compounding and necessitates a multifaceted approach for risk assessment and mitigation. Existing research, while highlighting FD occurrences and consequences in several hotspots, often focuses on meteorological drivers while neglecting the critical roles of large-scale atmospheric variability and anthropogenic climate change. This study addresses these gaps by employing a multivariate approach to investigate FD evolution characteristics in South Asia (SA), a region highly vulnerable to drought, and to quantify the contribution of atmospheric circulation and human-induced climate change to FD intensification.

Previous studies have examined flash droughts in various global regions, noting their sudden onset and rapid intensification. Research on FDs in Australia, South China, the central United States, and other areas has shown their significant impact on agriculture, water resources, and ecosystems. However, most research has focused on the meteorological drivers of FDs and their consequences. Fewer studies have examined the influence of large-scale atmospheric variability, such as land-atmosphere interactions and monsoon processes, on FD onset and intensity. The role of anthropogenic climate change in exacerbating FD risk in densely populated regions like South Asia has been largely unexplored. This study aims to address these knowledge gaps by providing a more comprehensive analysis of FD evolution in South Asia, considering the combined influence of atmospheric circulation patterns and human-induced climate change.

This study employs a multivariate probabilistic approach to analyze flash drought (FD) characteristics in South Asia (SA) from 1979 to 2021. Data sources include hourly root-zone soil moisture from ERA5 reanalysis (covering three layers: 0-7 cm, 7-28 cm, and 28-100 cm), and daily precipitation, surface air temperature, and evapotranspiration. To investigate the physical mechanisms driving FDs, monthly data of 500-hPa geopotential height, 850-hPa wind, 2 m temperature, sea surface temperature (SST), specific humidity, mean sea level pressure, cloud liquid water content, and 500-hPa vertical velocity were obtained from ERA5. Attribution analysis utilized monthly full-column soil moisture, temperature, precipitation, evapotranspiration, and SST from 10 CMIP6 GCMs under both all forcings (ALL) and natural forcings only (NAT) scenarios. FD events were identified based on a 5-day pentad mean soil moisture percentile, defining onset as a decline from >40th to <20th percentile with a rate exceeding 5% per pentad. FD features (frequency, severity, duration) were then calculated. A modified Mann-Kendall trend test analyzed trends in FD severity, frequency, and duration. A bivariate copula framework (using Gaussian, Frank, and Gumbel copulas) quantified the joint return periods (JRPs) of FD severity and duration. Fractional Attributable Risk (FAR) analysis, comparing ALL and NAT CMIP6 simulations, assessed the influence of anthropogenic climate change on FD onset speed. Spatial distributions of FAR were mapped and statistically analyzed using student's t-test, Wilcoxon rank-sum test, and Kolmogorov-Smirnov test.

The study revealed that flash droughts (FDs) in South Asia are more frequent and intense during the crop season (spring-summer transition and summer), particularly affecting central India, western Pakistan, and eastern Afghanistan. A modified Mann-Kendall trend test showed statistically significant increasing trends in FD frequency and severity in these regions, while Nepal, Bhutan, Bangladesh, and Sri Lanka showed decreasing trends. The median area affected by FDs was considerably larger during the spring-summer transition (24%) than the summer season (17%). Analysis of hydrometeorological anomalies indicated that higher-than-usual temperatures, reduced precipitation, high evapotranspiration, and low soil moisture often coincided with a high percentage of FD-affected landmasses. The bivariate copula analysis showed that the joint return period (JRP) of FD events was shorter for the spring-summer season, indicating more frequent severe and longer-lasting events. The most severe FD event in the spring-summer transition had a JRP of 50-100 years, lasting 70 days with a severity of 25.5, while the most severe summer event had a JRP of 50-100 years, lasting 50 days with a severity of 15.5. Attribution analysis using CMIP6 models revealed a substantial anthropogenic influence on the intensification of spring-summer FDs. The median fraction of attributable risk (FAR) was 60%, 80%, and 90% for Afghanistan, Pakistan, and India respectively. Analysis of atmospheric circulation patterns showed a dipolar pressure system during the spring-summer transition, leading to precipitation deficits, while the summer season was characterized by westward-migrating low-pressure systems and negative vertical velocity anomalies. These findings suggest that large-scale atmospheric circulation patterns play a key role in FD development.

This study's findings highlight the significant impact of both large-scale atmospheric variability and anthropogenic climate change on the intensification and increased frequency of flash droughts in South Asia. The results confirm the importance of considering both meteorological factors and human-induced climate change when assessing and mitigating drought risk. The increased frequency and severity of FDs during the crucial growing season underscore the significant threat to agriculture and food security in the region. The findings from the attribution analysis strongly suggest that climate change mitigation strategies are essential to reduce the future risk of FDs. The identified atmospheric circulation patterns provide valuable insights into the underlying mechanisms driving FD development, highlighting the need for improved predictive models and early warning systems. Further research could explore the combined impacts of drought and heat extremes, refining the understanding of FD-related risks and informing effective adaptation measures.

This study demonstrates the increasing frequency, severity, and spatial extent of flash droughts in South Asia, significantly impacted by both atmospheric circulation patterns and anthropogenic climate change. The high fractional attributable risk highlights the urgent need for climate change mitigation and adaptation strategies. Future research should focus on improving early warning systems, developing more sophisticated drought prediction models that incorporate the complex interactions between atmospheric circulation and land-surface processes, and exploring the cascading effects of droughts with other extreme events.

The study relies on reanalysis data and CMIP6 model outputs, which have inherent uncertainties. The attribution analysis focuses on the onset speed of FDs and may not fully capture the complexity of FD evolution. Further, the study primarily concentrates on the mean state of soil moisture during FD events, without detailed analysis of soil moisture variability across the vertical column. The choice of specific percentiles (20th and 40th) for defining FD onset could also influence the results.