Global distribution, trends, and drivers of flash drought occurrence

Flash droughts, characterized by rapid land surface desiccation, pose significant challenges to agriculture, economies, and societies. The speed of their intensification leads to substantial economic damage through reduced crop yields and livestock production impacts. Furthermore, rapid drought intensification significantly impacts ecosystems by increasing evaporative stress, contributing to compound extreme events like wildfires, water resource depletion, reduced air quality, and decreased food security. The sub-seasonal nature of flash droughts presents significant challenges for prediction. While regional studies exist on flash drought occurrence in the United States, Brazil, southern Africa, Spain, western Russia, and Australia, a crucial gap in knowledge remains concerning the global distribution, seasonal frequency, trends, and underlying drivers of these events. This study aims to address this gap by quantifying the global distribution of flash drought events between 1980 and 2015, analyzing their seasonal frequencies, determining trends in occurrence, and identifying the primary driving mechanisms behind their development. This global perspective builds upon previous regional analyses and provides a comprehensive understanding of flash drought susceptibility worldwide.

Previous research has focused on improving the detection, evaluation, and monitoring of flash drought through various approaches. These include analyzing subsurface soil moisture, atmospheric evaporative demand, evaporative stress using evapotranspiration (ET) and potential evapotranspiration (PET), and impact-based approaches. Studies have identified rapid drought intensification across various regions, including the United States, Brazil, southern Africa, Spain, western Russia, and Australia. However, a comprehensive global assessment of flash drought distribution, seasonal patterns, trends, and drivers was lacking until this current study.

This study utilized four global reanalysis datasets (MERRA, MERRA-2, ERA-Interim, and ERA5) spanning 1980–2015. Evapotranspiration (ET) and potential evapotranspiration (PET) data from these datasets were used to calculate the standardized evaporative stress ratio (SESR), representing the overall evaporative stress on the environment. SESR, similar to the evaporative stress index (ESI), is calculated as the standardized ratio of ET to PET. A comprehensive flash drought identification methodology was applied to SESR, incorporating multiple criteria related to rapid drought intensification and impact. This methodology considers the rate of change in SESR across pentads (five-day periods), alongside threshold values of SESR to define both the drought impact and the rate of intensification. The regions were defined using a 0.5° × 0.5° grid. Hotspots were identified using the composite analysis from all four data sets which were then bilinearly interpolated. Regions with high frequency of flash drought were identified for further temporal analysis, including seasonal patterns and trends. To assess the contribution of precipitation deficits and increased evaporative demand to flash drought development, standardized precipitation index (SPI) and standardized PET anomalies were calculated for each flash drought event within the 15 study regions. The Mann–Kendall test was used to assess statistically significant trends in yearly flash drought coverage. Regions that were too arid or cold were masked from the analysis. Aridity was determined using the aridity index (P/PET), excluding regions where the average annual aridity index was below 0.2 or where the average daily PET was <1 mm per day during the growing season.

The analysis revealed several key findings: 1. **Global Flash Drought Hotspots:** The study identified major flash drought hotspots primarily in the tropics and subtropics, including large parts of Brazil, the Sahel, the Great Rift Valley, and India. These regions experienced flash drought in 30–40% of the years studied. Mid-latitude hotspots were observed in the central United States, southwestern Russia, and northeastern China, with flash drought frequencies between 10–20%. The four major hotspots (Brazil, Sahel, Great Rift Valley, India) showed strong agreement among the four reanalysis datasets. 2. **Temporal Characteristics:** The seasonal patterns of flash drought varied regionally. Northern Hemisphere mid-latitude regions (excluding Iberian Peninsula and Asia Minor) showed peak flash drought frequency between May and July. Southern Hemisphere mid-latitudes exhibited different seasonal patterns. Tropical and subtropical regions demonstrated seasonal patterns influenced by their respective hemispheres, with peaks generally occurring during the growing seasons. Of the 15 study regions, eight were both regional flash drought maxima and significant agricultural areas. 3. **Trends in Flash Drought Coverage:** Six of the 15 study regions displayed a statistically significant increasing trend (p < 0.1) in flash drought coverage between 1980–2015, while three showed a significant decreasing trend. The magnitude of these trends varied significantly across regions. 4. **Drivers of Flash Drought Development:** On average across all study regions, large positive PET anomalies occurred during flash drought events at a similar rate to large precipitation deficits (33% vs 31%). The relative contribution of each driver varied regionally; for instance, European regions exhibited a higher frequency of PET anomalies, while American regions showed precipitation deficits as the main driver. Approximately 20% of flash drought events experienced both large precipitation deficits and PET anomalies concurrently. 5. **Relationship with Soil Moisture:** Analysis showed that soil moisture was depleted to the 20th percentile or lower in 11 out of 15 study regions during flash drought events, indicating that evaporative stress effectively captures rapid drought intensification leading to soil moisture depletion.

The results provide a comprehensive overview of global flash drought patterns. The identified hotspots align with previous research, though some regional variations exist, potentially due to differences in methodologies and variables. The findings highlight the almost equal contributions of precipitation deficits and enhanced evaporative demand in flash drought development, emphasizing the importance of considering both factors in future studies and predictions. The study also revealed regional variations in the dominant driver of flash drought, with PET anomalies prevailing in some areas and precipitation deficits in others. The observed trends highlight the potential influence of climate change on flash drought occurrence. Increasing trends in PET, as expected in a warming climate, could further exacerbate flash drought risk in susceptible regions. The relationship between flash droughts, soil moisture depletion, and agricultural impacts is strongly emphasized. The study acknowledges limitations due to using evaporative stress, as opposed to other hydrological variables.

This study provides a comprehensive analysis of global flash drought occurrence, trends, and drivers using a standardized evaporative stress ratio. The results reveal global hotspots, seasonal patterns, and significant trends in flash drought frequency and extent, highlighting the importance of both precipitation deficits and enhanced evaporative demand in flash drought development. Future research should focus on refining regional driver analysis, investigating the complex socioeconomic impacts of flash drought, and advancing sub-seasonal prediction capabilities to enhance resilience in vulnerable regions.

The study relies on evaporative stress as the primary indicator of flash drought, potentially limiting the generalizability of findings to other hydrological variables or identification methods. The relatively short study period (1980-2015) may affect the interpretation of long-term trends. The analysis does not consider all factors influencing flash drought development. Regional variations in the response to the primary drivers highlight that further research is needed at a local and regional level.