Liu et al. (2025) The spatiotemporal characteristics of extreme drought events in China from 1961 to 2022 via a copula function

Identification

• Journal: Weather and Climate Extremes
• Year: 2025
• Date: 2025-10-17
• Authors: Tingting Liu, Xiufang Zhu, Hongquan Sun, Mingxiu Tang
• DOI: 10.1016/j.wace.2025.100815

Research Groups

• State Key Laboratory of Remote Sensing and Digital Earth, Faculty of Geographical Science, Beijing Normal University, Beijing, China
• Key Laboratory of Environmental Change and Natural Disaster, Ministry of Education, Beijing Normal University, Beijing, China
• National Institute of Natural Hazards, Ministry of Emergency Management of China, Beijing, China

Short Summary

This study systematically analyzed extreme drought events in China from 1961 to 2022 using the Standardized Precipitation Index (SPI) and copula functions, revealing significant spatiotemporal variations in drought trends and severity across different regions. It highlights increased drought severity in Northeast and South China, while Northwest China and the Qinghai–Tibet Plateau experienced increased humidity.

Objective

• To identify drought events using run theory, calculate drought-related variables, and analyze their spatiotemporal distribution characteristics.
• To define extreme drought events (EDEs) as those with a joint probability of drought severity and duration less than 5%, calculate grid-by-grid thresholds for EDEs via copula functions, and assess the spatial heterogeneity of these thresholds.
• To compare the frequency, duration, and severity of EDEs in two periods (1961–1991 and 1992–2022) to discuss their spatiotemporal changes.
• To explore the dominant factors (duration or severity) of EDEs in different regions.

Study Configuration

• Spatial Scale: China, analyzed at a 0.25° × 0.25° grid resolution, and further divided into seven natural subregions.
• Temporal Scale: 1961 to 2022 (62 years), with monthly data. Comparisons were made between two periods: 1961–1991 and 1992–2022.

Methodology and Data

• Models used:
  • Standardized Precipitation Index (SPI) at a 3-month timescale.
  • Run theory (Yevjevich, 1967) for drought event identification (threshold: SPI < -0.5).
  • Copula functions (bivariate normal, t-copula, and Archimedean functions: Frank, Clayton, Gumbel) for joint probability modeling of drought duration and severity, and for determining extreme drought thresholds (joint exceedance probability < 5%).
  • Theil-Sen estimator for trend calculation.
  • Mann-Kendall test for statistical significance of trends.
  • Akaike Information Criterion (AIC) and Anderson-Darling goodness-of-fit test for optimal distribution selection.
  • Wilcoxon signed-rank test for comparing characteristic attributes between time periods.
• Data sources:
  • Monthly precipitation data from 1961 to 2022 from the CN05.1 grid dataset.
  • The CN05.1 dataset has a resolution of 0.25° × 0.25° and is derived from interpolation of observations from 2,416 national meteorological stations in China.

Main Results

• The SPI exhibited significant spatiotemporal heterogeneity across China from 1961 to 2022. Northwest China (Subregion 1) and the Qinghai‒Tibet Plateau (Subregion 6) showed increasing SPI trends (more humid), while Northeast China (Subregion 3) and South China (Subregion 7) showed decreasing SPI trends (increased drought severity).
• Drought events in western Northwest China (Subregion 1) and northern Qinghai‒Tibet Plateau (Subregion 6) had longer average durations (e.g., 3.93 months in Subregion 1) and greater average severities (e.g., maximum severity up to 10.52 in Subregion 1), exceeding national averages (3.47 months and 9.26).
• Extreme drought thresholds, determined by copula functions, showed clear spatial heterogeneity, with higher severity/duration thresholds in arid Northwest China (Subregion 1) and the Qinghai‒Tibet Plateau (Subregion 6).
• The frequency, duration, and severity of extreme drought events increased significantly in central and southern China, particularly in the subtropical humid region (Subregion 5), where 47% of grids experienced an increase in total occurrences, 54% in total months of occurrence, 64% in average severity, and 62% in maximum severity.
• The number of extreme droughts jointly caused by both duration and severity was greater than those dominated by either factor alone. While extreme droughts became less frequent nationally, their severity increased in specific regions like South China.

Contributions

• Provided a comprehensive and systematic assessment of extreme drought events in China from 1961 to 2022 by integrating the Standardized Precipitation Index (SPI) and copula functions.
• Introduced a novel definition of extreme drought events based on a joint exceedance probability of drought severity and duration (<5%), enabling dynamic, regionally specific thresholds that more accurately reflect spatial heterogeneity compared to traditional fixed-threshold methods.
• Analyzed the spatiotemporal changes in extreme drought characteristics over two distinct periods (1961–1991 and 1992–2022) and identified the dominant factors (duration/severity) contributing to extreme droughts in different regions.
• Offered a scientific basis for improved drought monitoring, risk assessment, and the development of more effective drought management strategies in China.

Funding

• National Key Research and Development Program of China (Grant No. 2021YFB3901201)
• National Natural Science Foundation of China (Grant No. 42077436)

Citation

@article{Liu2025spatiotemporal,
  author = {Liu, Tingting and Zhu, Xiufang and Sun, Hongquan and Tang, Mingxiu},
  title = {The spatiotemporal characteristics of extreme drought events in China from 1961 to 2022 via a copula function},
  journal = {Weather and Climate Extremes},
  year = {2025},
  doi = {10.1016/j.wace.2025.100815},
  url = {https://doi.org/10.1016/j.wace.2025.100815}
}