Cai et al. (2026) Interactive effects of severity and duration of compound dry–hot events on vegetation resistance time and recovery time in China

Identification

• Journal: Ecological Informatics
• Year: 2026
• Date: 2026-03-21
• Authors: Yunfei Cai, Anning Huang, Ying Huang, Wei Zhao, Xinsheng Zhu
• DOI: 10.1016/j.ecoinf.2026.103729

Research Groups

• School of Atmospheric Sciences, Nanjing University, Nanjing, China
• Nanjing Institute of Environmental Sciences, Ministry of Ecology and Environment of the People's Republic of China, Nanjing, China

Short Summary

This study developed a novel framework to systematically detect compound dry-hot events (CDHEs) and quantified their impacts on vegetation resistance and recovery times across China from 1982 to 2022. It found increasing trends in CDHE occurrences, severity, and duration, with severe and prolonged events significantly shortening forest resistance time while extending recovery time, and identified precipitation as the dominant driver of these temporal responses.

Objective

• To characterize the spatiotemporal patterns and trends of compound dry-hot events (CDHEs) affecting vegetation in China.
• To analyze how vegetation resistance time and recovery time vary across different climate zones and vegetation types in China.
• To investigate the interactive effects of CDHE severity and duration on vegetation resistance time and recovery time.

Study Configuration

• Spatial Scale: China, with gridded data resampled to 0.0833° (approximately 10 km) resolution. Analysis focused on vegetated areas, classified into four climate zones (arid, semiarid, sub-humid, humid) and four vegetation types (forest, grassland, shrubland, cropland).
• Temporal Scale: Growing seasons (April to October) from 1982 to 2022 (41 years). Data used monthly and semi-monthly (~15 days) temporal resolutions.

Methodology and Data

• Models used:
  • Compound Dry-Hot Index (CDHI = SPEI + SSMI - STI) for CDHE detection.
  • Standardized Precipitation Evapotranspiration Index (SPEI), Standardized Soil Moisture Index (SSMI), and Standardized Temperature Index (STI) as component indices.
  • Vine Copula framework (C-vine structure) for CDHI validation.
  • Least-squares detrending method for time series analysis.
  • Maximum value composite method for monthly NDVI aggregation.
  • XGBoost-SHAP method for identifying dominant drivers of resistance and recovery times.
• Data sources:
  • Near-surface (2 m) air temperature and soil moisture (0–7 cm, 7–28 cm, 28–100 cm depths) from ERA5-land reanalysis dataset (0.1° horizontal resolution).
  • Standardized Precipitation Evapotranspiration Index (SPEI) from the CHM_drought dataset (0.1° horizontal resolution).
  • Normalized Difference Vegetation Index (NDVI) from the GIMMS NDVI Version 1.2 dataset (0.0833° spatial resolution, semi-monthly).
  • Climate zone classification data from the National Earth Observation Data Center.
  • Vegetation distribution from the annual Chinese Land Cover Dataset (CLCD) (30 m spatial resolution).
  • Digital Elevation Model (DEM), population density (PD), and ecosystem service value (ESV) data from the Resource and Environment Science and Data Platform.

Main Results

• From 1982 to 2022, CDHE occurrences, severity, and duration affecting vegetation in China showed increasing trends across most regions, with significant hotspots in northern Xinjiang, the Loess Plateau, and northeastern Inner Mongolia. The annual occurrence rate increased by 0.03 events per decade, severity by 0.75 per decade, and average duration by 0.13 months per decade.
• Vegetation resistance exhibited a declining trend (−0.27 decade⁻¹), while resilience showed an increasing trend (0.009 decade⁻¹). Extreme CDHEs led to stronger resistance but lower resilience, whereas long-duration CDHEs resulted in both lower resistance and resilience.
• Vegetation resistance time and recovery time progressively increased from arid regions (mean 3.63 months and 3.94 months, respectively) to humid areas (mean 4.06 months and 4.49 months, respectively).
• Grasslands exhibited significantly shorter resistance time (average 3.54 months) and recovery time (average 3.95 months) compared to forests (average 4.18 months and 4.49 months), reflecting grasslands' higher resilience but lower resistance.
• The combined effects of severe severity and prolonged duration (≥3 months) CDHEs significantly altered vegetation responses: forest resistance time shortened by 2.2 months, and recovery time extended by 1.3 months, relative to mild and short-duration events.
• Precipitation was identified as the dominant driver of both vegetation resistance time (mean |SHAP| = 0.420) and recovery time (mean |SHAP| = 0.412) under CDHEs.

Contributions

• Developed a novel and comprehensive framework integrating multiple hydro-meteorological variables (precipitation, potential evapotranspiration, soil moisture, surface air temperature) to systematically detect CDHEs and quantify their impacts on vegetation resistance time and recovery time at the growing-season scale.
• Provided the first systematic, spatially comprehensive analysis of how vegetation resistance time and recovery time respond to the interactive effects of CDHE severity and duration across diverse vegetation types and climatic zones in China.
• Identified the dominant climatic drivers of vegetation resistance and recovery times under CDHEs using an advanced machine learning approach (XGBoost-SHAP), highlighting the critical role of precipitation.
• Offered spatially explicit scientific insights for ecological risk assessment, vegetation restoration planning, and climate-adaptive ecosystem management in China, particularly for vulnerable regions and vegetation types.

Funding

• National Natural Science Foundation of China (Grant 42375157)
• ‘GeoX’ Interdisciplinary Project of Frontiers Science Center for Critical Earth Material Cycling (Grant No. 20250301)
• CAS “Light of West China” Program (E129030101)
• Jiangsu University “Blue Project” outstanding young teachers training object
• Jiangsu Collaborative Innovation Center for Climate Change

Citation

@article{Cai2026Interactive,
  author = {Cai, Yunfei and Huang, Anning and Huang, Ying and Zhao, Wei and Zhu, Xinsheng},
  title = {Interactive effects of severity and duration of compound dry–hot events on vegetation resistance time and recovery time in China},
  journal = {Ecological Informatics},
  year = {2026},
  doi = {10.1016/j.ecoinf.2026.103729},
  url = {https://doi.org/10.1016/j.ecoinf.2026.103729}
}