Zhong et al. (2026) Copula-based composite drought index combining precipitation, temperature, and NDVI used for drought monitoring

Identification

• Journal: Theoretical and Applied Climatology
• Year: 2026
• Date: 2026-01-30
• Authors: Zhong Zhong, Qiang Zhao, Guoqing Sang, Jianwen Xue, Jianing Wang, Feng Lin
• DOI: 10.1007/s00704-026-06033-0

Research Groups

• School of Water Conservancy and Environment, University of Jinan, Jinan, China

Short Summary

This study developed a new Composite Drought Index (CDI) using nested copula functions to integrate precipitation, temperature, and the Normalized Difference Vegetation Index (NDVI) for improved drought monitoring in Liaoning Province, China. The CDI demonstrated high accuracy (over 87% hit rate) in identifying composite drought events and revealed a worsening trend of summer droughts, with over half the region projected to face more severe droughts in the future, except in spring.

Objective

• To develop and validate a novel Composite Drought Index (CDI) using nested copula functions that integrates precipitation, temperature, and NDVI for characterizing meteorological and agricultural drought in Liaoning Province, China.
• To analyze the spatio-temporal characteristics and trends of drought in Liaoning Province using the developed CDI.
• To identify the primary factors influencing the combined drought trends in Liaoning Province.

Study Configuration

• Spatial Scale: Liaoning Province, China (approximately 148,600 km²), with data resampled to a uniform spatial resolution of 0.05° × 0.05° (approximately 5.5 km).
• Temporal Scale: Monthly, seasonal, and annual scales from 2001 to 2021 for CDI construction and analysis. Some base data (CRUTSv4.06) dates back to 1901.

Methodology and Data

• Models used:
  • Nested Archimedean Copula functions (optimal: Clayton Copula) for multivariate joint distribution.
  • Kolmogorov-Smirnov test for optimal marginal distribution selection (optimal: Generalized Extreme Value distribution).
  • Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), Root Mean Square Error (RMSE) for copula function evaluation.
  • Pearson correlation analysis for index validation.
  • Sensitivity and Specificity metrics for drought detection performance.
  • Run theory for drought event characteristics (duration, severity, intensity).
  • Mann-Kendall nonparametric test and Theil-Sen median trend analysis for drought trends.
  • Livezey-Chen method and False Discovery Rate (FDR) for field significance testing.
  • Hurst exponent (R/S algorithm) for future trend prediction.
  • Geographical detector (factor and interaction detectors) for driving force analysis.
• Data sources:
  • Monthly precipitation (PRE), potential evapotranspiration (PET), and temperature (TMP): CRUTSv4.06 dataset from the Climatic Research Unit at the University of East Anglia.
  • Land surface temperature (LST): TRIMSLST dataset (1 km spatial resolution).
  • Normalized Difference Vegetation Index (NDVI): MOD13C2 vegetation index data (0.05° × 0.05° spatial resolution).
  • Monthly soil moisture (SM): TerraClimate global high-resolution climate dataset (approximately 4 km spatial resolution).
  • Digital elevation model (DEM), slope (SLP), and aspect (ASP): SRTM-90 M dataset (1 km spatial resolution).
  • Actual drought events, crop damage area, and grain yield: Meteorological Disaster Yearbook (2004–2021) and China Statistical Yearbook (2001–2021).
  • Population data: East View.
  • River network density data: National Geomatics Data Service System.

Main Results

• The Composite Drought Index (CDI) showed moderately strong correlations (R > 0.59) with SPI, SPEI, and VHI, and moderate correlations (R ≈ 0.4) with VCI, demonstrating its reliability in integrating meteorological and agricultural droughts.
• The CDI exhibited high sensitivity (all > 0.48) and very high specificity (all > 0.78) in identifying drought events, with an overall accuracy of 87% at the 3-month scale and 90% at the 12-month scale for combined meteorological and agricultural droughts.
• Spatially, southern and western mountainous areas of Liaoning Province experienced long-duration and high-severity droughts at the 3-month scale, while southwestern and southern coastal areas faced severe droughts at the 12-month scale. High drought intensity was concentrated in the northern plain.
• Temporally, summer droughts are worsening, while spring, autumn, and winter show trends of drought alleviation. However, future projections (based on Hurst index) suggest that over half of the study area, except for spring, may experience more severe droughts.
• NDVI was identified as the primary factor influencing interannual, summer, and autumn drought trends. Potential evapotranspiration (PET) was dominant in spring, and precipitation (PRE) in winter. Interactions between factors like NDVI, PRE, TMP, PET, and DEM significantly influenced the spatial heterogeneity of drought trends.

Contributions

• Proposes and validates a novel Composite Drought Index (CDI) based on nested copula functions, effectively integrating precipitation, temperature, and NDVI to provide a more comprehensive and accurate assessment of meteorological and agricultural droughts in complex hydrological regions.
• Demonstrates the CDI's superior performance in balancing meteorological and agricultural drought signals and its strong correlation with actual drought impacts on crop-affected areas and grain yields, addressing limitations of single-variable indices.
• Provides a detailed spatio-temporal analysis of drought characteristics, trends, and future persistence in Liaoning Province, offering valuable insights for regional drought management and climate change adaptation strategies.
• Identifies the key driving factors and their complex interactions influencing drought trends across different seasons and annually, enhancing the understanding of drought mechanisms in the study area.

Funding

• Shandong Provincial Natural Science Foundation [grant number ZR2024ME171]
• National Natural Science Foundation of China [grant number 41471160]

Citation

@article{Zhong2026Copulabased,
  author = {Zhong, Zhong and Zhao, Qiang and Sang, Guoqing and Xue, Jianwen and Wang, Jianing and Lin, Feng},
  title = {Copula-based composite drought index combining precipitation, temperature, and NDVI used for drought monitoring},
  journal = {Theoretical and Applied Climatology},
  year = {2026},
  doi = {10.1007/s00704-026-06033-0},
  url = {https://doi.org/10.1007/s00704-026-06033-0}
}