Zhao et al. (2025) A GNSS-based standardized index for near-real-time monitoring of the spatiotemporal evolution of droughts

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2025
• Date: 2025-11-08
• Authors: Qian Zhao, Lina Su, Ke Shi, Rongzi Chai, Ziang Zhu
• DOI: 10.1016/j.ejrh.2025.102906

Research Groups

• Institute of Earthquake Forecasting, China Earthquake Administration, Beijing, China
• Shaanxi Earthquake Agency, Xi’an, China
• Yunnan Provincial Basic Surveying and Mapping Technology Center, Kunming, China

Short Summary

This study introduces a novel GNSS-based Standardized Terrestrial Water Storage Anomaly Index (GNSS-STWSAI) that leverages daily vertical displacements from a dense GNSS network and a Gaussian Mixture Model to quantify hydrological droughts with high spatiotemporal resolution. The index provides near-real-time monitoring, revealing the lagged response of hydrological drought to meteorological forcing and enhancing understanding of monsoon-driven drought dynamics in Yunnan Province.

Objective

• To introduce and evaluate a novel GNSS-based Standardized Terrestrial Water Storage Anomaly Index (GNSS-STWSAI) for near-real-time, high spatiotemporal resolution monitoring of hydrological droughts in Yunnan Province, China.
• To quantify hydrological droughts using daily vertical displacements from a dense GNSS network and objectively classify drought severity using a Gaussian Mixture Model.
• To evaluate the GNSS-STWSAI against conventional meteorological indices and analyze drought events in relation to atmospheric circulation patterns influencing monsoon moisture transport.

Study Configuration

• Spatial Scale: Yunnan Province, China, covering an area of approximately 394,000 square kilometers. GNSS-derived Terrestrial Water Storage (TWS) variations are inverted on a 0.7° × 0.7° grid.
• Temporal Scale: January 2017 to December 2022 (6 years). GNSS-STWSAI is generated at daily resolution.

Methodology and Data

• Models used:
  • GNSS-based Standardized Terrestrial Water Storage Anomaly Index (GNSS-STWSAI)
  • Gaussian Mixture Model (GMM) for probability distribution fitting and drought severity classification
  • Elastic loading theory (Farrell, 1972) for inverting GNSS vertical displacements to TWS variations
  • Least-squares inversion with Laplacian regularization for TWS inversion
  • Circular disk loading model for mass changes inversion
  • Locally Estimated Scatterplot Smoothing (LOESS) for seasonal and trend decomposition of TWS time series
  • Bayesian Information Criterion (BIC) for optimal GMM component selection
  • Cross-wavelet and wavelet coherence analyses for dynamic relationships between drought indices
• Data sources:
  • GNSS data: Daily vertical displacements from 147 continuous stations (55 from Crustal Movement Observation Network of China (CMONOC), 92 from Yunnan Provincial Basic Surveying and Mapping Technology Center (YNCORS)).
  • Groundwater Wells Data: Daily water levels from 19 observation wells (China Earthquake Networks Center).
  • GLDAS data: Monthly Noah 2.1 model, 0.25° × 0.25° resolution (NASA).
  • GRACE data: Monthly CSR RL06 Mascon product, 0.25° × 0.25° resolution (University of Texas at Austin’s Center for Space Research).
  • Precipitation data: Daily gridded, 0.1° × 0.1° resolution (National Tibetan Plateau Data Center).
  • Meteorological drought indices:
    • Self-calibrated Palmer Drought Severity Index (scPDSI): Monthly, 0.5° × 0.5° resolution (University of East Anglia).
    • Standardized Precipitation Evapotranspiration Index (SPEI): Monthly, 0.5° × 0.5° resolution (University of East Anglia).
  • Climate indices: Monthly El Niño and Indian Ocean Dipole (IOD) indices (Australian Bureau of Meteorology).
  • Reanalysis data: Monthly ERA5-derived 850 hPa wind and sea level pressure (SLP) anomalies, 0.25° × 0.25° resolution (Copernicus Climate Data Store/ECMWF).
  • Official water resource records: Yunnan Provincial Water Resources Bulletin.

Main Results

• The GNSS-STWSAI, leveraging a dense GNSS network, provides daily, near-real-time drought monitoring with a spatial resolution of approximately 0.7°.
• The Gaussian Mixture Model (GMM) accurately fits the multimodal and skewed distribution of Terrestrial Water Storage Anomalies (TWSA) in Yunnan, achieving the lowest average RMSE (7.20) compared to Normal, GEV, Loglogistic, and Gamma distributions.
• GNSS-derived TWS exhibits higher average annual amplitudes (138.03 mm) and greater sensitivity to TWS changes compared to GRACE-TWS (112.70 mm) and GLDAS (90.54 mm).
• GNSS-TWS peaks approximately 48 days after precipitation, 9 days after GLDAS, and leads GRACE-TWS by a similar amount, indicating a more immediate response to TWS changes.
• Persistent and widespread hydrological drought conditions were identified in Yunnan Province from 2019 to 2022 by GNSS-STWSAI, consistent with observed precipitation deficits, reduced groundwater levels, and official total water resource records.
• Drought propagation analysis reveals a characteristic lag, with meteorological drought (SPEI) appearing first, followed by agricultural drought (scPDSI), and then hydrological drought (GNSS-STWSAI and GRACE-DSI).
• GNSS-STWSAI shows improved correlation coefficients with SPEI (approximately 19% higher) and scPDSI (approximately 67% higher) compared to GRACE-DSI, demonstrating its effectiveness in reflecting hydrological responses to meteorological drought.
• The severe 2019 drought was associated with a strong El Niño and positive Indian Ocean Dipole (IOD), while the 2022 drought coincided with a La Niña event and negative IOD, both modulating large-scale atmospheric circulation and moisture transport.

Contributions

• Development of a novel GNSS-based Standardized Terrestrial Water Storage Anomaly Index (GNSS-STWSAI) for near-real-time, high spatiotemporal resolution hydrological drought monitoring, utilizing daily GNSS vertical displacements.
• Introduction of a Gaussian Mixture Model (GMM) for objective and statistically robust classification of drought severity from GNSS-derived TWSA, effectively addressing the multimodal and skewed nature of TWS distributions.
• Demonstration of the enhanced capability of a dense GNSS network (147 stations) to capture finer-scale drought features and provide more reliable characterization compared to sparser networks.
• Detailed characterization of drought onset, propagation, and recovery in a topographically complex region, revealing the lagged response of hydrological drought to meteorological forcing.
• Provision of a valuable observational constraint for drought monitoring, risk assessment, and water resource management, contributing to a more comprehensive understanding of drought propagation processes.

Funding

• National Natural Science Foundation of China (grant numbers 42374008 and 42304004).

Citation

@article{Zhao2025GNSSbased,
  author = {Zhao, Qian and Su, Lina and Shi, Ke and Chai, Rongzi and Zhu, Ziang},
  title = {A GNSS-based standardized index for near-real-time monitoring of the spatiotemporal evolution of droughts},
  journal = {Journal of Hydrology Regional Studies},
  year = {2025},
  doi = {10.1016/j.ejrh.2025.102906},
  url = {https://doi.org/10.1016/j.ejrh.2025.102906}
}