Li et al. (2026) Reconstruction and evolution analysis of long-term terrestrial water storage anomalies in Xinjiang based on time series decomposition and multi-model coupling

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-07-13
• Authors: Zilong Li, Zhenxia Mu, J S Liu, Xiaoyan Qiu
• DOI: 10.1016/j.jhydrol.2026.136007

Research Groups

• College of Water Conservancy and Civil Engineering, Xinjiang Agricultural University
• Xinjiang Key Laboratory of Hydraulic Engineering Security and Water Disasters Prevention

Short Summary

This study proposes a framework to reconstruct long-term Terrestrial Water Storage Anomalies (TWSA) in Xinjiang using time-series decomposition, multi-model coupling, and machine learning. The reconstructed dataset spans from 1961 to 2022 and exhibits high consistency with existing records.

Objective

• Investigate the feasibility of reconstructing long-term TWSA in Xinjiang using a novel framework combining time-series decomposition, multi-model coupling, and machine learning.

Study Configuration

• Spatial Scale: Regional (Xinjiang)
• Temporal Scale: Monthly, from 1961 to 2022

Methodology and Data

• Models used:
  • VIC hydrological model
  • Machine learning algorithms for time-series decomposition and reconstruction
• Data sources:
  • GRACE/GRACE-FO satellite data
  • JPL and CSR Mascon products
  • Multi-source observation data (groundwater-level observations, CNRD v1.0, TerraClimate runoff depth)

Main Results

• The reconstructed TWSA dataset exhibits high consistency with existing records, with approximately 70% of grids having a correlation coefficient > 0.6.
• Independent validation based on groundwater-level observations shows that the standardized reconstructed groundwater storage anomalies are positively correlated with standardized groundwater-level anomalies in all validation grids (CC = 0.23–0.81).
• The reconstructed TWSA dataset reveals a predominantly downward trend from 1961 to 2022, with significant spatial heterogeneity.

Contributions

This study provides a novel framework for reconstructing long-term TWSA and contributes to the understanding of hydrological changes in Xinjiang by providing a reliable and continuous dataset spanning over six decades.

Funding

• This research was funded by the Xinjiang Key Laboratory of Hydraulic Engineering Security and Water Disasters Prevention (Grant No. XJLHESWD2023001).
• The study also received support from the College of Water Conservancy and Civil Engineering, Xinjiang Agricultural University (Grant No. XJAU-CWCCE2022002).

Citation

@article{Li2026Reconstruction,
  author = {Li, Zilong and Mu, Zhenxia and Liu, J S and Qiu, Xiaoyan},
  title = {Reconstruction and evolution analysis of long-term terrestrial water storage anomalies in Xinjiang based on time series decomposition and multi-model coupling},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.136007},
  url = {https://doi.org/10.1016/j.jhydrol.2026.136007}
}