Fu et al. (2025) Remote sensing-based monitoring of compound drought-waterlogging stress in groundwater-sensitive agroecosystems in arid regions

Identification

• Journal: Agricultural Water Management
• Year: 2025
• Date: 2025-09-22
• Authors: Di Fu, Xin Jin, Yanxiang Jin, Xufeng Mao, Na Yao
• DOI: 10.1016/j.agwat.2025.109826

Research Groups

• School of the Geographical Science, Qinghai Normal University, China
• MOE Key Laboratory of Tibetan Plateau Land Surface Processes and Ecological Conservation, China
• Qinghai Forestry Engineering Supervision Center Co., Ltd., China

Short Summary

This study developed a remote sensing-based framework to monitor groundwater-driven compound drought-waterlogging stress in arid Groundwater-Sensitive Agroecosystems (GWSA) in Northwest China. By downscaling root-zone soil moisture data to 30 m resolution and employing novel indices, the research revealed significant spatial heterogeneity in stress patterns and identified groundwater level, precipitation, and temperature (via snowmelt) as key drivers, offering critical insights for agricultural resilience.

Objective

• To assess the effectiveness of the established remote sensing-based method for extracting compound drought-waterlogging stress in GWSA.
• To characterize the spatiotemporal patterns of compound drought-waterlogging stress in the Gahai Irrigation District GWSA.
• To identify the factors influencing the spatiotemporal changes in compound drought-waterlogging stress.

Study Configuration

• Spatial Scale: Gahai Irrigation District, Northwest China (97°20′ to 97°35′E and 37°00′ to 37°20′N), with data downscaled to 30 meters spatial resolution.
• Temporal Scale: Daily data from 2018 to 2022 for monitoring, with some ancillary data from earlier periods (e.g., soil type 1995, soil texture 2013–2014).

Methodology and Data

• Models used:
  • Random Forest (RF) algorithm for downscaling 1 km root-zone soil moisture (RZSM) to 30 m resolution.
  • Soil Moisture Condition Index (SMCI-index) for identifying drought (SMCI-index < 0.4 for ≥10 days) and waterlogging (SMCI-index > 0.6 for ≥3 days) events.
  • Double Stress Index (DSI), calculated as the harmonic mean of drought and waterlogging event frequencies, for identifying compound stress.
  • Vegetation Condition Index (VCI) and Vegetation Drought Index (VDI) for drought monitoring validation.
  • SWAT-MODFLOW model (from Jin et al., 2022) for waterlogging monitoring validation benchmark.
• Data sources:
  • In-situ observations: Precipitation, air temperature (Delingha Meteorological Station), runoff of the Bayin River (Qinghai Provincial Department of Water Resources), groundwater depth (Delingha Hydrological Station).
  • Soil moisture products: SMCI1.0-dataset (1 km daily, 100 cm depth RZSM) from National Tibetan Plateau Data Center. SMAP L4 (9 km 3-hourly RZSM, 0–100 cm). GLDAS 2.2 (0.25° daily RZSM). GLEAM 4 (0.1° daily RZSM).
  • Satellite imagery & derived products: MOD11A1/MYD11A1 (1 km daily Land Surface Temperature - LST). MOD09GA/MYD09GA (500 m daily surface reflectance for NDVI, NDWI, kNDVI, EVI, GI). ASTER GDEM (30 m for DEM, Slope, Aspect).
  • Reanalysis & other datasets: MSWEP v2 (0.1° daily precipitation). China dataset of soil properties (900 m for sand, silt, clay, porosity, soil organic matter). Spatial Distribution Data of Soil Types in China (1 km). ERA5-Land (0.1° daily soil temperature, albedo, volume of water in soil, surface/subsurface runoff, evapotranspiration).

Main Results

• A 30 m daily root-zone soil moisture (DRZSM) product for 2018–2022 was successfully generated using a Random Forest algorithm and 33 downscaling factors.
• The SMCI-index, derived from DRZSM, showed strong consistency with VCI (Probability of Detection > 80%, correlation coefficient r = 0.98) and VDI (r = 0.86) for drought monitoring. Waterlogging detection achieved a 69.27% spatial overlap with hydrological model simulations and a strong negative correlation with observed groundwater depth (r = -0.86 for maximum frequency, r = -0.80 for medium-to-high frequency areas).
• Compound Drought-Waterlogging Stress exhibited significant spatial heterogeneity:
  • Moderate stress (13.82 km²) dominated the central GWSA, primarily driven by waterlogging-induced soil degradation.
  • Severe stress (6.04 km²) occurred along GWSA boundaries, characterized by alternating drought-waterlogging dominance.
• Key drivers of compound stress were groundwater level, precipitation, and temperature.
  • Temperature showed a paradoxical effect, where increasing temperatures reduced drought areas and increased waterlogging areas, attributed to enhanced snowmelt-driven groundwater recharge.
  • Groundwater depth was the dominant factor influencing waterlogging dynamics (importance scores 0.193–0.256).
  • Precipitation exhibited lagged effects on total drought and waterlogging areas, mediated by the groundwater system.
• A temporal shift in dominant stress was observed, transitioning from waterlogging-dominant (2018–2019) to drought-dominant (2020–2022) regimes, with 2020 marking a critical transition (waterlogging areas decreased by 76%, drought areas surged by 471%).

Contributions

• Proposes a novel remote sensing-based framework for monitoring groundwater-driven compound drought-waterlogging stress in arid GWSA, addressing a gap in existing literature that often focuses on surface water-driven events.
• Develops a high-resolution (30 m) daily root-zone soil moisture product (100 cm depth) by effectively downscaling multi-source data, which significantly mitigates precipitation interference and directly captures groundwater-regulated stress signals.
• Introduces the SMCI-index and Double Stress Index (DSI) with physiologically relevant time lag thresholds for precise identification and classification of compound stress events.
• Provides critical insights into the spatiotemporal characteristics and complex driving mechanisms of compound stress in GWSA, including the counter-intuitive role of temperature via snowmelt recharge.
• Offers actionable scientific support for enhancing agricultural resilience, disaster risk prevention, and optimizing agricultural water management, particularly advocating for integrated "surface water-groundwater" governance in arid and semi-arid regions.

Funding

• National Natural Science Foundation of China (grant no. 42161020 and 42201174)
• Natural Science Foundation of Qinghai Province, China (grant no. 2023-ZJ-943J)

Citation

@article{Fu2025Remote,
  author = {Fu, Di and Jin, Xin and Jin, Yanxiang and Mao, Xufeng and Yao, Na},
  title = {Remote sensing-based monitoring of compound drought-waterlogging stress in groundwater-sensitive agroecosystems in arid regions},
  journal = {Agricultural Water Management},
  year = {2025},
  doi = {10.1016/j.agwat.2025.109826},
  url = {https://doi.org/10.1016/j.agwat.2025.109826}
}