Zhang et al. (2026) Applying geostatistical electrical resistivity tomography and a water content estimation model for loess spatial mapping

Identification

• Journal: Environmental Earth Sciences
• Year: 2026
• Date: 2026-03-30
• Authors: Huiqi Zhang, Yue Liang, Tian-Chyi Jim Yeh, Rifeng Xia, Linli Li, Zhiwei Sun, Bin Zhang
• DOI: 10.1007/s12665-026-12916-2

Research Groups

• National Engineering Research Center for Inland Waterway Regulation, Chongqing Jiaotong University, China
• Department of Hydrology and Atmospheric Sciences, University of Arizona, USA
• Tianjin Survey and Design Institute for Water Transport Engineering Co. Ltd, China
• Chongqing Yufa Water Conservancy Research Institute Co.,Ltd, China

Short Summary

This study developed a novel piecewise model for estimating loess volumetric water content (θ) from electrical resistivity (ρ) data, significantly improving accuracy, especially in low-moisture zones. Coupled with geostatistical electrical resistivity tomography (GERT), this method effectively mapped the spatial distribution of θ in a loess slope, outperforming traditional techniques for geological hazard mitigation.

Objective

• To characterize the spatial distribution of volumetric water content (θ) in a loess slope using geostatistical electrical resistivity tomography (GERT) combined with laboratory measurements.
• To develop a precise quantitative piecewise model linking loess electrical resistivity (ρ) and θ, addressing underestimation issues in high ρ regions.

Study Configuration

• Spatial Scale: A loess slope in Baota District, Yan’an City, Shaanxi Province, China. Field sampling area of 6 meters × 6 meters × 5 meters depth. Loess layer depth up to 5 meters.
• Temporal Scale: Field sampling conducted in September 2019. GERT survey and laboratory measurements were performed.

Methodology and Data

• Models used:
  • Geostatistical Electrical Resistivity Tomography (GERT) based on the successive linear estimator (SLE).
  • Piecewise model for θ estimation from ρ, incorporating a critical resistivity threshold (79.4 Ω·m) to differentiate continuous and discontinuous conduction regimes.
  • Global power-law model (for comparison).
  • Kriging interpolation (for TDR comparison).
  • Empirical model by Sun et al. (2020) for silty loam (for comparison).
• Data sources:
  • Laboratory measurements: 30 undisturbed loess samples collected from 0 to 500 cm depth, analyzed for electrical resistivity (ρ) and gravimetric/volumetric water content (w/θ).
  • Field GERT survey: Data acquired from 6 monitoring wells (subsurface electrodes at 0.25 m intervals) and 25 surface electrodes (1 m grid over 4 m × 4 m area).
  • Borehole data: Used for validation of GERT-derived θ distributions.
  • Simulated TDR data: Used for comparative accuracy assessment.

Main Results

• Vertical heterogeneity in θ was observed in the 0–500 cm loess layer, with average gravimetric water content (w) declining from 14.0% to 9.3% in the shallow layer (0–250 cm) and stabilizing around 8.9% in deeper layers.
• Dry density (d_d) progressively increased from 1370 kg·m⁻³ to 1540 kg·m⁻³ with depth.
• A critical electrical resistivity threshold of 79.4 Ω·m was identified, marking the transition of pore water from continuous to discontinuous states.
• The developed piecewise model significantly reduced θ estimation errors in high ρ regions compared to traditional global models, with the mean absolute error (MAE) decreasing by 19.3% and the root mean squared error (RMSE) by 15.5%.
• GERT-derived θ distributions, using the piecewise model, showed moderate consistency with borehole data (RMSE = 1.98%), accurately identifying a vertical infiltration interface at 150–200 cm depth.
• The proposed method outperformed TDR spatial interpolation (MAE = 1.70%, RMSE = 3.24%) and regional empirical models (MAE = 2.34%, RMSE = 2.69%) in terms of θ prediction accuracy.

Contributions

• Developed a novel piecewise model that accurately links loess electrical resistivity and volumetric water content by accounting for the transition between continuous and discontinuous pore water conduction regimes, addressing underestimation biases of traditional models in low-moisture conditions.
• Demonstrated the superior performance of Geostatistical Electrical Resistivity Tomography (GERT) combined with the piecewise model for high-resolution, large-scale spatial mapping of volumetric water content in loess slopes.
• Provided a practical and accurate technical solution for monitoring moisture distribution, crucial for geological hazard mitigation (e.g., landslide early warning) and eco-hydrological regulation in the Chinese Loess Plateau.
• Elucidated the underlying physical mechanisms of electrical conduction in loess, highlighting the roles of pore water connectivity, compaction density, and electrical double-layer effects on resistivity-water content relationships.

Funding

• National Natural Science Foundation of China (52379097, 52509138)
• Guangxi Science and Technology Program (AA23062023)
• Water Conservancy Science and Technology Project of Jiangxi Province (202426ZDKT27)
• US NSF award EAR 1931756
• Global Expert award through Tianjin Normal University from the Thousand Talents Plan of Tianjin City

Citation

@article{Zhang2026Applying,
  author = {Zhang, Huiqi and Liang, Yue and Yeh, Tian-Chyi Jim and Xia, Rifeng and Li, Linli and Sun, Zhiwei and Zhang, Bin},
  title = {Applying geostatistical electrical resistivity tomography and a water content estimation model for loess spatial mapping},
  journal = {Environmental Earth Sciences},
  year = {2026},
  doi = {10.1007/s12665-026-12916-2},
  url = {https://doi.org/10.1007/s12665-026-12916-2}
}