Wang et al. (2025) A UAV-based method for root zone soil moisture modeling of different farmland scale with grain and economic crops

Identification

• Journal: Agricultural Water Management
• Year: 2025
• Date: 2025-11-03
• Authors: Jichao Wang, Hongwei Huang, H.H.S. Ariyasena, Jian Na Zhao, Xinyue Zhang, Xuerui Gao, Zhao Xi-ning, Yangzi Zhao
• DOI: 10.1016/j.agwat.2025.109932

Research Groups

• College of Water Resources Science and Engineering, Taiyuan University of Technology, Taiyuan, Shanxi, China
• College of Water Resources and Architectural Engineering, Northwest A&F University, Yangling, Shaanxi, China
• Institute of Soil and Water Conservation, Northwest A&F University, Yangling, Shaanxi, China
• College of Soil and Water Conservation Science and Engineering, Northwest A&F University, Yangling, Shaanxi, China
• College of Resources and Environment, Northwest A&F University, Yangling, Shaanxi, China
• School of Economics, Management, and Law, Shaanxi University of Technology, Hanzhong, Shaanxi, China

Short Summary

This study developed an integrated UAV-based remote sensing and Remote Sensing-based Water Balance Assessment Tool (RWBAT) model to accurately estimate root zone soil moisture (RZSM) for four crop types in the Loess Plateau, demonstrating high simulation accuracy, especially at deeper soil depths.

Objective

• To apply the Remote Sensing-based Water Balance Assessment Tool (RWBAT) model for the first time to analyze root zone soil moisture under different vegetation conditions.
• To construct high-resolution (5 m) spatial and temporal maps of RZSM by integrating UAV-based remote sensing with the RWBAT model.
• To enhance the RWBAT model by combining normalized, high-resolution vegetation indices with in-situ and meteorological data for effective estimation of RZSM across diverse land cover types.
• To evaluate the advantages of the current study’s UAV-based RZSM estimation methods over traditional methods.

Study Configuration

• Spatial Scale: A 0.77 km² farmland area in Shaanxi Province, China, divided into 5 m x 5 m grids. Soil moisture was measured at depths from 0 to 140 cm (every 20 cm). Four crop types were studied: winter wheat, maize, rapeseed, and apple trees.
• Temporal Scale: Data collection and simulation spanned from March to September 2023, with 13 specific data acquisition dates. Meteorological data were recorded every 30 minutes.

Methodology and Data

• Models used:
  • Remote Sensing-based Water Balance Assessment Tool (RWBAT) model.
  • Linear regression models for Leaf Area Index (LAI) estimation.
  • Monte Carlo algorithm for RWBAT model parameter calibration.
• Data sources:
  • UAV multispectral imagery (DJI MATRICE 300 RTK with Mica Sense Red Edge-MX camera, 13.89 cm ground sampling distance).
  • Field-measured Leaf Area Index (LAI) using a direct measurement method at 30 sampling sites.
  • Field-measured soil moisture data (drying method and TDR-Trime) at 30 sampling sites, 0-140 cm depth.
  • Meteorological data (precipitation, solar radiation, air temperature, relative humidity, wind speed) from a nearby weather station.
  • Soil type and property data from the China soil database.
  • Nine vegetation indices (e.g., NDVI, EVI, SAVI, DVI) derived from UAV imagery.

Main Results

• Crop-specific LAI estimation models, built using optimal vegetation indices, achieved R² values ranging from 0.60 (apple trees) to 0.87 (maize), with RMSE values from 0.39 m²/m² to 3.49 m²/m².
• The RWBAT model demonstrated high simulation accuracy for multi-depth soil moisture, particularly at 120–140 cm depth, with R² values of 0.91 (wheat), 0.76 (apple trees), 0.78 (rapeseed), and 0.80 (maize).
• Model calibration yielded R² values for key parameters ranging from 0.65 to 0.74, and RMSE values from 0.016 m³/m³ to 0.034 m³/m³.
• High-resolution (5 m pixel size) RZSM distribution maps were successfully generated, showing clear banded stratification patterns influenced by crop types and their LAI/NDVI variability.
• Sensitivity analysis revealed that increases in relative humidity and precipitation enhance soil moisture across all crops, with precipitation having a more pronounced influence at deeper soil depths. Temperature increases reduced soil moisture due to enhanced evapotranspiration.

Contributions

• First successful application and calibration of the RWBAT model for estimating root zone soil moisture across diverse crop types (wheat, maize, rapeseed, apple trees).
• Developed a novel integrated UAV-RWBAT approach for generating high-resolution (5 m) spatial and temporal RZSM maps, addressing limitations of coarse-resolution satellite data.
• Enhanced the RWBAT model by effectively integrating high-resolution UAV-derived vegetation indices with in-situ and meteorological data, improving RZSM estimation accuracy, especially in deeper soil layers (>100 cm).
• Demonstrated a process-based modeling approach that reduces overfitting risks inherent in purely data-driven methods and provides crucial insights for precision agricultural water management.

Funding

• National Key Research and Development Program of China (2022YFD1900700)
• National Natural Science Foundation of China (42377346)

Citation

@article{Wang2025UAVbased,
  author = {Wang, Jichao and Huang, Hongwei and Ariyasena, H.H.S. and Zhao, Jian Na and Zhang, Xinyue and Gao, Xuerui and Xi-ning, Zhao and Zhao, Yangzi},
  title = {A UAV-based method for root zone soil moisture modeling of different farmland scale with grain and economic crops},
  journal = {Agricultural Water Management},
  year = {2025},
  doi = {10.1016/j.agwat.2025.109932},
  url = {https://doi.org/10.1016/j.agwat.2025.109932}
}