Chen et al. (2025) Monitoring sub-canopy inundation dynamics in global croplands: An unexplored application of SWOT satellite data

Identification

• Journal: Agricultural Water Management
• Year: 2025
• Date: 2025-12-12
• Authors: Yongzhe Chen, Shunlin Liang, Huanjun Liu, Phuping Sucharitakul, Xuejing Leng, Husheng Fang, Wenyuan Li, Han Ma, Jianglei Xu, Yichuan Ma, Lichang Yin
• DOI: 10.1016/j.agwat.2025.110075

Research Groups

• Jockey Club STEM Lab of Quantitative Remote Sensing, Department of Geography, The University of Hong Kong, Hong Kong, China
• The University of Hong Kong ‒ Shenzhen Institute of Research and Innovation, Shenzhen, China
• State Key Laboratory of Black Soils Conservation and Utilization, Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences, Changchun, China
• Department of Biology, Faculty of Science, Chiang Mai University, Chiang Mai, Thailand
• Jiangsu Key Laboratory of Soil and Water Processes in Watershed, College of Geography and Remote Sensing, Hohai University, Nanjing, China
• School of Remote Sensing and Information Engineering, Wuhan University, Wuhan, China
• State Key Laboratory of Ecological Safety and Sustainable Development in Arid Lands, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Urumqi, China

Short Summary

This study develops the first method for year-round monitoring of sub-canopy inundation dynamics (non-inundated, partially-inundated, fully-inundated) in global croplands using SWOT satellite KaRIn coherent power, demonstrating its robustness across diverse climate zones and improving upon existing remote sensing limitations.

Objective

• To develop the first algorithm capable of estimating the temporal variation in area fractions of different inundation statuses (i.e., non-inundated, partially-inundated, and fully-inundated) for paddy and non-paddy croplands across diverse climate zones.
• To demonstrate the algorithm’s robustness across four globally representative regions.

Study Configuration

• Spatial Scale: Four globally representative regions (Sanjiang Plain, southern Jiangsu Province, eastern Arkansas, northeast Thailand), covering diverse climate zones and cropland types. Data aggregated to approximately 60 meters × 60 meters resolution.
• Temporal Scale: July 1st, 2023 to December 31st, 2024 (study period for SWOT data). Focus on year-round dynamics.

Methodology and Data

• Models used:
  • Gaussian Mixture Model (GMM) for classifying inundation statuses and detecting noise in coherent power data.
  • Median Absolute Deviation (MAD) filter for identifying outlier standard deviation values and abnormal wind conditions.
  • Polynomial functions for fitting surfaces representing coherent power thresholds as a function of incidence angle and Normalized Difference Vegetation Index (NDVI).
• Data sources:
  • SWOT satellite Ka-band Radar Interferometer (KaRIn) Level 2 KaRIn high-rate water mask pixel cloud (L2HRPIXC) data (Coherent Power - COP, Incidence Angle - INC).
  • Harmonized Landsat-8 and Sentinel-2 (HLS) v2.0 data for Normalized Difference Vegetation Index (NDVI).
  • Local, high-resolution land cover maps (e.g., Cropland Data Layers in USA, SERVIR 30 m land cover maps for Southeast Asia) or custom cropland maps for stable cropland identification.
  • Multi-Source Weighted-Ensemble Precipitation (MSWEP) v2.8 data for precipitation.
  • MODIS Net Evapotranspiration Gap-Filled 8-Day L4 Global 500 m SIN Grid V061 product for evapotranspiration.
  • Ground-truth photographs (PhenoCam photos), farmer interviews, and published literature for validation.
  • CYGNSS-based Berkeley-RWAWC V3.1 dataset for comparison.

Main Results

• The study developed the first algorithm for effective year-round cropland inundation monitoring across diverse climate zones, leveraging SWOT KaRIn coherent power (COP) to distinguish non-inundated, partially-inundated, and fully-inundated fields.
• The method systematically mitigates confounding factors such as incidence angle, vegetation water content, wind speed variability, and noise, establishing COP thresholds using Gaussian Mixture Models.
• Validation across four globally representative regions showed high consistency with ground-truth photographs (17 out of 18 matches), farmer interviews, and published literature.
• For non-paddy croplands, the estimated fully-inundated area fraction consistently remained low (below 15% in simpler systems, below 30% in others), with predominantly non-inundated states except for periods of partial inundation.
• Paddy field inundation regimes varied significantly by region but aligned well with local practices and cropping calendars, capturing key inundation periods for different rice cultivation methods.
• The SWOT-based method demonstrated superior performance in detecting inundation beneath dense, mature crop canopies compared to the CYGNSS-based Berkeley-RWAWC dataset, which often underestimated inundated areas, particularly partial inundation and inundation under tall-stem rice varieties.

Contributions

• Presents the first algorithm to estimate the temporal variation in area fractions of non-inundated, partially-inundated, and fully-inundated statuses for both paddy and non-paddy croplands globally using publicly available SWOT satellite data.
• Introduces a novel application of SWOT KaRIn coherent power (COP) for sub-canopy inundation monitoring in vegetated areas, addressing a critical gap where such returns are typically excluded in standard SWOT Level 2 products.
• Develops a robust methodology to systematically mitigate and control the impacts of incidence angle, vegetation water content, wind speed variability, and residual noise on COP signals, enabling accurate inundation status determination.
• Provides detailed, spatiotemporally consistent information on cropland inundation dynamics, which can significantly refine agricultural hydrological models, improve estimates of irrigation water consumption, and reduce uncertainties in global greenhouse gas (methane) emission budgets.
• Offers a valuable proxy for oxygen-deficient soil conditions in non-paddy croplands, enhancing the accuracy of crop yield prediction models, especially in regions prone to waterlogging.

Funding

• Hong Kong Jockey Club Charities Trust Global STEM Professorship Scheme (grant no. GSP 225)
• Open Research Program of the International Research Center of Big Data for Sustainable Development Goals (grant no. CBAS2022ORP01)

Citation

@article{Chen2025Monitoring,
  author = {Chen, Yongzhe and Liang, Shunlin and Liu, Huanjun and Sucharitakul, Phuping and Leng, Xuejing and Fang, Husheng and Li, Wenyuan and Ma, Han and Xu, Jianglei and Ma, Yichuan and Yin, Lichang},
  title = {Monitoring sub-canopy inundation dynamics in global croplands: An unexplored application of SWOT satellite data},
  journal = {Agricultural Water Management},
  year = {2025},
  doi = {10.1016/j.agwat.2025.110075},
  url = {https://doi.org/10.1016/j.agwat.2025.110075}
}