Monitoring sub-canopy inundation dynamics in global croplands: An unexplored application of SWOT satellite data

Inundation dynamics within croplands, particularly paddy fields, profoundly influence irrigation water consumption, greenhouse gas emissions, and crop yields. The traditional practice of continuous flooding irrigation in paddy fields has been abandoned in many regions worldwide, contributing to reduced water consumption and methane emissions. However, regional information on paddy field irrigation/inundation regime is typically limited to coarse estimates derived from meta-analyses or government reports, as no existing method enables spatiotemporally consistent monitoring of inundation beneath crop canopies worldwide. This limitation arises from the inability of optical sensors to detect water beneath dense canopies, substantial variability of SAR backscatter coefficients across crop growth stages, and the limited availability of full-polarization SAR data. Here, we develop the first method for effective year-round cropland inundation monitoring across diverse climate zones. Our method leverages coherent power (COP) rather than water level measurements from the Ka-band Radar Interferometer (KaRIn) aboard the SWOT satellite to distinguish between non-inundated, partially-inundated, and fully-inundated fields. By systematically mitigating and controlling confounding factors that can affect COP signals, including incidence angle, vegetation water content, wind speed variability and noise, we establish COP thresholds for different inundation statuses under a variety of conditions using Gaussian Mixture Models. Validated across four globally representative regions, the method's results are consistent with ground-truth photographs (17/18 match), farmer interviews, and published literature. A comparison with the CYGNSS-based Berkeley-RWAWC dataset, which spans 37.4°S 37.4°N, demonstrates a better performance of our SWOT-based method. The estimated cropland inundation dynamics can provide valuable support for improved agricultural water management worldwide. • First universal method for cropland year-round sub-canopy inundation monitoring. • Explores a new application of SWOT satellite data. • Validation in four representative regions worldwide demonstrates a high accuracy. • Outperforms existing CYGNSS-based inundation mapping method. • Provides vital data for improved agricultural water management.