Rice inundation assessment using polarimetric UAVSAR data

Authors: HUANG, XIAODONG - Applied Geosolutions, Llc; RUNKLE, BENJAMIN - University Of Arkansas; ISBELL, MARK - Isbell Farms; MORENO-GARCIA, BEATRIZ - University Of Arkansas; Reba, Michele; TORBICK, NATHAN - Applied Geosolutions, Llc

Submitted to: Earth and Space Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/15/2021
Publication Date: 2/7/2021
Citation: Huang, X., Runkle, B., Isbell, M., Moreno-Garcia, B., Reba, M.L., Torbick, N. 2021. Rice inundation assessment using polarimetric UAVSAR data. Earth and Space Science. 8. https://doi.org/10.1029/2020EA001554.
DOI: https://doi.org/10.1029/2020EA001554

Interpretive Summary: Irrigated rice requires intense water management for cultivation under typical agronomic practices. Cost effective tools to improve the efficiency and assessment of water use are a key need for industry and resource managers. In this research we employ two different techniques-model, based decomposition and machine learning, to map flooded rice using time-series data obtained from unmanned aerial vehicles (i.e. Uninhabited Aerial Vehicle Synthetic Aperture Radar or UAVSAR). Simultaneous field measurements of water depth were collected during the 2019 crop season using water level sensors in commercial rice fields across the study site in east-central, Arkansas, USA. Outcomes show that the two techniques used were able to identify rice flooding status. Both techniques identified similar parameters and achieved an overall accuracy of 80%. These outcomes are useful to farmers and resource managers as they provide accurate information of crop water conditions over large areas efficiently. The growth of UAVSAR data and upcoming missions for further data collection should enhance our ability to monitor inundated conditions that are required to track rice water volume.

Technical Abstract: Irrigated rice requires intense water management under typical agronomic practices. Cost effective tools to improve the efficiency and assessment of water use is a key need for industry and resource managers to scale ecosystem services. In this research application we employ model-based decomposition and machine learning to map inundated rice using a time-series of polarimetric, L-band Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) data. Simultaneous ground truth observations recorded water depth inundation status during the 2019 crop season using instrumented fields across the study site in east-central, Arkansas, USA. A three component model-based decomposition generated metrics representing surface-, double bounce-, and volume- scattering along with a shape factor, randomness factor, and the Radar Vegetation Index (RVI) to characterize crop inundation status independent of growth stage and under canopy cover. Machine learning (ML) comparisons employed Random Forest (RF) using the UAVSAR derived parameters for identifying cropland inundation status across the region. Outcomes show that RVI, proportion of the double-bounce within total scattering, and the relative comparison between the double-bounce and the volume scattering have moderate to strong mechanistic ability to identify rice inundation status with Overall Accuracy (OA) achieving 75%. The use of relative ratios further helped mitigate the impacts of far range incidence angles. The RF approach, which requires training data, achieved a higher OA and Kappa of 88% and 71%, respectively, when leveraging multiple SAR parameters. The growth of polarimetric L-band availability should enhance cropland inundation metrics beyond open water that are required for tracking water quantity at field scale over large areas.