Improved integrated framework for flooded crop damage and recovery assessment: A multi-source earth observation and participatory mapping in Hadejia, Nigeria.
Journal:
Journal of environmental management
Published Date:
Apr 30, 2025
Abstract
Flooding has increasingly significant adverse effects on global food security, and there is a lack of a framework to effectively integrate remote sensing with survey data for accurate damage and recovery assessment. Also, optical satellite images for flood mapping face cloud interference, and free synthetic aperture radar (SAR) lack the temporal frequency needed to capture flooding dynamics. This study developed a new framework for modelling crop damage, loss, and recovery due to flash flooding using time-series multi-sensor satellite images. Crop recovery from remote sensing was validated with extensive participatory mapping data. Crop damage and recovery were assessed during Nigeria's 2020 and 2022 floods. Consistency was found between farmer-reported losses and remote sensing-based damage assessments: 91 % of farmers reporting total crop loss had no recovery. Flood maps and crop recovery assessments achieved over 90 % accuracy, demonstrating the reliability of multi-source optical and SAR satellite images combined with a machine learning technique. Severe flood damage was evident, with only 13 % and 16 % of flooded cropland recovered in 2020 and 2022, respectively. The integrated approach developed in this study eliminates uncertainties in other remote sensing techniques, overcomes participatory mapping limitations, and offers scalability for national-level implementation, providing critical information for post-disaster planning, farmer compensation, and sustainable agricultural practices to enhance food security in a changing climate.