Assessing Effects of Wetland Soil Moisture on Flood Progress in Kentucky Using Multi-Temporal Satellite Data

Low-lying agricultural land along the western segment of the Ohio River has been subject to intense flooding frequently from local streams. These flooded and waterlogged farmlands have suffered severe economic losses due to destroyed or inaccessible crop fields, delayed seeding and missed planting and harvesting opportunities. Rapid flood mapping and monitoring of their progress is crucial for crop health assessment, crop insurance and agriculture policy making. Among the various hydrological factors that affect the magnitude and rate of flooding, soil moisture content, surface topography and soil permeability are often considered to be the most critical factors.

The focus of our study is assessing the influence of soil moisture content on rapid surface runoff and flooding along the Ohio River through the integration of microwave and optical satellite datasets. We first generated flood progression maps for target areas along the Ohio River using a time-series of Sentinel-1 C-band and TerraSAR-X X-band images. We used the Cropland Data Layer and K-Mean Classification approach to isolate the flooded areas from the ground features such as open water bodies and urban areas. We measured the normalized difference vegetation indices and land surface temperatures to obtain soil moisture estimates from optical/infrared MODIS satellite data and ground measurements. Our initial results suggest a good correlation between initial soil moisture contents and ground surface runoff. Our study also demonstrates the applicability of microwave and optical sensors in rapid flood progress assessments for agricultural applications.