Is Ephemeral Surface Water Essential to Modeling Groundwater of a Leveed River Floodplain?

Large river floodplains are some of the most active hydrologic settings on Earth. The forces driving this dynamism can negatively affect levees and exacerbate agriculture runoff. Numerical modeling helps manage these concerns. However, the number of flooded areas can rapidly increase with model size, which increases uncertainty if sufficient field observations are not available. Fortunately, remote sensing innovations are helping fill in the missing data. The Sentinel-1 array is a prime example, which collects 10-meter resolution radar data every 12-days across the globe. Several studies have validated surface water extents from Sentinel-1, but few have incorporated that information into hydrologic modeling.

This project evaluates the influence of Sentinel-1 derived surface water on the Middle Mississippi River valleys groundwater system. The most recent topobathemetry DEM for this floodplain was sampled with perimeter pixels of each Sentinel-1 identified open-water area. The interpolated perimeter elevations represent hydraulic head boundaries of two-way exchange between coupled numerical flow models of groundwater (MODFLOW-2005) throughout the valleys unconsolidated materials, and surface water (HEC-RAS) through the river corridor (channel and connected floodplain). These integrated models simulate conditions from December 2019 to August 2020, using a timestep synchronized with the Sentinel-1 repeat period of 12-days.

Particle tracking results (MODPATH-6) show considerably increased rates (up to +50 cm/day, σ ±5 cm/day), and changing azimuth directions (up to ±6.3°, σ ±1.1°) of groundwater flow when utilizing the Sentinel-1 data. These dynamics were most prominent near sudden changes in topographic relief, such as along levees and channel banks. These findings provide novel insights of groundwaters impact on levee stability, underseepage, and consequently the transport of nutrients throughout this agriculture dominated landscape. This research also raises awareness about innovative remote sensing methods that can enhance various types of hydrologic modeling around the world.