Toward a Testable Theory of Large, Lowland Floodplain Inundation

Floodplain inundation and drainage is a spatially and temporally sporadic process of the hydrology of the Amazon River. Predicting flooding hazards to recognize consequences of land use, land cover, and climatic changes is crucial for a world that has an ever increasing demand and reliance on fresh water. Using in-situ measurements to assess the vast, diffusive flow conditions and storage changes along the Amazon floodplain is essentially not possible, whereas, a technologically advanced method utilizing two-dimensional height mapping must be used to understand changes in h, dh/dx, and dh/dt. We modeled a section of the River Solimoes above Itapeua, Brazil using Lisflood and topographic data from the Shuttle Radar Topography Mission (SRTM). Lisflood, a raster based flood modeling code, uses a one dimensional kinematic wave approximation for channel flow and is solved bu an explicit finite difference scheme. Floodplain inundation is modeled by a two dimensional diffusion wave. By simulating flooding along the River Solimoes, we are demonstrating that there is high potential for the simulations to be consistent with actual changes across floodplain channels. We seek simulations that predict h, dh/dx, and dh/dt across large, lowland floodplains particularly in support of the proposed Water Elevation Recovery (WatER) satellite mission. WatER will measure the elevations of water surfaces with a sub-100 m pixel spacing, thus our testable theory will provide relationships between water level measurements and the responsible flux and mass balances.