Evaluating the Effects of Inundated Land Area on Downstream Flow Using the LIS-VIC Model

The Red River basin of the North is prone to flooding and surface ponding due to the flat terrain, low soil permeability and influence of soil freeze/thaw. The spatial and temporal variability of water exchange between channels and uplands in these low-gradient areas are poorly understood, and quantifying surface storage in these locations is necessary in order to improve accuracy of flood forecasting in downstream locations. This study is part of a larger project in conjunction with the North Central River Forecast Center (NCRFC) to better account for the surface ponding in order to improve streamflow predictability. In this study the Land Information System (LIS) integrated Variable Infiltration Capacity (VIC) model, recently updated to represent surface water storage, is used to quantify temporary surface storage and analyze the effects on downstream peak flow in order to enhance understanding of the spatial and temporal variability of exchange between uplands and stream channels. The LIS-VIC model is calibrated and validated using observed streamflow and Moderate Resolution Imaging Spectroradiometer (MODIS) inundated land area (ILA) time series. The model’s performance is then evaluated for three different parameterizations: 1) No representation of surface water ponding, 2) Surface ponding represented by the land surface hypsometry, and 3) Surface ponding represented by the topographic wetness index (TWI) distribution. The synergy between the LIS-VIC model and ILA provides better understanding on the effects inundated uplands have on downstream flooding. This capability creates potential for future research on the assimilation of ILA products into the LIS-VIC model to support streamflow forecasting applications.