Monday, 10 January 2005
Use of MODIS snow cover imagery for streamflow and reservoir storage forecasts in the Snake River basin
We describe an approach to seasonal streamflow forecasting for the Snake River, a major tributary of the Columbia River. Our method employs the Variable Infiltration Capacity (VIC) macroscale hydrology model, which is modified to allow adjustment of the model's initial snow state using Moderate Resolution Imaging Spectroradiometer (MODIS) imagery for winters 2000-2004. Seasonal and two-week streamflow forecasts made on March 1, April 1, April 15, May 1 and May 15 were evaluated retrospectively using archived MODIS imagery for 2000-2003. We evaluated short lead (two week) streamflow forecasts in real-time for a subset of the above forecast dates in 2004. In general, reduction in mean absolute error between predicted streamflow and observed streamflow, caused by inclusion of MODIS data, was greater for the spring two-week forecasts than for the longer seasonal forecasts, and there was some indication that error reduction for the seasonal forecasts was higher for forecasts made later in the spring. The effect of MODIS updating on reservoir storage volume forecasts using the SnakeSim monthly reservoir management model was also evaluated. For reservoirs where the model performed well in retrospective simulations, storage forecast errors were generally reduced or unchanged as a result of MODIS updated streamflow forecasts. In addition, we evaluated updated seasonal and two-week streamflow forecasts for two dates during Spring 1997 using the snow cover product from the National Operational and Hydrologic Remote Sensing Center (NOHRSC).