Page Banner

United States Department of Agriculture

Agricultural Research Service

Research Project: Application of a physically-based distributed snowmelt model in support of reservoir operations and water management

Location: Watershed Management Research

Project Number: 2052-13610-010-05-I
Project Type: Interagency Reimbursable Agreement

Start Date: Feb 27, 2013
End Date: Dec 31, 2016

This project will focus on the Bureau of Reclamation’s need for hydrologic modeling tools that can optimize reservoir management decisions to improve planning and management of water supplies in both the short- and long-term. The work is broken up into two phases broken down in three annual stages apiece. The first phase addresses short-term river forecasts (i.e. current to 3-day advance) while the second phase evaluates model performances over longer durations (i.e. decadal scale). In both phases the latest advances in snow accumulation and melt modeling, much of it developed at the USDA-ARS Northwest Watershed Research Center and contained in the Isnobal snow model, will be applied to the Boise River Basin. There are three large reservoirs in this stretch of the Boise River. This will be the first application of a physically-based distributed snow model in an operational forecasting setting.

Phase 1 will assess the suitability of these modeling tools to operational needs and evaluate their performance against currently available streamflow products. Stage one will be the application of Isnobal, a physically-based distributed snow model, in an operational setting - the Boise River Basin. The model will be forced with currently available data including meteorological stations operated by the Natural Resources Conservation Service, Reclamation, and Sawtooth National Forest Avalanche Center. Simulated snowcover will be evaluated against satellite-derived observations and melt inputs will be compared to measured discharge. In stage two Isnobal will be coupled to a soil storage or routing model in order to convert Isnobal-simulated surface water inputs to stream discharge. Once coupled, predicted streamflow will be evaluated against observations. Stage three will be a proof of concept of the abilities of the coupled Isnobal routing model to forecast short-term (1 - 3 days) reservoir inflows. Short-term weather forecasts will be manually input to the coupled model. Predicted inflows will be provided as a test product to Reclamation and other interested parties. Simulated inflows will be evaluated against other forecasting products and observations. Stages 1 – 3 will coincide with FYs 2013 through 2015. Phase 2 will evaluate how the high-resolution, physically-based distributed snow model compares to the simplified snow routines commonly applied in long-range, regional climate impact studies. In stage one routing and evapotranspiration routines from two commonly applied hydrologic models with simplified snow representations – DHSVM and VIC – will be assimilated into the high-resolution snow model. Stage two will compare the multiple combinations of snow and hydrologic models to ascertain the sensitivity of streamflow simulations and hence predicted reservoir inflows to the level of detail provided in the respective simulations. We will assess the effects of scale in both the hydrologic and snow components of the modeling process as well as how increasingly complex levels of process representations affect these simulations. Stage three will entail further testing of the working hypotheses developed in stage two, an assessment of potential biases introduced by model scales and complexities, and further recommendations for simulating climate change impacts on snowmelt-dependent water supplies in the Western United States. Stages 1 – 3 of Phase 2 will coincide with FYs 2014 – 2016

Last Modified: 09/25/2017
Footer Content Back to Top of Page