Ensemble Compute Applications In Hec-Hms

Ensembles represent a way to pool predictions from several models and characterize the uncertainty associated with both the inherent climate variability and model differences. The Hydrologic Modeling System (HEC-HMS) is developed by the U.S. Army Corps of Engineers, Hydrologic Engineering Center. One of the new capabilities in software version 4.11 is Ensemble Compute. The new Ensemble Compute feature simplifies the process of obtaining and analyzing ensemble results. This presentation will showcase the new Ensemble feature by first providing an overview of its usage and available documentation, followed by two case study examples: (1) post-fire hydrology forecasting with different precipitation scenarios and (2) hydrologic model with climate prediction data from Coupled Model Intercomparison Project (CMIP 5). .

In this example, runoff response is predicted with numerical weather forecast models coupled to calibrated hydrology models for the Gallinas Creek watershed affected by the recent New Mexico fires. Debris and sediment from the fire scars could cause flooding in the watershed during the monsoon season. Five gridded precipitation scenarios were downloaded from the NOMADS Climate and Weather Model Archive at the National Oceanic and Atmospheric Administration and used to estimate flood flows for each of the five meteorologic ensemble members.

Practitioners and researchers routinely analyze output from multiple climate models to better evaluate uncertainty and improve accuracy of hydrologic predictions. Numerous climate models are available, some producing locally downscaled data that can be directly used with established hydrologic models. In this example, we run an Ensemble analysis of ten of the downscaled CMIP5 datasets representing ten different climate scenarios. The ten temperature and precipitation datasets are input into the calibrated HEC-HMS model for Tule River Basin - a 390 square mile basin above Schafer Dam in Tulare County, CA. The model uses the Structural Discretization method with Gridded Temperature Index for snowmelt computation, and the Gridded Precipitation and Interpolated Temperature meteorological methods to produce precipitation and temperature grids. We analyzed both ‘hindcasting’ (comparing the model prediction for a period in the past to observed results), and future model outputs (up to 2099), looking at potential trends in precipitation, temperature, snow accumulation and stream flows.