Authors: Gabriel Lewis, Taylor Dixon, Julie Kalansky, Ming Pan – CW3E; Brett Whitin, Alan Haynes, Peter Fickenscher – CNRFC; Rob Hartman – Robert K. Hartman Consulting Associates
Title: Novel Approaches to Improving Operational Flood and Water Supply Forecasting Across the Sierra Nevada
Abstract: The California-Nevada River Forecast Center (CNRFC) provides daily hydrologic forecasts at nearly 200 locations across two states. These forecasts provide vital support to the National Weather Service (NWS) flood warning program and the water resources management programs. For example, the California Department of Water Resources (DWR) serves over 41 million people living within these flood-prone basins, including the $17 billion farming economy throughout the Central Valley. As anthropogenic climate change alters both the severity and frequency of atmospheric rivers (ARs), which provide up to half of the precipitation across California, operational agencies are having to recalibrate their planning and forecasting models to account for stronger storms. Traditionally, the CNRFC has calibrated their models over the past 40 years of historic discharge and weather data to provide daily streamflow and seasonal water supply forecasts. In recent years, the CNRFC has coupled their Sacramento-Soil Moisture Accounting (SAC-SMA) and SNOW-17 models with the Hydrologic Ensemble Forecast Service (HEFS) to help quantify uncertainties in their streamflow forecasts. HEFS combines near-term weather forecasts from the Global Ensemble Forecast System (GEFS) and historical weather possibilities (climatology) out to 365 days to supplement DWR’s Bulletin 120 water supply forecasts. The probabilistic streamflow forecasts from HEFS operate seamlessly across lead-times of hours to seasons, and thus enhance both the flood warning and water supply aspects of the DWR and CNRFC missions. However, HEFS is still a relatively young technology and stands to benefit from improvements over time, specifically to its ability to accurately and reliably forecast the fast and heavy runoff that ARs often generate.
Here, we compute HEFS hindcasts over 1980-2023 for the observed GEFS meteorology, produced by the Center for Western Weather and Water Extremes (CW3E) at Scripps Institute of Oceanography (SIO). We use these novel sources of input variables (forcings) for the CNRFC hydrologic models to calculate streamflows and compare agreements between simulations and observations over the past 40 years. We also compare our HEFS forecasts and hindcasts with the DWR Bulletin 120 seasonal water supply forecasts and observed April-July volume totals. We strive to continually work with operational partners like the CNRFC and DWR to develop and enhance techniques for improving hydrologic forecasts using state-of-the-art measurements, meteorological inputs, and models, and ultimately to help provide the best possible streamflow and water supply forecasts across the region.