Reconstruction of historical California flood forecasts
Abstract
Floods are a consistent threat to the population and infrastructure of California, which has highly variable precipitation and flood extremes relative to the rest of the conterminous U.S. The warming climate is expected to accelerate the frequency of climate extremes and exacerbate the flooding problem. Short-term forecasting is of key importance to mitigating the human suffering and economic losses caused by floods. WRF-Hydro (aka National Water Model or NWM) provides new technology for producing flood forecasts. However, it has long been suspected that NWM-based flood forecasts are less accurate in California than forecasts produced using the Sacramento Soil Moisture Accounting model (Sac-SMA). Here, we evaluate the ability of the hydrologic core of NWM (the Noah-MP model) to produce accurate flood forecasts compared with Sac-SMA at a set of index stations across the state. We were provided with historical flood forecasts produced by Sac-SMA over the last ~20 years by the California-Nevada River Forecast Center (CNRFC). We reconstructed the Sac-SMA forecasts, and produced parallel reforecasts based on Noah-MP. In general, our Sac-SMA-based reconstructed flood forecasts were consistent with the archived forecasts produced by CNRFC. We found that in most cases, with appropriate calibration and selection of physics options, the Noah-MP reforecasts were of comparable (although not superior) accuracy to those produced using Sac-SMA.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFM.H33C..05M