J1B.3 A Statewide, Weather-Regime based Stochastic Weather Generator for California

Monday, 29 January 2024: 9:00 AM
340 (The Baltimore Convention Center)
Nasser Najibi, Cornell University, Ithaca, NY; and A. Weyant, A. J. Perez, W. Arnold, A. Schwarz, A. Gershunov, R. Maendly, and S. Steinschneider

Water resource planners in California must prepare for the evolving effects of climate change. However, there is significant uncertainty about how the climate will evolve over the coming decades. Despite consensus that California’s future will be warmer and precipitation extremes will intensify, the rate of change is less clear, as are other elements like changes in average precipitation and modes of regional atmospheric circulation. Furthermore, California experiences extreme, natural variability in hydroclimate, which is certain to increase and also must be considered in future planning efforts. This work develops a weather-regime based stochastic weather generator for California that can create large climate ensembles to support bottom-up water resources planning under future climate uncertainty. The model provides a way to translate various signals of climate change from global climate models (GCMs) into traces of weather that are tailored for vulnerability-based planning efforts, but which are also linked to a process-based understanding of major controls over regional climate. We demonstrate that the weather generator can reproduce the climate across the entire state of California very well based on a large suite of performance metrics, including extreme precipitation and drought across multiple spatiotemporal scales. The model also compares favorably to state-of-the-science downscaled GCM simulations in terms of reproducing extremes. The weather generator is used to create an ensemble of 1000-year climate scenarios at a 6 km resolution across California representing various signals of climate change, including those directly related to warming (i.e., thermodynamic changes such as increased temperatures and more intense precipitation) and more nuanced signals related to shifts in atmospheric circulation (i.e., dynamic climate change). The resulting scenarios reveal that even in the absence of anthropogenic climate change, climate extremes can far exceed those experienced in the historical record. However, when even modest amounts of anthropogenic climate change are also considered, the combined impacts on extreme events are severe and highlight that significant investment in water resources systems will almost certainly be needed to sustain adequate water services across the state.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner