Subseasonal‐to‐Seasonal Hindcast Skill Assessment of Ridging Events Related to Drought Over the Western United States

November 23, 2020

CW3E scientist Peter Gibson, along with researchers from CW3E (Michael DeFlorio and Luca Delle Monache), NASA’s Jet Propulsion Laboratory (Duane Waliser and Alexander Goodman) and NASA’s Global Modeling and Assimilation Office (Andrea Molod) recently published an article in the Journal of Geophysical Research: Atmospheres titled “Subseasonal‐to‐Seasonal Hindcast Skill Assessment of Ridging Events Related to Drought Over the Western United States”.

Subseasonal‐to‐Seasonal (S2S) forecasts of atmospheric rivers, precipitation and drought are scientifically challenging, yet are of immense value for stakeholders and decision makers in the water resource management sector. While forecasting precipitation directly on S2S timescales has proven difficult in past research, this study approaches the problem from a different angle – to focus on persistent atmospheric ridging events that are well known to cause drought across the Western United States. In prior research, the authors have developed a ridge detection algorithm that automatically tracks ridging events associated with drought, which can be applied seamlessly across very large observational, weather and climate model datasets.

By applying this ridge detection algorithm, the authors probed the extent to which drought relevant ridging events can be forecasted on S2S timescales (2-6 weeks ahead) in state-of-the-art dynamical forecast models. For certain models, and certain aspects of ridging, the authors found evidence of skillful ridging forecasts across Weeks 3 and 4, and more modest skill across Weeks 5 and 6 lead time. The authors also explored conditions in the climate system that tend to help extend prediction skill in the forecast models as well as conditions that tend to reduce skill. This information is valuable for future research and product development, for bias correcting known errors in the output from forecast models and improving S2S prediction skill.

Figure 1. Probabilistic Brier Skill score (BSS) for predicting above normal occurrence of ridge types (N, S, and W ridge types) across a 2-week period for lead times of weeks 1-2, weeks 3-4 and weeks 5-6. Skill scores above zero indicate model skill compared to a reference prediction. The left panels (a, c) show skill assessed relative to a reference climatology defined by the historical median ridge type frequency across the particular 2-week period. Panels (b,d) show skill assessed relative to randomly sampling (with replacement) from the reference distribution defined by the historical distribution of each ridge type frequency found in reanalysis.

Based on this research, experimental forecasts are being issued in real time by CW3E and are publicly available. This S2S forecast product, as well as the underpinning research, supports CW3E’s Strategic Plan goals for revolutionizing seasonal outlooks of extreme events in North America and their impacts on floods, drought, hydropower, and the economy. This research received funding and support from the California Department of Water Resources.

Gibson, P. B., Waliser, D. E., Goodman, A., DeFlorio, M. J., Delle Monache, L., & Molod, A. (2020). Subseasonal‐to‐seasonal hindcast skill assessment of ridging events related to drought over the Western United States. Journal of Geophysical Research: Atmospheres, 125, https://doi.org/10.1029/2020JD033655.