Predictability of Extreme Precipitation in Western US Watersheds Based on Atmospheric River Occurrence, Intensity, and Duration

We quantified the relationship between atmospheric rivers (ARs) and occurrence and magnitude of extreme precipitation in western U.S. watersheds, using ARs identified by the Atmospheric River Tracking Method Intercomparison Project and precipitation from a high-resolution regional climate simulation. Our analysis shows the potential of ARs in predicting extreme precipitation events at a daily scale, with Gilbert Skill Scores of similar to 0.2. Monthly extreme precipitation amount in west coast watersheds is closely related to AR intensity, with correlation coefficients of up to 0.6. The relationship between ARs and precipitation is most significant in the Pacific Northwest and California. Using a K-means clustering algorithm, ARs can be classified into three categories: weak ARs, flash ARs, and prolonged ARs. Flash ARs and prolonged ARs, though accounting for less than 50% of total AR events, are more important in controlling extreme precipitation patterns and should be prioritized for future studies of hydrological extreme events. Plain Language Summary Atmospheric rivers (ARs) are the elongated atmospheric bands that transport huge amount of water vapor toward the poles. Various studies have attributed ARs as a major cause of extreme precipitation and thus flooding in the western United States, but the reverse question is seldom investigated: when there is an AR, what is the likelihood of having an extreme precipitation event? In our study, we show that ARs can be used to predict the occurrence of extreme precipitation events at daily scale. We also show that the intensity of ARs is well correlated with extreme precipitation amount at a monthly scale. The results suggest that extreme precipitation prediction can potentially benefit from AR forecasts, which have a longer lead time than precipitation forecasts. Regarding the vast amount of ARs as identified from various AR-tracking algorithms, we used a machine learning-based approach to highlight ARs that are strongly correlated with hydrological extreme events. These ARs usually feature long duration or high intensity, and focusing on them may reveal a clearer and most robust relationship between ARs and hydrological extreme events.