Temporal Clustering of Atmospheric Rivers and Precipitation Impacts in Southern South America

Landfalling Atmospheric Rivers (ARs) are often linked with extreme precipitation events which cause flooding and other hydrometeorological hazards. When ARs occur in close succession, or cluster temporally, these compounding events can result in heightened impacts. Here, we quantify the temporal clustering of historic wintertime (MJJA) ARs over the western coast of Southern South America. We identify AR events using the Guan and Waliser (2019) detection algorithm which employs a percentile-based IVT threshold. We apply a statistical framework utilizing the Ripley-K function to compare observed AR clustering rates in reanalysis data with synthetic, randomly distributed sequences of ARs. Using this method we can determine the regions and timescales at which sequences of AR events show temporal clustering at rates greater-than-random, and identify significant clusters of ARs in the historic record. In Southern South America, we find significant temporal clustering of ARs in the regions between 33-40 S (1-1.5 clusters per season) and 45-55 S (1-2 clusters per season). The Northern region contains clusters up to 25 days long with 3-4 AR events per cluster, and the Southern region has longer-lived cluster durations up to 40 days, each containing 3-4.5 events on average. Notably, the region from 40-45 S exhibits little to no significant clustering despite a high frequency of landfalling ARs. In regions with significant clustering, 30-60% of all ARs are part of a cluster, and clusters account for up to 40% of seasonal precipitation. Future work will include analyzing the influence of ENSO and other large-scale circulation patterns on AR clustering in this area. While this study focuses on Southern South America, it provides a framework that can be applied to similar regions and the results offer a historical context for understanding successive ARs that will serve as a baseline when studying future precipitation extremes.