Reconstructing extreme precipitation in Northern California using tree-ring based moisture proxies across the Western United States

Extreme precipitation and consequent floods are one of the most common natural disasters impacting California, but they also constitute a significant portion of the available water supply. This motivates the need to better understand the past frequency and inter-annual persistence of these events, including the atmospheric rivers (ARs) that cause them, in order to help contextualize potential trends in extremes linked to climate change. In this study, we hypothesize that extreme precipitation in Northern California can be reconstructed by a network of tree-ring based moisture proxies that record the path of landfalling ARs as they deliver moisture to Northern California and then further into the interior of the Inter-Mountain West (IMW). We use a Hidden Markov Model (HMM) to develop an annual index for the primary mode of cold-season extreme precipitation counts across 67 gages in Northern California. This index is regressed against two tree-ring based moisture proxy datasets that reflect cold-season precipitation and summertime soil moisture, respectively. These regressions are conducted with tree-ring based data across Northern California only and across 5 states along the West Coast and in the IMW, in order to isolate the benefits of tree-ring based data further into the continental interior. We conclude by analyzing a reconstruction of regional extreme precipitation occurrences in Northern California that extends back to ~500 CE, and comparing pre-instrumental extreme events highlighted by our reconstruction with records of past extremes in the written and sediment records. Initial results show that tree-ring based data further into the continental interior are informative of extremes in Northern California, and that reconstructions can be significantly improved by using tree-ring based data tailored for the cold-season. Furthermore, our reconstruction corroborates major flood events recorded in the sediment record, while also highlighting previously unidentified extremes.