资源环境科技发展态势分析平台

Tree‐ring records of pre‐instrumental hydroclimate, which contribute needed context for understanding recent drought and flood events, typically provide one value per year that represents the entire year or one particular season. This may introduce a seasonal bias to the records or omit seasonally‐variable moisture. Here, I use tree‐ring records to reconstruct precipitation and runoff in 28 U.S. West Coast watersheds for running one‐, three‐, six‐, and twelve‐month intervals. When compared on a yearly basis, the Monthly and Four‐Season models have higher overall skill and better extreme capture in most basins than the Cool‐Warm and Annual models. The Monthly and Four‐Season versions also decrease model error in years with more intense precipitation and retain more variability in the pre‐instrumental period. Improved capture of year‐round moisture can provide a more complete representation of the pre‐instrumental past and strengthen capacity to study shorter‐duration and season‐specific events like atmospheric rivers.