Using Pacific Ocean climatic variability to improve hydrologic reconstructions

Abstract

Summary Reconstructions of hydrologic variables are commonly created using tree-ring chronologies. Incorporating climate signals and regionally specific sea surface temperature (SST) indices can greatly improve reconstructive skill. This study performed several reconstructions in the Upper Green River Basin (UGRB), located in the northern portion of the Upper Colorado River Basin. Reconstructed snowpack (April 1st snow water equivalent (SWE)) for the UGRB will provide information on the long-term variability of snowpack and assist in determining the effectiveness of current weather modification efforts. Previous attempts to reconstruct regional snowpack in the UGRB were unsuccessful due to a limited number of tree-ring chronologies. This study increased the number of stations used to represent regional SWE and expanded the spatial area of the study. This resulted in the first reconstruction of UGRB SWE using tree-ring chronologies (R2 = 0.34). Next, the predictor variable pool is increased by adding climate signals (Southern Oscillation Index (SOI), Pacific Decadal Oscillation (PDO)) and Pacific Ocean SSTs from regions that were determined to be teleconnected with UGRB SWE. Singular value decomposition (SVD) was performed on Pacific Ocean SSTs and UGRB SWE to identify coupled regions of climate (SSTs) and hydrology (SWE). Stepwise linear regression was used to identify the best predictor combinations. Including climate data (SOI) in the model improved the R2 value from 0.34 to 0.57. The addition of Pacific Ocean SST data further improved the skill of the model (R2 = 0.63). Reconstructions using Pacific Ocean SST data identified by SVD and climate signals explained a higher degree of SWE variance than reconstructions using tree-ring chronologies and climate data, indicating that Pacific Ocean SSTs and climate signals are useful predictors for snowpack reconstruction in the UGRB. This methodology is transferable to other regions.