Identifying and Comparing CMIP5 Ensembles and the Resultant Hydrologic Conditions for the Colorado River Basin

Tuesday, 16 December 2014
Noe Isaac Santos1,2, William P Miller3, Thomas C Piechota2, Kenneth C Nowak1 and Daniel Bunk1, (1)Bureau of Reclamation Boulder City, Boulder City, NV, United States, (2)UNLV, Las Vegas, NV, United States, (3)NOAA Colorado Basin River Forecast Center, Salt Lake City, UT, United States
Climate projections for the Colorado River Basin (CRB) provide multiple scenarios that illustrate an uncertain and highly variable future. It is hypothesized that climate models which capture the historical sea surface temperature (SST) and climate relationships for the CRB might garner more weight or consideration when analyzing future conditions. This study analyzes the historical and downscaled Coupled Model Intercomparison Project 5 (CMIP5) climatology and hydrology data, along with SST data, using singular variable decomposition (SVD) to identify regions of significant correlation between the Pacific Ocean and the CRB.

SVD is a statistical tool capable of isolating important modes of variability when identifying coupled relationships between datasets involving gridded arrays and time series. SVD analysis was performed between historical SST and basin temperature, precipitation, and runoff conditions. The historical period analysis was performed prior to performing SVD analysis on all future climate projections to establish a baseline for the coupled SST and climate/hydrology relationship. An inter-model comparison reveals variable regions of significance and correlation magnitudes when all CMIP5 models are analyzed throughout the historical period. The models that captured certain historical features/relationships were then analyzed as a subset against the full ensemble to determine if any refinement or consensus in future projected conditions exists.

This study has identified a subset of CMIP5 models hypothesized to replicate the coupled SST and climate/hydrology conditions in the CRB. With this knowledge, a comparison between subset and ensemble model projections provides insight into the magnitude of the climate and hydrology projection differences for the CRB.

This study was performed to develop a subset of climate projections and to determine the differences between ensemble and subset climate model projections. The results of the study can potentially assist CRB resource managers in determining the applicability of climate and hydrology projections and reduce the uncertainty of future conditions. The outcome of this study will further assist with characterizing the range of possible climate, hydrology, and drought conditions in the Colorado River Basin.