Impacts of climate change on the hydrologic response of headwater basins in Colorado

Abstract

The headwater basins of Colorado are heavily relied upon for freshwater resources on an annual basis. However, knowledge concerning generation of such resources, and implications of climate change on their availability in the future, is not well understood. Thus, this research has been undertaken to develop, calibrate, and test a comprehensive process-based model in four mountainous watersheds of Colorado, and investigate the potential impacts of changing climate on hydrologic response in these basins. Specifically, the four study watersheds considered for analysis include the Cache la Poudre, Gunnison, San Juan and Yampa River basins. Calibration of the model compared several parameter optimization techniques for performance in each of the study basins, which included the more common Shuffled Complex Evolution – University of Arizona (SCE-UA) method and a Markov Chain Monte Carlo (MCMC) method known as the Gibbs Sampler Algorithm (GSA). Fully calibrated and tested models were driven by a suite of 112 climate projections, downscaled both spatially and temporally, and were run on a daily time-step for a period of 90 years from 2010 – 2099. Results from model calibration indicate GSA outperformed SCE-UA in a majority of the study basins, in addition to revealing promising results from a two-stage method that combined the strengths of the two techniques. Error statistics showed very good (Nash-Sutcliffe coefficient of efficiency >0.75 and relative error <+/-10%) performance of monthly streamflow simulations compared to naturalized flows at the outlet of each watershed over a period of 16 years (1990 – 2005). Additionally, the models provided satisfactory results for simulating monthly streamflow at multiple sites nested within each watershed, which increased confidence in model parameterization and representation of dominant hydrologic processes. Results indicate that on an average annual basis, 55% – 65% of precipitation goes to evapotranspiration, and lateral flow contributes to between 64% and 82% of gross water yield. Results from future simulations over the course of the 21st century indicate inconsistent responses in streamflow to increasing temperature and variable precipitation projections. However, results did show consistency in the Yampa River basin, where 71 out of 112 future projections resulted in statistically significant (α<0.1) positive trends of average annual streamflow. Furthermore, all study basins exhibited a decreasing ratio of precipitation to potential evapotranspiration from emissions scenario ensemble averages, which suggest Colorado basins will become more arid over the 21st century. Future forecasting of water availability in Colorado may benefit from this research, as specific climate projections were provided that resulted in consistent responses (increasing and decreasing) in streamflow across all watersheds. Implications of this study are considerable, as management of water resources, both within the state and across the West, will be affected by freshwater availability in headwater basins of Colorado in the future.