Understand and Quantify Long-Term Hydroclimate Trends in the Highland
Lakes Watersheds via Spatially Distributed Hydrological Modeling
Abstract
The Highland Lakes on the Colorado River in Texas, specifically lakes
Buchanan and Travis, supply water for the Austin metropolitan area. The
population of the Austin area is projected to reach 4.5 million by 2040,
which will lead to an increased water demand. The goal of this study is
to generate a long-term hydroclimate dataset using hydrological
modeling, and to examine the observed trends and their causes to help
understand the changes in runoff in the lower Colorado River basin. The
lower Colorado River basin includes the main stem of the Colorado River
and four sub-watersheds – Pecan Bayou, San Saba River, Llano River and
Pedernales River. The Distributed Hydrology Soil Vegetation Model
(DHSVM) includes a 300-meter spatial resolution and a 12-hour time step
from 1950-2018 to simulate soil moisture, evapotranspiration, runoff and
streamflow. The model parameters are calibrated during the period of
1951-1990 by comparing the simulated streamflow with observed
naturalized flows using three statistical criteria: relative bias (RB),
coefficient of determination (R2), and the
Nash-Sutcliffe efficiency (NSE). The simulated streamflows are further
validated against naturalized flows over the period of 1991-2015. The
streamflow trends, along with other hydroclimate variables, are examined
and compared across all watersheds during the study period. In addition,
sensitivity analyses are conducted to evaluate how changing temperatures
and precipitation will affect streamflow in the watersheds. This
long-term hydroclimate dataset for the lower Colorado River basin not
only provides a comprehensive understanding of historical trends, but
also serves as a baseline for future analysis.