Regional climate change and its impacts on water resources in Texas.

摘要

This dissertation aims to: (1) simulate different climate change scenarios of Texas for a 25-year period from 2006 to 2030; and (2) investigate the impacts of climate change on the availability of water resources in Texas. Specifically, this dissertation attempts to provide answers to the set of questions listed below. (1) What is the trend of climate change in Texas from 2006 to 2030? Would Texas become drier or wetter in these years? (2) Based on the simulated results of climate change in Texas, what is the consequent geographic distribution of changes in precipitation and temperature in Texas for the period of 2006--2030 compared to known historical data? (3) Based on results from integrated climate-hydrologic modeling and analysis, what are the spatial and temporal variations of the availability of water resources across Texas in the 25-year period? (4) Among the sixteen water planning regions in Texas, which ones will receive more rainfall and hence see more water, and which ones will experience less rainfall?;This dissertation uses an integrated climate-hydrologic modeling approach to answer the questions listed above. Due to coarse spatial resolutions, outputs from global climate model are inappropriate for hydrologic analysis at a regional level. Therefore, this dissertation uses a regional climate model, RegCM3.1, to produce high-resolution regional climate data over Texas by dynamically downscaling outputs from a global climate model. In addition, this study uses the Variable Infiltration Capacity (VIC) macroscale hydrologic model to generate surface runoffs based on results from the regional climate simulations.;The simulation procedure involved four steps: Firstly, an 18-year (1982--1999) Perfect Boundary Condition (PBC) Run was performed based on the NNRP reanalysis 2 data to evaluate and calibrate the performance of RegCM3.1. Secondly, continuous present-day 25-year (1971--1995) Control Run and future 25-year SRES A1B scenario (2006--2030) Future Scenario Run, over the Texas domain were conducted based on the NCAR CCSM3 6-hourly outputs. Thirdly, to remove the biases introduced by the global climate model and make simulation results applicable for subsequent hydrological modeling, this dissertation used a "quantile-based" bias correction approach to transform values of climate variables simulated by the Control Run and the Future Scenario Run. In the fourth and final step, the potential impacts of climate change on the availability of water resources in Texas were quantified by driving the VIC hydrologic model with bias-corrected current and future simulated climate conditions.;In the PBC Run, RegCM3.1 was found to perform well for the study area after calibration and optimization, especially on an annual basis. In the historical climate simulation based on the reanalysis data, RegCM3.1 simulates reasonable precipitation and surface air temperature over Texas. In the Control Run, output extracted from CCSM3 20th Century experiment during the period from 1971 to 1995 was used as the initial and lateral boundary conditions of RegCM3.1. By comparing with CRU observed data in the same period, the results suggested that the simulation of surface air temperature tends to be realistic, but uncertainties in precipitation still seem high. In the Future Scenario Run, CCSM3 simulated SRES A1B scenario served as the initial and lateral boundary conditions of RegCM3.1. Predictions of climate change for the 25 years from 2006 to 2030 were made by comparing the results between the Future Scenario Run and the Control Run.;Through dynamically downscaling climate scenarios to finer spatial and temporal scales, this dissertation demonstrated a range of possible climate change scenarios that may occur in Texas in the next two decades or so. Simulation results suggest that we will experience a generally warming trend in Texas for the 25 years from 2006 through 2030. In addition, results suggest that the magnitude of warming differ between the northern part and the southern part of Texas. Furthermore, analysis results indicate that annual precipitation in Texas will likely increase in the study period with some spatial variations across the state. The VIC model outputs based on simulated, bias corrected climate forcings demonstrate that there will be an increase in annual surface runoff in Texas on average, in the next 25 years. (Abstract shortened by UMI.)