Date of Completion

9-14-2015

Embargo Period

3-12-2016

Keywords

Hydrological Modeling, climate change, flood, drought, northeast, precipitation intensity, evapotranspiration, VIC, clm

Major Advisor

Guiling Wang

Associate Advisor

Christine Kirchhoff

Associate Advisor

Emmanouil N. Anagnostou

Associate Advisor

Amvrossios C. Bagtzoglou

Associate Advisor

David Bjerklie

Field of Study

Environmental Engineering

Degree

Doctor of Philosophy

Open Access

Open Access

Abstract

The objective of this dissertation research is to better understand the hydrological impacts of climate variability and climate change. This objective is first addressed in a two-part study focusing on the Northeast US using the Connecticut River Basin as a case study. Changes to the hydrological cycle are investigated for the past several decades using precipitation and river discharge data from observations and soil moisture and evapotranspiration (ET) from the VIC hydrological model. From 1950-2011 a clear increase of precipitation intensity is identified, together with increasing precipitation amount, discharge, runoff ratios, and soil moisture. The ET trend is negligible. This study of the past is followed by projections of the future using the VIC model driven by a bias-corrected climate for the period of 2046-2065 from three climate models. The projected future changes that had not yet manifested in the past include enhanced ET for all four seasons and a change to the seasonality of snow melt and discharge. There are also indications of wetter winters, changing characteristics of flood events, and a consistently increasing mean intensity of precipitation which continues from the past analysis. Compared to the past, the future foods are projected to be less frequent but last longer.

Among all hydrological variable, ET is the most difficult to simulate. In this dissertation research, an innovative approach to improving the accuracy of ET estimations is developed, which combines hydrological models with data derived from satellite remote sensing including leaf area index and ET. This model-data integration leads to a more accurate reconstruction of historic river flow and different future hydrological trends that include an increase of summer droughts.

This dissertation research also explores the mechanisms underlying the recently discovered decline of the ET trend in many regions focusing on the continental U.S. using the Community Land Model 4.5. Experimental simulations are conducted to isolate the effects of the most influential factors on ET. It is found that the changing characteristics of precipitation, precipitation amount in particular, are the primary cause of the ET trend decline. The roles of wind speed and temperature changes are found to be negligible.

COinS