Date Available

7-28-2021

Year of Publication

2020

Document Type

Doctoral Dissertation

Degree Name

Doctor of Philosophy (PhD)

College

Engineering

Department/School/Program

Civil Engineering

Advisor

Dr. James F. Fox

Abstract

Variance decomposition is the partitioning of different factors affecting the variance structure of a response variable. The present research focuses on future streamflow and sediment transport processes projections as the response variables. The authors propose using numerous climate factors and hydrological modeling factors that can cause any response variable to vary from historic to future conditions in any given watershed system. The climate modeling factors include global climate model, downscaling method, emission scenario, project phase, bias correction. The hydrological modeling factor includes hydrological model parametrization, and meteorological variable inclusion in the analysis. This research uses a wide spectrum of data, including climate data of precipitation and temperature from GCM results, and observations of meteorological data, streamflow and spring flow data, and sediment yield data. This research focuses on employing an off-the-shelf hydrological model and developing different numerical models (using MATLAB) for simulating sediment transport processes and water movement in an epigenetic karst system. With regards to variance decomposition, the approach is to use a mixed statistical method of linear and nonlinear analysis by means of analysis of variance (ANOVA) and artificial neural networks (ANN) respectively. All the computational tools that will be used to perform the statistics are provided by SPSS software.

Two study sites are considered in this work including South Elkhorn watershed and Cave Run watershed. South Elkhorn watershed is a fluvial system and is located in Lexington, Kentucky, USA. This system is characterized as a wet, temperate region in the central and eastern United States where a change in the climate is projected. The mean streamflow, extreme streamflow, and sediment processes forecast are investigated in this watershed. Royal Spring watershed is a fluviokarst system and is adjacent to the South Elkhorn watershed. In this watershed we investigate the water pathway connectivity as well as the impact of climate change on the mean annual spring flow and streamflow.

Analysis of variance results indicate that the difference in forecast and hindcast mean streamflow predictions is a function of GCM type, climate model project phase, and downscaling approach. Predicted average monthly change in streamflow tends to follow precipitation changes and result in a net increase in the average annual precipitation and streamflow by 10% and 11%, respectively, when comparing historical period (1980-2000) to the future period (2045-2065). Results show that the relative change of streamflow maxima was not dependent on systematic variance from the annual maxima method versus peak over threshold method. However, it was dependent all climate modeling factors. Ensemble projections forecast an increase of streamflow maxima of 51% for 100-year streamflow event. Hydrologic model parameterization was the greatest source of variance impacting forecasted sediment transport variables. Hydrologic inputs from climate change including forecasted precipitation, temperature, relative humidity, solar radiation and wind speed all impacted sediment transport. Ensemble average forecasts sediment yield to increase by 14% for the Upper South Elkhorn watershed. The numerical model of the Cave Run/ Royal Spring watershed suggests 30 to 45% of surface stream discharge originates from in-stream swallet reversal and hillside springs. Also, the hydrology of the floviokarst system might be altered by the impact of climate change where an increase in the surface flow and spring flow is projected to be 8.8% and 12.2%, respectively. The results show that the change in pathway connectivity is important on seasonal bases and follows the seasonal change in precipitations.

Digital Object Identifier (DOI)

https://doi.org/10.13023/etd.2020.319

Funding Information

National Science Foundation (NSF) (2017-2020) (#163288)

Higher Committee for Education Development (HCED) (2013-2018)

Download

Share

COinS