Date of Award

Spring 2023

Project Type

Thesis

Program or Major

Earth and Environmental Sciences

Degree Name

Master of Science

First Advisor

Anne Lightbody

Second Advisor

Michael Palace

Third Advisor

Shane Csiki

Abstract

Understanding a river's morphology can help scientists measure erosion, improve understanding of flood hazards and better estimate riparian habitat. While there have been many developments in the field of fluvial geomorphology, there are still many challenges that exist in efficiently and accurately quantifying various important river metrics, such as the location of channel banks. Conducting field surveys of riverbanks is expensive and time consuming, leading many scientists to shift towards remote sensing techniques to measure river morphological features. This project aims to extract bankfull width measurements automatically from high-resolution Digital Elevation Models (DEMs) of rivers located in New Hampshire and Vermont. Six rivers located throughout Vermont and New Hampshire were identified, based on the availability of previous channel field surveys and high-resolution DEM data. The USGS stream channel and floodplain toolbox was used as well as a combination of other tools such as Terrain Analysis Using Digital Elevation Models (Tau DEM) and Whitebox Geospatial Analysis Tool (Whitebox GAT), to automatically generate bank points and bankfull width estimates of these six rivers. Specific input parameters for each river were evaluated to produce accurate widths. The best parameters were used to create a model to rapidly estimate bankfull widths of New Hampshire and Vermont rivers with similar topographic signatures. This study found that 49% of bankfull widths derived from remote sensing were within the 2.95 meters of estimated uncertainty among all study rivers. However, no statistically significant difference between bankfull width and bed substrate or bank substrate was found among all the rivers.

Download

Share

COinS