Publication Date
5-2025
Date of Final Oral Examination (Defense)
2-24-2025
Type of Culminating Activity
Dissertation
Degree Title
Doctor of Philosophy in Ecology, Evolution, and Behavior
Department
Biological Sciences
Supervisory Committee Chair
Megan Cattau, Ph.D.
Supervisory Committee Member
Nancy Glenn, Ph.D.
Supervisory Committee Member
Alejandro Flores, Ph.D.
Supervisory Committee Member
Ernest Keeley, Ph.D.
Abstract
Western United States ecosystems are undergoing transformations driven by a shifting climate and associated disturbances. Remote sensing data and ecohydrologic models are essential for assessing these impacts and developing management strategies that are responsive to changing conditions. This dissertation investigates vegetation cover through remote sensing approaches and hydrological processes using ecohydrologic models from the hillslope to the watershed scale in Idaho, USA. In the first chapter, I utilized high-resolution multispectral imagery acquired from Unoccupied Aerial Systems (UAS) to quantify vegetation cover (litter, herbaceous, bare ground, shrub, and tree) within two sagebrush communities within the Reynolds Creek Experimental Watershed, Idaho. A disagreement was observed between fractional vegetation cover estimates derived from UAS imagery and those obtained from coarser-resolution Landsat satellite imagery. UAS provided a higher accuracy for rangeland fractional cover estimation compared to those satellite-derived products. This finding highlights the potential benefits of employing UAS imagery for more accurate assessments of semi-arid vegetation cover at the hillslope scale. In the second chapter, I integrated high-resolution vegetation cover and a physics-based hydrological model, the Rangeland Hydrology and Erosion Model, to quantify post-fire changes in runoff and erosion at the hillslope scale within the Johnston Draw, a headwater watershed in the study area, after a prescribed fire. The model results demonstrated an increase in runoff and soil loss immediately following the fire event. However, these post-fire impacts were observed to decline eight months after the burn. High-intensity rainfall events occurring immediately following a fire generate the highest runoff and erosion, posing the greatest risk to watershed health and demanding critical management intervention. In the third chapter, I developed and evaluated Random Forest models for predicting daily summer stream temperatures at 23 United States Geological Survey stream gages within Idaho. These models exhibited good predictive accuracy (RMSE between 0.4 to 1.2 ℃). To gain deeper insights into the factors driving stream temperature variability, a paired air and stream temperature signal analysis was employed to quantify the contributions of groundwater and dam influences on the observed stream temperature patterns. High-elevation, small streams have low amplitude ratios, indicating greater thermal buffering capacity. This dissertation develops novel methods for: (1) monitoring vegetation cover in semi-arid environments; (2) assessing the impacts of prescribed fire on hillslope runoff and erosion; and (3) developing predictive models for daily summer stream temperature to support cold-water fish habitat management.
DOI
10.18122/td.2370.boisestate
Recommended Citation
Huang, Tao, "Assessing Semi-Arid Vegetation Cover, Post-Fire Hydrology, and Stream Temperature Using Remote Sensing and Modeling" (2025). Boise State University Theses and Dissertations. 2370.
10.18122/td.2370.boisestate
Comments
https://orcid.org/0000-0001-9004-6130