Award Date

December 2015

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Environmental Science

First Committee Member

Haroon Stephen

Second Committee Member

Krystyna A. Stave

Third Committee Member

Helen R. Neill

Fourth Committee Member

David K. Kreamer

Number of Pages

248

Abstract

Recent variations in meteorological conditions indicate the earth’s climate is changing in ways that can impact delicate ecological balances in sensitive regions. These impacts threaten the essential services provided by such ecosystems. Determining how climate changes are affecting the biosphere is essential to adapt and mitigate harmful consequences. In order to mitigate the negative effects of climate change and adapt to shifting ecological resource constraints, it is imperative to locate such changes and determine vulnerability of ecological resources to changing environmental conditions.

Identifying climate driven ecological changes faces numerous challenges given the reliance on vegetation indices as the primary measure of vegetative surface cover. It is essential that gaps in the understanding and reliability of these indices be addressed. In this research, the problem of determining climate driven ecological change is addressed using remote sensing techniques to find trends in an ecologically sensitive region over the last 30 years. Relation between climate and vegetation trends is studied and a methodology that does not rely upon a single measure for determining vegetative changes is developed. The comprehension of the scenario when both the red and near-infrared bands shift in the same direction is developed and related to compositional changes. The performance of vegetation indices and transforms is determined using survey data. The non-parametric Mann-Kendall (MK) trend test is used to establish the presence of trends. The research study is conducted in the alpine ecosystem.

New insights into climate driven changes were gained using a novel study design that incorporated several vegetation indices, tasseled cap transforms and spectral mixture analysis. Surface reflectance throughout the watershed has declined in all six Landsat bands over the last three decades. At the same time, temperatures have demonstrated a statistically significant rise. Vegetative composition changes throughout the study area were identified including widespread declines in needle leaf shrubs. Composition changes are related to red-shift translations as a result of variations in the vegetative structural density. An elevational relationship was found in sparsely vegetated areas with declining vegetation in the lower half of the watershed and vegetative increases in the upper half of the watershed

Correlation of surface reflectance red-shift trends to actual changes in vegetative surface cover was tested against long-term field survey data. Red-shift stretch and downward translation demonstrated close agreement with predicted vegetation increases while red-shift compression and upward translation was not as effective a predictor of vegetative declines. Vegetation indices and tasseled cap brightness and greenness transforms performed remarkably well in predicting vegetation trends. Moreover, a poor performance of the tasseled cap wetness index in predicting vegetation trends is revealed. An additional important finding of this research is the close correlation between the USGS CDR surface reflectance dataset and older dark object subtraction methods for calculating surface reflectance.

It is essential to understand the impact of climate change on the environment in order to deal with potentially severe and rapid consequences brought about by reduced ecosystem productivity. This research provides a useful insight about ecological impacts of climate change using remote sensing.

Keywords

Climate Change; Remote Sensing

Disciplines

Environmental Sciences

Language

English


Share

COinS