Document Type

Dissertation - Open Access

Award Date

2020

Degree Name

Doctor of Philosophy (PhD)

Department

Geospatial Science and Engineering

First Advisor

Xiaoyang Zhang

Keywords

greenness, MODIS, phenology, remote sensing, Wildfire

Abstract

Land surface phenology (LSP) characterizes the seasonal dynamics in the vegetation communities observed for a satellite pixel and it has been widely associated with global climate change. However, LSP and its long-term trend can be influenced by land disturbance events, which could greatly interrupt the LSP responses to climate change. Wildfire is one of the main disturbance agents in the western United States (US) forests, but its impacts on LSP have not been investigated yet. To gain a comprehensive understanding of the LSP responses to wildfires in the western US forests, this dissertation focused on three research objectives: (1) to perform a case study of wildfire impacts on LSP and its trend by comparing the burned and a reference area, (2) to investigate the distribution of wildfire impacts on LSP and identify control factors by analyzing all the wildfires across the western US forests, and (3) to quantify the contributions of land cover composition and other environmental factors to the spatial and interannual variations of LSP in a recently burned landscape. The results reveal that wildfires play a significant role in influencing spatial and interannual variations in LSP across the western US forests. First, the case study showed that the Hayman Fire significantly advanced the start of growing season (SOS) and caused an advancing SOS trend comparing with a delaying trend in the reference area. Second, summarizing >800 wildfires found that the shifts in LSP timing were divergent depending on individual wildfire events and burn severity. Moreover, wildfires showed a stronger impact on the end of growing season (EOS) than SOS. Last, LSP trends were interrupted by wildfires with the degree of impact largely dependent on the wildfire occurrence year. Third, LSP modeling showed that land cover composition, climate, and topography co-determine the LSP variations. Specifically, land cover composition and climate dominate the LSP spatial and interannual variations, respectively. Overall, this research improves the understanding of wildfire impacts on LSP and the underlying mechanism of various factors driving LSP. This research also provides a prototype that can be extended to investigate the impacts on LSP from other disturbances.

Format

application/pdf

Number of Pages

216

Publisher

South Dakota State University

Rights

In Copyright - Non-Commercial Use Permitted
http://rightsstatements.org/vocab/InC-NC/1.0/

Available for download on Friday, December 31, 2021

Share

COinS