PARENT SESSION
Oral Session 7: GIS / Remote Sensing I.
Presiding: E Ellis and P Valko
Monday, August 2, 8:00 AM to 11:30 AM, Meeting Room B 116.
An approach to examine woody plant proliferation using satellite imagery and ancillary data.
Xiao, Jingfeng*,1, Moody, Aaron1, 1 Department of Geography, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
ABSTRACT- The invasion of trees and shrubs into grasslands, or woody plant proliferation, is a common phenomenon in arid and semi-arid regions in the Southwest US. Woody plant proliferation has caused a significant carbon sink, and hence has important implications for the carbon cycle. Satellite remote sensing provides a valuable tool for studying woody plant proliferation due to its consistent spatial and temporal coverage. We explored an approach to examine the rate and spatial extent of woody plant proliferation by combining multitemporal Landsat TM/ETM+ imagery, high-resolution orthoimagery, and spectral mixture analysis. A multiple-endmember spectral mixture model was used to estimate the fraction of vegetation within each TM/ETM+ pixel. A phenological logic along with spectral mixture analysis was used to differentiate woody plants from herbaceous vegetation within each pixel. We then examined the rate of woody plant proliferation during the past decade using the estimated fractions of woody and herbaceous vegetation from the early 1990s to present. High-resolution orthoimagery and air photos were used for both selection of training areas and validation of the results. Our results have implications for improving our understanding of the rate and spatial extent of woody plant proliferation and net changes in regional woody plant cover in arid and semi-arid ecosystems in the Southwest US.
Key words: remote sensing, spectral mixture analysis, woody plant proliferation, Landsat