thumbnail

Multi-scale remote sensing sagebrush characterization with regression trees over Wyoming, USA: laying a foundation for monitoring

International Journal of Applied Earth Observation and Geoinformation

By:
, , , and
DOI: 10.1016/j.jag.2011.09.012

Links

Abstract

agebrush ecosystems in North America have experienced extensive degradation since European settlement. Further degradation continues from exotic invasive plants, altered fire frequency, intensive grazing practices, oil and gas development, and climate change – adding urgency to the need for ecosystem-wide understanding. Remote sensing is often identified as a key information source to facilitate ecosystem-wide characterization, monitoring, and analysis; however, approaches that characterize sagebrush with sufficient and accurate local detail across large enough areas to support this paradigm are unavailable. We describe the development of a new remote sensing sagebrush characterization approach for the state of Wyoming, U.S.A. This approach integrates 2.4 m QuickBird, 30 m Landsat TM, and 56 m AWiFS imagery into the characterization of four primary continuous field components including percent bare ground, percent herbaceous cover, percent litter, and percent shrub, and four secondary components including percent sagebrush (Artemisia spp.), percent big sagebrush (Artemisia tridentata), percent Wyoming sagebrush (Artemisia tridentata Wyomingensis), and shrub height using a regression tree. According to an independent accuracy assessment, primary component root mean square error (RMSE) values ranged from 4.90 to 10.16 for 2.4 m QuickBird, 6.01 to 15.54 for 30 m Landsat, and 6.97 to 16.14 for 56 m AWiFS. Shrub and herbaceous components outperformed the current data standard called LANDFIRE, with a shrub RMSE value of 6.04 versus 12.64 and a herbaceous component RMSE value of 12.89 versus 14.63. This approach offers new advancements in sagebrush characterization from remote sensing and provides a foundation to quantitatively monitor these components into the future.

Geospatial Extents

Additional Publication Details

Publication type:
Article
Publication Subtype:
Journal Article
Title:
Multi-scale remote sensing sagebrush characterization with regression trees over Wyoming, USA: laying a foundation for monitoring
Series title:
International Journal of Applied Earth Observation and Geoinformation
DOI:
10.1016/j.jag.2011.09.012
Volume
14
Issue:
1
Year Published:
2012
Language:
English
Publisher:
Elsevier
Contributing office(s):
Earth Resources Observation and Science (EROS) Center
Description:
12 p.
Larger Work Type:
Article
Larger Work Subtype:
Journal Article
Larger Work Title:
International Journal of Applied Earth Observation and Geoinformation
First page:
233
Last page:
244
Country:
United States
State:
Wyoming