Skip to main content
U.S. flag

An official website of the United States government

Analyzing Landsat time-series data across adjacent path/rows and across multiple cycles of FIA: Lessons learned in southern Missouri

Informally Refereed

Abstract

The North American Forest Dynamics (NAFD) Program is assessing disturbance and regrowth in the forests of the continent. These forest dynamics are interpreted from per-pixel estimates of forest biomass, which are produced for a time series of Landsat 5 Thematic Mapper (TM) and Landsat 7 Enhanced TM Plus images. Image data are combined with sample plot data from the Forest Inventory and Analysis (FIA) program using Random Forests, a tree-based estimation method implemented here in the R statistical environment. The NAFD approach is based on a sample of image Path/Rows, resulting in most images being disjunct from and independent of other images in the sample. Increases in sample intensity and needs for assessing forest dynamics over geographic extents larger than a single image are leading to increased frequency of adjacent, overlapping images in the sample. We assessed the consistency of estimates of forest biomass and classification of forest/nonforest in southern Missouri, USA, across space and time, for adjacent images in Path 25/Row 34 and Path 24/Row 34, and for coincident images in Path 25/Row 34 acquired in 2000 and 2007. Results were consistent across space and time, implying consistency of both Landsat and FIA data, and supporting the NAFD image sample strategy and subsequent augmentation with overlapping images.

Parent Publication

Keywords

North American Forest Dynamics, NAFD, Landsat, consistency analyses, forest inventory, FIA

Citation

Nelson, Mark; Healey, Sean; Moser, W. Keith; Hansen, Mark; Cohen, Warren; Hatfield, Mark; Thomas, Nancy; Masek, Jeff. 2009. Analyzing Landsat time-series data across adjacent path/rows and across multiple cycles of FIA: Lessons learned in southern Missouri. In: McWilliams, Will; Moisen, Gretchen; Czaplewski, Ray, comps. Forest Inventory and Analysis (FIA) Symposium 2008; October 21-23, 2008; Park City, UT. Proc. RMRS-P-56CD. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 11 p.
https://www.fs.usda.gov/research/treesearch/33340