Abstract
In this paper, I describe a set of procedures that automate forest disturbance mapping using a pair of Landsat images. The approach is built on the traditional pair-wise change detection method, but is designed to extract training data without user interaction and uses a robust classification algorithm capable of handling incorrectly labeled training data. The steps in this procedure include: i) creating masks for water, non-forested areas, clouds, and cloud shadows; ii) identifying training pixels whose value is above or below a threshold defined by the number of standard deviations from the mean value of the histograms generated from local windows in the short-wave infrared (SWIR) difference image; iii) filtering the original training data through a number of classification algorithms using an n-fold cross validation to eliminate mislabeled training samples; and finally, iv) mapping forest disturbance using a supervised classification algorithm. When applied to 17 Landsat footprints across the U.S. at five-year intervals between 1985 and 2010, the proposed approach produced forest disturbance maps with 80 to 95% overall accuracy, comparable to those obtained from traditional approaches to forest change detection. The primary sources of mis-classification errors included inaccurate identification of forests (errors of commission), issues related to the land/water mask, and clouds and cloud shadows missed during image screening. The approach requires images from the peak growing season, at least for the deciduous forest sites, and cannot readily distinguish forest harvest from natural disturbances or other types of land cover change. The accuracy of detecting forest disturbance diminishes with the number of years between the images that make up the image pair. Nevertheless, the relatively high accuracies, little or no user input needed for processing, speed of map production, and simplicity of the approach make the new method especially practical for forest cover change analysis over very large regions.
Highlights
Rapid assessment of forest disturbance is an important source of information for studies ranging from global environmental change to local forest management planning [1], [2]
The approach relies on the ability to automatically capture this signal without user interaction and use it in a robust supervised classification algorithm that can handle incorrectly labeled training data
Accuracy of Classifications The evaluation of the forest change maps derived from the automated procedure includes an assessment of the overall map accuracy, which measures the proportion of individual footprint area that is classified correctly into two categories
Summary
Rapid assessment of forest disturbance is an important source of information for studies ranging from global environmental change to local forest management planning [1], [2]. When viewed from a disturbance perspective, changes in forested areas are important for understanding fluxes of carbon and water between the biosphere and the atmosphere. North American forests have been viewed as a net carbon sink, the magnitude of the sink is uncertain and requires careful assessment of land use history including harvest [3], [4], [5], [6], [7]. Either alone or in combination with field studies, has made great contributions to documenting land use history in forested landscapes [8]. The literature on remote sensing-based change detection studies in forested areas is rich with a wide range of case studies, methods, and applications in virtually every type of environment [9], [10], [11], [12], [13], [14], [15]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
