Mapping forest alliances and associations using fuzzy systems and nearest neighbor classifiers

Abstract

This is an author-produced, peer-reviewed version of this article. © 2009, Elsevier. Licensed under the Creative Commons AttributionNonCommercial-NoDerivatives 4.0 International License (https://creativecommons.org/licenses/by-nc-nd/4.0/). The final, definitive version of this document can be found online at Remote Sensing of Environment, doi: 10.1016Environment, doi: 10. /j.rse.2007 1 NOTICE: This is the author's version of a work accepted for publication by Elsevier. Changes resulting from the publishing process, including peer review, editing, corrections, structural formatting and other quality control mechanisms, may not be reflected in this document.Changes may have been made to this work since it was submitted for publication. AbstractThe study and management of biological communities depends on systems of classification and mapping for the organization and communication of resource information. Recent advances in remote sensing technology may enable the mapping forest plant associations using image classification techniques. But few areas outside Europe have alliances and associations described in detail sufficient to support remote sensing-based modeling. Northwestern Montana has one of the few plant association treatments in the United States compliant with the recently established National Vegetation Classification system. This project examined the feasibility of mapping forest plant associations using Landsat Enhanced Thematic Mapper data and advanced remote sensing technology and image classification techniques.Suitable reference data were selected from an extensive regional database of plot records. Fifteen percent of the plot samples were reserved for validation of map products, the remainder of plots designated as training data for map modeling. Key differentia for image classification were identified from a suite of spectral and biophysical variables. Fuzzy rules were formulated for partitioning physiognomic classes in the upper levels of our image classification hierarchy. Nearest neighbor classifiers were developed for classification of lower levels, the alliances and associations, where spectral and biophysical contrasts are less distinct.Maps were produced to reflect nine forest alliances and 24 associations across the study area. Error matrices were constructed for each map based on stratified random selections of map validation samples. Accuracy for the alliance map was estimated at 60%. Association classifiers provide between 54 and 86% accuracy within their respective alliances. Alternative techniques are proposed for aggregating classes and enhancing decision tree classifiers to model alliances and associations for interior forest types.

Tree Species Classification in Temperate Forests Using Formosat-2 Satellite Image Time Series

“…This result was not necessarily expected, the pixel size (8 m) being small enough to detect individual tree and thus, the variety of species within a single stand. Alternative approaches based on soft classification [66], spectral unmixing [67] or the use of small training sets containing mixed pixels [61] could be used to account for uncertainty in these forests. Finally, our evaluation method is independent, in the sense that pixels used to train and validate the classifications are strictly distinct.…”

Recognising fuzzy vegetation pattern: the spatial prediction of floristically defined fuzzy communities using species distribution modelling methods

“…While there is a long history of floristic analysis, the mapping of patterns derived through such analysis remains a challenge (Ohmann & Gregory ; Triepke et al. ). Fuzzy communities can overlap, and a single point in space may share the properties of multiple communities.…”

Using climate projections to assess ecosystem vulnerability at scales relevant to managers

“…Insolation provides an inference of physical site variables including incoming energy, the primary driver for ecological processes and a strong predictor of vegetation potential (Dubayah and Rich ; Dubayah and Rich , Triepke et al. ). Solar insolation values were derived from a tri‐shade surface of the region, where each pixel was represented by growing season solar input as a computation of three sun angles in the spring, summer, and autumn (ESRI ).…”