CGRG Bibliography of Canadian Geomorphology
Search Results
Author : Fraser, J.Z.
Date : 2004.
Title : Simulation of landscape succession dynamics based on delineation of Integrated Terrain Units: Oak Ridges Moraine, Ontario.
Publication : 49th Annual Meeting of the Geological Association and the Mineralogical Association of Canada. May 12-14, 2004. Brock University, St. Catharines, Ontario.
Issue :
Page(s) :
Abstract
The application of Ecological Land Classification (ELC) systems has become increasingly valuable for life science inventories, natural resource planning and landscape scale restoration initiatives. Due to the decreasing cost of remote sensed imagery, and the demand for current data for ecological models, the development of efficient and cost-effective image classification techniques has become an important area of study. However, there are a number of challenges in generating remotely-sensed ELC interpretations including, evaluating accuracy across multiple processes to effectively integrate remote sensed image classification with other more finely resolved interpretive information such as air photo interpretation and site based field investigation of landscape components. We report here on a methodology developed for classifying ELC units in southern Ontario using a 'top-down' remote-sensing approach and a 'bottom-up' spatially-nested Ecological Land Classification field sample design. We have adopted knowledge-based, spatial decision-making approaches in creation of an ecological classification level derived from geomorphic terrain pattern analysis applied to digital elevation data. First, an unsupervised classification is performed for ELC units using mid-resolution (5 meter IRS) satellite imagery. These units are further delineated using interpretation techniques including: 1) detailed aerial photograph interpretation 2) delineation of geomorphically defined terrain units based on feature extraction from data including elevation models and hydrological information, and 3) deriving temporal changes from Normalized Difference Vegetation Index (NDVI) values. The automated delineation of terrain units and the accuracy of the classification are evaluated through air photo interpretation and field evaluation which then serves as the basis for further refinement of classification procedures and decision-making algorithms.
Bibliography of Canadian Geomorphology