CGRG Bibliography of Canadian Geomorphology
Search Results
Author : Brown, O.; Utting, D.; Little, E.; Grunsky, E.; and Harris, J.
Date : 2005.
Title : Remote predictive and enhanced surface-features mapping using analysis techniques on RADARSAT-I and LANDSAT TM data on northern Baffin Island, Nunavut, Canada.
Publication : 35th Annual International Arctic Workshop. March 9-12, 2005. Timms Centre for the Arts, University of Alberta, Edmonton, Alberta, Canada.
Issue :
Page(s) :
Abstract
Developments in analysis techniques for remotely sensed data have significant potential for making field work more efficient and cost-effective. Analysis of remotely sensed data has been undertaken by the Canada—Nunavut Geoscience Office, in collaboration with the Geological Survey of Canada, to provide field geologists with information suitable for advanced planning of the North Baffin Geoscience Project. Supervised classifications of LANDSAT TM (30m resolution) and multi-beam RADARSAT-1 (25m resolution) data have been conducted to help predict surficial material. Principal Components Analysis (PCA) of multi-beam RADARSAT-1 data has been conducted to produce enhanced surface-features maps to assist with preliminary interpretations of the glacial history. Supervised classifications were first conducted in the northwest corner of NTS 37G, which has been mapped using traditional means (i.e., air photos and ground truthing). Supervised classification techniques were able to discern areas of till, colluvium, alluvium and water. An “Other” class represented bedrock, glaciolacutrine, glaciofluvial and marine sediments because these materials were not easily differentiated. A number of different classifications techniques and combinations of data were tested. The predictive map was then compared to a simplified version of the surficial geology map derived through traditional means. Similarities and differences between the resulting maps were visually identified. Most of the discrepancies occur because spectral characteristics of material classes may overlap due to similarities in material properties. Also, supervised classification techniques classify “pixel by pixel”, whereas traditional mapping groups areas of relatively homogeneous materials into polygons. Spectral signatures, classification techniques and data combinations that produced the best-fit predictive map in the test area will be used to make remote predictive maps in contiguous map areas NTS 37E and 37H. PCA was conducted on RADARSAT-1 standard beam mode imagery to highlight surface-features and assist in interpreting the glacial history of the area. Different filtering techniques and histogram stretches will be examined to determine which method best accentuates surface features. The component images will be combined with the predictive surficial geology maps, LANDSAT TM imagery and air photo mosaics to assist with preliminary interpretations of glacial history and ice-movement chronology that will be tested during the field component of the program. Through the use of these two predictive mapping methods, preliminary glacial histories can be quickly interpreted at the regional scale greatly assisting the geologists during advanced planning stages of field operations.
Bibliography of Canadian Geomorphology