GEOSCAN, résultats de la recherche

Menu GEOSCAN


TitreDeveloping and testing surficial sediments classification methods using remote predictive mapping, Repulse Bay area, Nunavut
AuteurWityk, U; Ross, M; McMartin, I; Campbell, J; Grunsky, E; Harris, J
Source33rd Canadian Symposium on Remote Sensing, abstracts; par Canadian Symposium on Remote Sensing; 2012 p. 57
Année2012
Séries alt.Secteur des sciences de la Terre, Contribution externe 20120002
Réunion33rd Canadian Symposium on Remote Sensing; Ottawa; CA; juin 11-14, 2012
Documentlivre
Lang.anglais
Mediaen ligne; numérique
Formatspdf
ProvinceNunavut
Sujetsdépôts glaciaires; elements glaciaires; techniques de cartographie; télédétection; imagerie par satellite; géophysique; géologie des dépôts meubles/géomorphologie; Cénozoïque; Quaternaire
ProgrammeGisements polymétalliques - Presqu'île Melville (Nunavut), GEM : La géocartographie de l'énergie et des minéraux
LiensAbstracts (PDF, 1.22 MB)
LiensOnline - En ligne
Résumé(disponible en anglais seulement)
New techniques are emerging to support the surficial geological mapping of vast northern regions at scales appropriate for mineral exploration and related land use management. Remote Predictive Mapping (RPM) techniques help streamline the process in the initial mapping stages by generating first-order maps from analysis and classification of satellite imagery and limited field observations. These techniques are applied to map the nature and distribution of surficial sediments west of Repulse Bay, Nunavut, located within an active mineral exploration area of the Western Churchill Geological Province. .
A multi-data approach using LANDSAT 7 TM and SPOT 4/5 imagery and field-based data is employed to determine the optimal combination of data to refine classification of surficial materials and maps with quantitative measures of accuracy. The imagery and data interpretation explore advantages of using different spatial and spectral resolutions.
Two methods are used to determine the optimal class combination and produce the most accurately classified maps. The first combines remote sensing with traditional surficial mapping to determine where confusion occurs. The second is based on statistics which calculate separability of classes prior to classification/predictive mapping as well as confusion between and within mapped surficial material types after the classification. Both methods are used and compared to discover optimal class combinations - the first method qualitatively, and the second quantitatively. Preliminary statistical results illustrate that exposed marine sediments, carbonate-rich tills, organics and boulder terrains are most accurately classified; while confusion occurs between remaining tills, sand and gravel and bedrock. Grouping of non-carbonate till sub-classes into a single till class has generated maps that are more geologically consistent.
GEOSCAN ID290979