Publication Cover
Canadian Journal of Remote Sensing
Journal canadien de télédétection
Volume 38, 2012 - Issue 3
94
Views
7
CrossRef citations to date
0
Altmetric
Research Article

Effects of incidence angles on mapping accuracy of surficial materials in the Umiujalik Lake area, Nunavut, using RADARSAT-2 polarimetric SAR images. Part 2. Polarimetric analysis

, , , , , & show all
Pages 404-423 | Received 05 Apr 2011, Accepted 13 Mar 2012, Published online: 05 Jun 2014
 

Abstract

Our study assesses the effect of incidence angle on classifications obtained using various polarimetric classifiers applied to polarimetric RADARSAT-2 synthetic aperature radar (SAR) images for mapping surficial materials in Arctic Canada. The RADARSAT-2 polarimetric SAR images were acquired over the Umiujalik Lake test area of Nunavut in three west-looking descending beam modes (FQ1, FQ12, and FQ20) with increasing respective incidence angles. Polarimetric analyses included computation of polarimetric signatures, Wishart supervised classification, as well as Wishart–H/, Wishart–H//A, and Freeman–Wishart unsupervised classifications. Polarimetric signatures helped to understand class separability as a function of the scattering mechanisms of the surficial materials considered in this study. The medium incidence angle (FQ12) image produced the best overall classification accuracy (48.7%) for the Wishart supervised classification. In general, the Freeman–Wishart unsupervised classification produced better areal distribution of surficial materials with the FQ12 and FQ20 images than with the steep angle FQ1 image. More sophisticated classification algorithms are required to combine the multibeam RADARSAT-2 polarimetric SAR images with other geospatial data such as optical images and digital elevation model data. The influences of variable environmental conditions (moisture and temperature) on mapping accuracy of surficial materials also require further research.

Notre étude évalue l'effet de l'angle d'incidence sur la classification d'images RSO polarimétriques RADARSAT-2, obtenue avec différents algorithmes de classification polarimétrique pour cartographier les dépôts de surface de l'Arctique canadien. Les images RSO polarimétriques RADARSAT-2 ont été acquises sur la zone-test du Lac Umiujalik au Nunavut, selon une orbite descendante avec une visée vers l'ouest selon trois modes (FQ1, FQ12, et FQ20) qui avaient des angles d'incidence croissants. L'analyse polarimétrique comprend le calcul des signatures polarimétriques, la classification supervisée Wishart et les classifications non supervisées Wishart–H/, Wishart–H//A et Freeman–Wishart. Les signatures polarimétriques aident à comprendre la séparabilité des classes en fonction des mécanismes de diffusion des dépôts de surface considérés dans l’étude. L'image acquise avec un angle d'incidence moyen (FQ12) a produit la meilleure précision globale de classification (48.7 %) avec la classification supervisée Wishart. En général, la classification non supervisée Freeman–Wishart a produit une meilleure distribution spatiale des dépôts de surface avec les images FQ12 et FQ20 qu'avec l'image d'angle d'incidence faible (FQ1). Des algorithmes de classification plus sophistiqués sont requis pour combiner les images RSO polarimétriques RADARSAT-2 avec d'autres données géospatiales, comme des images optiques et des données de modèle numérique de terrain.

Acknowledgements

The Geological Survey of Canada (GSC) funded this study as a contribution to the Northeast Thelon Compilation under the Northern Uranium for Canada and Remote Predictive Mapping (RPM) projects of the Geomapping for Energy and Minerals (GEM) Program. We thank the Canadian Space Agency for providing RADARSAT-2 data, Deborah Lemkow (GSC) for invaluable assistance in ordering the data, and Dr. Bill Morris at McMaster University for supporting development of the digital elevation model. Profs Brigitte Leblon and Dave Lentz are each supported by NSERC Discovery grants. This manuscript benefited from the input of two anonymous reviewers.

Reprints and Corporate Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

To request a reprint or corporate permissions for this article, please click on the relevant link below:

Academic Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

Obtain permissions instantly via Rightslink by clicking on the button below:

If you are unable to obtain permissions via Rightslink, please complete and submit this Permissions form. For more information, please visit our Permissions help page.