Polarimetric scattering from shallow firn and forests with snow cover
2010 (English)In: Proceedings of the ESA Living Planet SymposiumBergen, Norway: (ESA SP-686, December 2010) / [ed] H. Lacoste-Francis, Noordwijk, Netherlands: European Space Agency , 2010Conference paper (Other academic)
In this paper the potential for inferring shallow firn depth from polarimetric SAR (PolSAR) data at L- and C-band is investigated. Using ALOS PALSAR and Radarsat-2 SAR imagery, and field data including Ground Penetrating Radar profiles and shallow cores, we investigate the spatial distribution of backscatter and decompose backscatter using polarimetric methods to analyse how polarimetric scattering is affected by firn depth near the firn line. The investigation is aimed at a more refined delineation of glacier firn lines and a better understanding of scattering from firn, superimposed ice and the bare ice facies. We found that PolSAR can be used to infer shallow firn thicknesses up to depths of at least 2 m water equivalent (m w.e.) and that old and contemporary firn surfaces can be differentiated using PolSAR. Contrary to many previous investigations the importance of surface scattering in the firn area is also emphasised in the scattering decompositions. Volume scattering was found to have a secondary or tertiary importance. This has important implications for the analysis of backscatter using semi-empirical models.The effect of snow depth on backscatter in pro-glacial a sub-Arctic forest and its potential for improving forest mapping is also discussed. Snow depth data were acquired by manual probing and snowpit measurements. In addition forest stand densities were assessed in situ and NDVI and tasseled cap transformations were made in optical remote sensing data (SPOT-4) to parameterise the forest. Scatterer decomposition and pedestal height products were found to be related to snowpack depth. It was not possible to separate the influences of snow cover and forest structure due to the partial dependence of the former on the latter. Nevertheless it can be concluded that PolSAR improves our ability to map the forest margins of low density, sub-Arctic forests. Our findings have implications for the implementation of algorithms for the exploitation of future SAR missions including Sentinel-1.
Place, publisher, year, edition, pages
Noordwijk, Netherlands: European Space Agency , 2010.
polarimetry, synthetic aperture radar, snow, firn
Research subject Physical Geography
IdentifiersURN: urn:nbn:se:su:diva-53592ISBN: 978-92-9221-250-6OAI: oai:DiVA.org:su-53592DiVA: diva2:391016
ESA Living Planet Symposium
Presented in the Polarimetry Session of the conference. Funded by Rymdstyrelsen and C.F. von Horn Foundation.2011-01-242011-01-242011-01-24Bibliographically approved