Publications

Keck, T; Preusker, R; Fischer, J (2017). Retrieving snow and ice characteristics by remotely sensed emissivity using the multi-view brightness temperature within 8 mu m to 14 mu m. REMOTE SENSING OF ENVIRONMENT, 201, 181-195.

Abstract
This study provides a method determining the characteristics of snow and ice with remote sensing by converting top of atmosphere brightness temperature to surface emissivity which depends on wavelength and observation angle with respect to snow and ice properties. Envisat's Advanced Along-Track Spectral Radiometer (AATSR) features dual-view thermal infrared measured brightness temperature. Using bands 11 mu m and 12 mu m in nadir and forward (55) view and a total column water vapour (TCWV) product from Medium Resolution Imaging Spectrometer (MERIS), we obtain four measurements per pixel. We retrieve three surface emissivities epsilon(lambda,theta) per pixel calculated from simulated surface temperatures related to 11 mu m/nadir view values. We define emissivity "classes" for different snow grain sizes from angular and spectral field measurements of snow grain size and emissivity above snow and ice from Hori et al. (2006, 2007): fine, medium, coarse, suncrust, and ice. Remaining pixels are either indistinct between classes, unclassified, or invalid. Temperatures above the freezing point label pixels as wet. Analysing 26 AATSR scenes in 2007 and 2008, we retrieve a high portion of classification in Greenland, Antarctica, and the sea ice of the Hudson Bay. Steep and heterogeneous topography may cause invalid and unclassified pixels in Eurasia. Close to coasts we generally find a higher snow temperature and a large number of wet pixels.

DOI:
10.1016/j.rse.2017.09.006

ISSN:
0034-4257