作者: Claire E. Bulgin , Steinar Eastwood , Owen Embury , Christopher J. Merchant , Craig Donlon
DOI: 10.1016/J.RSE.2013.11.022
关键词:
摘要: Abstract We present a Bayesian image classification scheme for discriminating cloud, clear and sea-ice observations at high latitudes to improve identification of areas clear-sky over ice-free ocean SST retrieval. validate the against manually classified dataset using Advanced Along Track Scanning Radiometer (AATSR) data. A three-way near-infrared textural feature improves classifier accuracy by 9.9% nadir only version cloud clearing used in ATSR Reprocessing Climate (ARC) project latitude regions. The gives similar numbers ice scenes misclassified as but significantly more cases are correctly identified (89.9% compared with 65% ARC). also demonstrate potential including information from 0.6 μm channel be extent surface temperature retrieval 77.7% an overall 96%.