A Review of the Applications of Sar Polarimetry and Polarimetric Interferometry–An Esa Funded Study (original) (raw)

This paper presents a literature review conducted under an ESA funded study entitled "Applications of SAR Polarimetry". This study was jointly conducted by QinetiQ, AELc, University of Rennes 1, SarVision and Vexcel UK. The overall study aim was to review, assess and validate the benefits of using polarimetry for land cover classification and sea ice classification. The potential of polarimetric interferometry for vegetation parameter retrieval was also assessed and demonstrated. A literature review on classification techniques for polarised SAR data was undertaken. The objective of this review was to choose the most promising techniques prior to their evaluation for land and sea ice applications. This paper presents the objectives, the key issues, the conclusions and recommendations of the review. 1 STUDY BACKGROUND The objective of this study is to investigate techniques that will support applications of SAR polarimetry in ESA missions in the Earth Watch programme, and also missions in which ESA has a strategic interest such as Terrasar, Radarsat 2 and Alos Palsar. In particular, to investigate techniques for the following application areas: agriculture and land use classification, sea ice monitoring. Additionally, the applications of the new technique that combines SAR polarimetry with SAR interferometry (POLinSAR) are also investigated. The selected application areas are important. The mapping of agricultural crops and land cover has been identified by both commercial service providers and by the European Commission as a key product area. Sea-ice classification, for navigation, is already effectively an operational application with current sensors, i.e. single frequency and single polarisation SARs. However the use of polarimetric data is anticipated to resolve ambiguities that can arise with single channel data. Polarimetric data can also provide information on physical characteristics such as soil moisture and forest bio mass. When polarimetry is combined with interferometry further information becomes available such as tree height. In order to study the wide rage of polarimetric topics involved the following consortium was formed: QinetiQ Ltd (Prime Contractor), AELc, SarVision BV, University of Rennes and Vexcel UK Ltd. This paper describes the results of a literature survey on techniques for the classification of polarimetric data. This survey provides a background to specific areas that are presented separately at this Workshop [1, 2, 3], it also provides a summary of the basic issues and problems that arise in the classification of polarimetric data which are relevant to the round table discussions. Polarimetric interferometry is a largely self contained new technique and is described in [4] 3 REVIEW 3.1 Key issues in the classification of SAR data There are technical problems and conceptual problems associated with image classification. The technical issues have undergone progressive developments over a number of years. The conceptual issues have always been present, but have generally been masked behind classification errors and the effects of data resolution. Both of these issues are introduced below.