Identification of Snow Using Sar Polarimetry Techniques

The full polarimetric L-band ALOS-PALSAR data of snow cover area in Himalayan region have been analyzed based on various component scattering mechanism models and all model results are compared. Various polarimetric descriptors (Viz. degree of polarization, co-polarization coherence, total power, phase difference, polarization fraction) are also applied for snow discrimination. Based on these techniques, a new method i.e. Radar Snow Index (RSI) has been developed for total snow discrimination from other targets. * Corresponding author 1. INDRODUCTION 1.1.1.1 Mapping of snow and ice covered areas is important for many applications such as prediction of floods, snowmelt runoff modeling, water supply for irrigation and hydropower stations, weather forecasts and understanding climate change. Snow cover mapping through ground surveys are nearly impossible to carry out in the high relief Himalayan region, and aerial photographic surveys are insufficient. Visible and near-Infrared (NIR) remote sensing techniques are proved to be promising for snow cover mapping. But, in the presence of cloud cover and different weather conditions visible and NIR fail in acquiring snow cover information. Microwave remote sensing has an advantage over visible and NIR techniques due to its all weather capability, penetration through clouds and independence of sun illumination. Microwave observations of snow are sensitive to surface moisture variations, and thus, provide useful information concerning the variation in its physical state. Conventional SAR data at intermediate and low frequency (e.g. C -, L-band) is useful to detect only wet snow from other targets (Venkataraman et al., 2008). Due to high penetration capability and low attenuation of dry snow cover at intermediate and low frequency, dry snow cover behaves like transparent media. Hence discrimination of total (wet + dry) snow cover using intermediate and low frequency SAR data, still remains a problematic research approach (Venkataraman et al., 2008; Martini, 2005). The full polarimetric SAR data does contain more information than the corresponding single or dual polarization SAR data. Full polarimetric data gives an optimization of the polarimetric contrast and other polarimetric parameters which may be very useful for accurate target discrimination between snow and non-snow covered areas. Therefore, SAR polarimetry may produce great potential to map snow-covered areas and polarimetric SAR data analysis is also extremely useful to develop methodology to discriminate snow from other surface features. The main job proposed under this study, to investigate and develop innovative snow cover identification method by exploring and integrating various polarimetric SAR decomposition model parameters. Due to the limited availability of fully polarimetric PALSAR data, investigation is confined to Badrinath region in Himalayas. This is used as a test-site for an extensively snow-covered area having many glaciers of varying dimensions that act as a huge fresh water reservoir. The test-site covers Panpatia, Satopanth, Bhagirath Kharak, and Suraji Bank glaciers. Numerous smallsized glaciers also occur within the neighborhood . The area falls between latitude 30o 30’ N and 31 15’ N and longitude between 79 15’ E and 79 o 30’ E. In this study, the full polarimetric L-band ALOS PALSAR data (acquisition date May 12) are used for snow discrimination over Badrinath area, Uttarakhand in Indian Himalayan region. ALOS-AVNIR data (acquisition date, May 6, 2007) is used to interpret snow-covered area and non snow-covered area and it helps in the selection of the training sample of different features for supervised classification. 2. METHODS AND TECHNIQUES 2.1 Target Decomposition Models SAR polarimetry can be important to snow study and to discriminate snow and non snow cover. In this study, to find out L-band PALSAR capability for discriminating snow from other targets and to classify PALSAR data, existing target decomposition models have been applied. In literature, both coherent and incoherent target decomposition theorems are available viz coherent decomposition: Pauli and Krogager decomposition and incoherent decomposition: decomposition of Freeman and Durden (1998), Cloude and Pottier (1997), and Yamguchi (four component scattering model) (2007) etc.. Most of target in our test site is natural, thus incoherent decomposition may provide good results. In this investigation, incoherent target decomposition models have been applied. The incoherent decomposition theorems can be expressed as International Archives of the Photogrammetry, Remote Sensing and Spatial Information Science, Volume XXXVIII, Part 8, Kyoto Japan 2010