snow
65th EASTERN SNOW CONFERENCE
Fairlee (Lake Morey), Vermont, USA 2008
Evaluation of Pixelbase and Subpixel Methods for Snow Cover
Studying in Regional Scale
SEPIDEH DADASHI, 1 ALI-AKBAR MATKAN, 2 PARVIZ ZIAIIAN, 3
AND DAVOOD ASHORLO 4
ABSTRACT
Monitoring snow cover changes are important to water resource researches because it supplies electricity and freshwater. Although operational snow cover mapping by optical sensors has been used for more than 4 decades in the world but In Iran, traditional data collection methods are still common. Using these methods have a lot of problems. Usually the ground observation network is not dense enough to provide required information - especially in rugged regions- and also mountainous situation makes difficult accessible to ground stations.
In this research, Modis satellite images after changing their PDS format to L1B, used for snow cover studying in Karaj and Latian Basins. These Basins that bounded between 51° 20' to 51° 36' longitude and 35° 52' to 36° 11' latitude supply consumption water for capital city of Iran.
Two methods used in this research. NDSI used as a Pixel-base Method that determines snow/no snow pixels and linear spectral unmixing algorithm used for Pixelbase method.
Results shows that subpixel method in comparison to NDSI – NDSI.0.4 that is proposed for global scale monitoring- has better result but using NDSI after atmospheric correction and changes its threshold, has best result. In order to atmospheric correction, Bulk correction method used.
For subpixel method, Endmembers determined using IRS- P6 image and MNF algorithm used for decrease noises. Thereafter, spectral curves determined for each endmembers and finally spectral unmixing performed. For evaluation the accuracy of both of the pixelbase and subpixel methods, IRS-P6 Images – 23.5 m spatial resolution - used.
Keywords: snow cover; pixelbase; subpixel; NDSI, linear spectral unmixing; MODIS; Karaj and
Latian Basins
Fairlee (Lake Morey), Vermont, USA 2008
Evaluation of Pixelbase and Subpixel Methods for Snow Cover
Studying in Regional Scale
SEPIDEH DADASHI, 1 ALI-AKBAR MATKAN, 2 PARVIZ ZIAIIAN, 3
AND DAVOOD ASHORLO 4
ABSTRACT
Monitoring snow cover changes are important to water resource researches because it supplies electricity and freshwater. Although operational snow cover mapping by optical sensors has been used for more than 4 decades in the world but In Iran, traditional data collection methods are still common. Using these methods have a lot of problems. Usually the ground observation network is not dense enough to provide required information - especially in rugged regions- and also mountainous situation makes difficult accessible to ground stations.
In this research, Modis satellite images after changing their PDS format to L1B, used for snow cover studying in Karaj and Latian Basins. These Basins that bounded between 51° 20' to 51° 36' longitude and 35° 52' to 36° 11' latitude supply consumption water for capital city of Iran.
Two methods used in this research. NDSI used as a Pixel-base Method that determines snow/no snow pixels and linear spectral unmixing algorithm used for Pixelbase method.
Results shows that subpixel method in comparison to NDSI – NDSI.0.4 that is proposed for global scale monitoring- has better result but using NDSI after atmospheric correction and changes its threshold, has best result. In order to atmospheric correction, Bulk correction method used.
For subpixel method, Endmembers determined using IRS- P6 image and MNF algorithm used for decrease noises. Thereafter, spectral curves determined for each endmembers and finally spectral unmixing performed. For evaluation the accuracy of both of the pixelbase and subpixel methods, IRS-P6 Images – 23.5 m spatial resolution - used.
Keywords: snow cover; pixelbase; subpixel; NDSI, linear spectral unmixing; MODIS; Karaj and
Latian Basins