Abstract
In order to monitor snow-cover dynamics in the Tana River Basin in Northern Fennoscandia, SPOT VEGETATION (VGT) images of the snowmelt seasons of 1998 and 1999 were used to identify snow-covered areas, employing an algorithm that was originally developed for data from the Moderate Resolution Imaging Spectroradiometer (MODIS). This algorithm is based on the Normalized Difference Snow Index (NDSI), which usually is calculated from the green and mid-infrared bands. In the absence of a green band, the applicability of this algorithm to VGT data from the red and mid-infrared bands was tested by comparing NDSI values with a corresponding Landsat Thematic Mapper (TM) image. The best agreement was found with slightly lower threshold values for the NDSI. Comparison of the snow-cover estimates also allowed testing of the performance of the NDSI-based algorithm in partially snow-free conditions. By applying the algorithm to ten-day syntheses of VGT images, the moment of snow disappearance could be registered for each 1 × 1 km pixel in the study area. The results were largely consistent with observations at meteorological stations in the area, confirming the effectiveness of VGT images and the algorithm employed in monitoring snow-cover depletion patterns.
Original language | English |
---|---|
Pages (from-to) | 2933-2949 |
Number of pages | 17 |
Journal | International Journal of Remote Sensing |
Volume | 25 |
Issue number | 15 |
DOIs | |
Publication status | Published - 3 Jun 2004 |
Externally published | Yes |