Author(s): Hussainzada Wahidullah; Han Soo Lee; Vinayak Bhanage
Linked Author(s):
Keywords: Snow cover extent; MODIS; Water resource management; Confusion matrix; Sustainability
Abstract: Snow cover and mountain glacier are the main source of water in Afghanistan. In this study, the MODIS NDSI snow cover C6 were used for the snow cover mapping over Balkhab River basin in the Northern river basin, Afghanistan. Average cloud coverages in the MODIS Aqua and Terra images were high with 37.6% and 31.8%, respectively, from 1st January 2010 to 31st December 2018, and then were improved by spatial and temporal combinations; Aqua and Terra combination (A/T), Temporal combination of (A/T) products with preceding (A/T +1day) and previous days (A/T -1day). The 3 steps combination reduced the average cloud coverage down to 8.9%. Validation and confirming the accuracy of remote sensing data is important. Since historical observed data is not available for the study area, the Landsat 8 images with less than 20% cloud cover for the months with higher snow cover were chosen for the validation. In addition to the direct pixel-by- pixel comparisons, confusion matrix is utilized for validation. A confusion matrix is a summary of prediction results on a classification problem to describe the performance of a classification model. In direct comparison, 86.47% of the pixels are matching. In the confusion matrix, the overall accuracy (ACC) is increased to 89.37% because the cloud pixels are also land or snow in real. The MODIS final product of this study can be used for a ssessment of the contribution of snow-melt water to the river stream, and the effect of climate change on the amount of snow accumulation during study period.
Year: 2020