Estimation of snow depth in GIS environment from observation points on Z Gali region: A case study of NW Himalaya

Abstract

The study's attempt is to investigate snow depth estimation in a GIS environment for observation points in the NW Himalaya. Using ASTER-GDEM with a 30 m resolution, snow density was mapped. The snow depth was collected on different days in the winter season, especially from January to March 2014. For this study, we used nine spatially distributed observation points. Out of nine spatially distributed observation points, eight locations were observed, and one out of each location was used for the cross-validation method. IDW and Kriging, with semivariogram models Gaussian and Linear, were attempting to derive the snow depth from routine manual observation points in the entire area. The study makes an effort to measure the 3D, 2D, and elevation distance of target points from other observation points, which improves the quality of the study. Three error estimation parameters were used: mean absolute error, mean percentage error, and root mean square error by cross-validation test. The result revealed that IDW values (32.51, −37.87, 33.94), Kriging Gaussian (52.54, −1.42, 51.45), and Kriging Linear (38.72, −38.66, 40.11). The minimum and maximum errors are then compared for five days at one observatory point in order to make a decision. According to the study, continually verifying the amount of snow in neighboring avalanche-prone areas offers useful information for assessing the integrity of the snowpack.

Keywords

• GIS
• Interpolation
• Observation points
• Snow depth
• Validation