SNOW AVALANCHE HAZARD MAPPING USING HIGH RESOLUTION REMOTE SENSING IMAGES AND DATA & PROCESS DRIVEN MODELLING

Abstract

Accurate observations of snow and avalanche parameters are important inputs in the process of avalanche hazard estimation, forecasting, and mitigation. In the present research high resolution estimation of the two important parameters snow depth and avalanche deposits, related to avalanche hazard assessment and validation were considered. The snow depth of an avalanche-prone region is directly linked with avalanche occurrence likelihood and the detection of avalanche deposits is used for the avalanche hazard model validation and eminent hazard assessment of the avalanche-prone regions. This study presents the high-resolution snow depth and avalanche deposits mapping of avalanche-prone regions of the Manali-Dhundi area, Himachal Pradesh (HP), India, using repeated UAV Photogrammetric surveys. The images captured by an RGB camera mounted in a fixed-wing UAV (e-Bee X) were used to generate the high-resolution Digital Surface Model (DSM) and Orthophotos of the study area. A snow depth map was generated by differencing co-registered snow-covered DSM and bare surface DSM. UAV-derived snow depth values have been validated with the point snow depth measurements of Wireless Sensor Networks (WSN) and manual observations. The UAV-retrieved snow depth values were found in good correlation with field-measured snow depth (R=0.95). A novel framework using UAV RGB images is proposed for avalanche deposits detection and analysis. In the first part of the proposed framework, an OBIA-CNN (Object-Based Image Analysis-Convolution Neural Network) method is applied to UAV images for the detection of avalanche deposits. In the second step the detailed accuracy analysis of the detected deposits based on their count, area, size similarity, and shape similarity with respect to reference deposits has been carried out. The precision of the model was found 1.0, the value of recall was 0.88 and the F-1 score was 0.93. The overall object shape-matching accuracy of the detected deposits was also found high (0.92). Finally, the surface area and snow volume-based characterization of the detected deposits was performed. The regular monitoring (before and after every snow storm) of the snow depth and avalanche deposits using the UAV-based method is quick, cost-effective, and provides accurate snow and avalanche conditions for operational use and planning in snow-bound regions.