Abstract

Release of snow avalanche from a mountain slope depends on various parameters such as snow cover, terrain and meteorological conditions of the region. The precise information of avalanche occurrence in terms of its location and extent is essentially important for hazard mapping and for avalanche occurrence feedback. In the present study, various techniques have been explored for automatic detection and mapping of snow avalanche debris for a part of Western Himalayan region using Sentinel-2 satellite data. Spectral signatures of avalanche and non-avalanche snow collected from the field spectroradiometer survey are used for identifying suitable spectral bands of Sentinel-2 for avalanche debris detection. Techniques such as Ratio Method, Gray Level Co-occurrence Matrix, a new proposed index, i.e. Avalanche Debris Index and Object-Based Image Analysis (OBIA) are applied on satellite images to retrieve the avalanche debris. Retrieved avalanche debris are further compared with the manually digitized avalanche occurred boundaries. The OBIA method has been found the most suitable for avalanche debris detection and mapping using the medium resolution satellite data.

摘要

山坡雪崩的释放取决于各种参数，例如该地区的积雪、地形和气象条件。雪崩发生的位置和范围的精确信息对于灾害测绘和雪崩发生反馈至关重要。在本研究中，已经探索了各种技术，用于使用 Sentinel-2 卫星数据对喜马拉雅西部部分地区的雪崩碎片进行自动检测和绘图。从现场光谱辐射计调查中收集的雪崩和非雪崩雪的光谱特征用于识别适合雪崩碎片检测的 Sentinel-2 光谱带。比值法、灰度共生矩阵等技术，一种新提出的指标，即 雪崩碎片指数和基于对象的图像分析 (OBIA) 应用于卫星图像以检索雪崩碎片。检索到的雪崩碎片进一步与手动数字化的雪崩发生边界进行比较。已发现 OBIA 方法最适合使用中分辨率卫星数据进行雪崩碎片探测和绘图。