Abstract
This work presents a multi-temporal and multi-source approach for glacier cover classification, i.e. bare soil, glacier ice, firn, and snow. The method is based on Hidden Markov Model (HMM) and Support Vector Machine (SVM) and can handle different kinds of satellite virtual constellations composed of high-resolution optical and/or SAR platforms. The proposed method is tested on a Sentinel-1 time series acquired over the Ortler Alps and the obtained classification map time series is used to extract the temporal behavior of the snowline and estimate the equilibrium line altitude (ELA) of the glaciers.