EXTRACTING RELEVANCE FROM SAR TEMPORAL PROFILES ON A GLACIER AND AN ALPINE WATERSHED BY A DEEP AUTOENCODER

Avalanches are vital for glacier survival, yet their impact is difficult to quantify. We used low-cost cameras and drones to monitor an avalanche cone in the French Alps for two years. By accounting for ice flow, we found that avalanches can deposit 30 meters of snow annually – 50 times more than normal snowfall. This high-frequency data reveals that these cones fill until reaching a specific steepness, after which new snow slides further down to the base.

Sub-annual glacier surface velocity derived from remote sensing images are receiving increasing attention. In the meantime, systematic seasonal errors (i.e. biases) have been reported in time series derived from optical images on various landforms, such as landslides or dunes. The extent to which such biases affect glacier velocity maps remains poorly investigated. Here, we propose characterizing the amplitude and spatial distribution of these seasonal biases.

Slow-moving landslides respond to seasonal climate variations, but displacement time series from optical satellite imagery often contain illumination-related biases that obscure true signals. This study assesses methods to mitigate seasonal errors and reveals kinematic changes and controlling factors of a large slow-moving landslide in the Argentinean Andes.

While global annual glacier velocities are openly accessible, sub-annual velocity time series are still lacking. This hinders our ability to understand flow processes and the integration of these observations in numerical models. We introduce an open source Python package called TICOI (Temporal Inversion using linear Combinations of Observations, and Interpolation) to fuse multi-temporal and multi-sensor image-pair velocities produced by different processing chains to produce standardized sub-annual velocity products.

Avalanches are important for the mass balance of mountain glaciers, but few data exist on where and when they occur and which glaciers they affect the most. We developed an approach to map avalanches over large glaciated areas and long periods of time using satellite radar data. The application of this method to various regions in the Alps and High Mountain Asia reveals the variability of avalanches on these glaciers and provides key data to better represent these processes in glacier models.