Object tracking through time: the example of a rock avalanche on a glacier

International audience "Just as the availability of satellite remote sensing data increases, so do the demands for extracting geospatial information in an automated manner at different scales. Within the research project MORPH (http://morph.zgis.at/), new object-based methods for mapping, monit...

Full description

Bibliographic Details
Main Authors: Tiede, Dirk, Friedl, Barbara, Hölbling, Daniel, Dittrich, Jirathana, Aubrey Robson, Benjamin
Other Authors: Department of geoinformatics Z_GIS, University of Salzburg, University of Salzburg, Department of geography, University of Bergen (UiB), Centre d'Etudes Spatiales de la BIOsphère (CESBIO), Office national d'études et de recherches aérospatiales (ONERA), Espace pour le développement (ESPACE DEV), Société T.E.T.I.S
Format: Conference Object
Language:English
Published: HAL CCSD 2018
Subjects:
Online Access:https://hal.univ-reunion.fr/hal-01958407
Description
Summary:International audience "Just as the availability of satellite remote sensing data increases, so do the demands for extracting geospatial information in an automated manner at different scales. Within the research project MORPH (http://morph.zgis.at/), new object-based methods for mapping, monitoring and modelling spatial-temporal dynamics of surface morphology are developed by an integrated analysis of various remote sensing data (optical, SAR, DEMs). The focus is on the multi-scale investigation of landslides and volcanic deposits in study areas in southern Iceland, which are highly dynamic in their geomorphic evolution and characterized by progressive mass displacements and surface morphology changes. The integrated use of remote sensing data from different sensors allows for the mapping of geomorphological features and the monitoring of surface morphology changes at high spatial and temporal resolutions. Within this study, we investigate a specific methodological topic of object-based image analysis (OBIA), namely the tracking of (in this case slow) moving objects through time based on image data from different optical satellite sensors. We are focusing on a large rock avalanche, which occurred in 1999 in the southeastern part of the Vatnajökull ice cap. Debris from this event was deposited on the Svöludalsjökull outlet glacier. The extent of the debris cover on the glacier is automatically delineated using object-based rule-sets on different Landsat images of the last 18 years. Building on a strategy to link moving objects geographically between different images based on directional overlap and a space-temporal model of the candidate object class, we estimate the speed of the debris and the change in size of the debris deposit. Results are compared to a glacier velocity estimation using SAR offset tracking with TerraSAR-X data. Such an automated change analysis of rock avalanche deposit transportation on a glacier will not only help in documenting the rock avalanche-fate itself, but also provides an ...