The potential of earth observation for the analyses of land surface dynamics

Die Vortragsreihe GeographieDiskurse befasst sich im Sommersemester mit dem Thema "γεωγραφία (geographia): Earth Description – Earth Observation". Prof. Claudia Künzer spricht am 9.3.2021 um 12:00 in ihrem Online-Vortrag über "The Potential of Earth Observation for the Analyses of Land Surface Dynamics".

At present an unrivalled amount of remotely sensed earth observation data is globally available in satellite data archives. Imagery from medium-, high- and highest resolution optical, multispectral, and radar sensors is waiting to be exploited to reveal its full potential. The USA and Europe have set the pace in sharing earth observation resources; amongst others via the free and open Landsat data archives spanning nearly four decades, as well as data of the European Sentinel missions in orbit since 2014. Along with this go increasing capabilities to handle and explore big data of dense temporal coverage. Comprehensive geographic analyses combining the findings of earth observation data analyses with ancillary data and in-situ ground knowledge propel earth observation beyond simple mapping activities. Whereas in past decades remote sensing data were analysed based on the spectral-reflectance-defined ‘finger print’ of individual surfaces, it is now the ‘temporal fingerprints’ of our land surface which open up novel pathways of data analyses. Nowadays – more than ever – remote sensing as a methodological discipline can contribute to addressing society-relevant challenges. Especially in highly dynamic regions earth observation can support the sustainable management of natural resources via long-term monitoring schemes and the provision of sophisticated, planning-relevant information. The talk will present related examples from recent analyses of land surface dynamics globally, and will also shed light on the potential of future prognosis based on long term trends, as well as a completely new analyses sphere enabled via deep learning and object extraction methods in very highest resolution data archives.

Nach oben scrollen