Forests play an essential role in the terrestrial carbon cycle and disturbances are an important factor in the dynamics of carbon cycling in ecosystems. Thus, improved monitoring of forest disturbances and recovery is required to understand disturbance regimes and their impacts in order to better quantify regional and global carbon dynamics. More accurate quantification of net CO2 exchange at global scales can improve our understanding of the feedbacks between the terrestrial biosphere and the atmosphere in the context of global change and facilitate climate policy-making.
Remote sensing (RS) is an essential component of current carbon observation assets, for detecting trends and variability in land components of the carbon cycle. RS data volumes and computation methods have recently led to rapid advances in our understanding of disturbances and RS is the only means for the mapping of ecosystem properties at larger scales capturing essential ecosystem processes. In addition, new RS products have become available (e.g. Hansen et al., 2013, FAO-FRA, 2010, De Sy et al., in progress) in order to better understand forest disturbances and recovery dynamics at global level. RS offers the possibility to assess the dynamics of terrestrial ecosystems at spatial scales suitable for capturing essential ecosystem processes, as well as change events such as natural and human induced disturbances. In this context, the thesis candidate will focus on the development of a satellite remote sensing based approach for creating maps containing a spatial classification of disturbances and regrowth frequencies in forest ecoystems. The FAO-FRA (FAO, 2010) dataset will be used to train and validate the developed approach.
- Explore the use of Random Forest (Breiman, 2001) to assess the suitablity of the high resolution global maps of the 21st century forest cover change (Hansen et al., 2013) as a proxy for the FAO-FRA dataset; and
- Explore other parameters such as disturbance indices, vegetation indices, biomass, Land Surface Temperature, and surface reflectance to improve the spatial classification approach.
- GRS – 32306 - Advanced Earth Observation
- Good knowledge in scripting is an asset (e.g. R)
Theme: Modelling & visualisation, Integrated Land Monitoring