Today some of our EAGLE students presented their internship and innovation laboratory projects. Very interesting topics and they obviously applied and deepened their remote sensing knowledge a lot.
Julia Sauerbrey: Prediction of Organic Matter Content from Sentinel-2 Imagery in Germany
Marina Reiter: Conservation in Tasmania: Green Leaf Project and World Class Reserve Systems
Sarah Nolting: Regional Scale Statistical Mapping of Snow Avalanche Likelihood and its Combination with an Optical Remote Sensing Based Avalanche Detection Approach – First Attempts for the Province of South Tyrol (Italy)
The following internship and innovation laboratory projects were presented today:
Karsten Wiertz did his internship at the Białowieza national park on “Spatio-temporal analysis of tree mortality and gaps in the Białowieza Forest using high resolution imagery”.
Jakob Schwalb-Willmann did his internship at the Max-Planck-Institute in Radolfzell on “moveVis GUI: Development of a browser-based web application for animating movement data”.
Marcus Groll worked during his internship at DLR DFD on “Configuration of augmented & virtual Reality content with Unity 3d”.
Pilar Endara analyzed in her internship at DLR DFD the “Spatio-temporal analysis of precipitation and NDVI in Somalia”.
Additionally we heard about the innovation laboratory of Jakob Schwalb-Willmann at EURAC „Sentinel-2 based forest change detection across the Alps“.
Analyzing species-environment interaction is feasible using various data and method. An increasing technology is the tracking of animals and especially its linkage to remote sensing, as covered in AniMove.org. However, with this technology new challenges have to be dealt with, e.g. the decrease in accuracy of tracking devices in dense vegetation. In this innovation laboratory you can explore the error margins of tracking devices in relation to the landcover and explore options to solve it and being able to provide new improved methods to link animal movement data and remotely sensed environmental information.
This study will examine whether the application of hybrid classifiers increases the classification accuracy in comparison to a single classifier. A combination between parametric and non-parametric classifiers will be applied and their performance will be assessed. The student is expected to gain a deep knowledge of applied Machine Learning algorithms within this innovation laboratory.
The data provided by aerial imagery are amongst the oldest sources of spatially explicit information for modern-time environmental management. These data are often captured over landscape-level domains using overlapping flight stripes to enable stereo photogrammetric analysis based on parallax and obtaining information on vertical forest structure. As compared to the state-of-the-art LiDAR data, the stereo photogrammetric analysis is associated with a number of advantages and disadvantages for creating surface models, including the absence of bare earth ground data and usual geometric distortions (disadvantage) as well as reasonable acquisition costs and provision of multispectral information (advantages). The idea behind this work within an innovation laboratory is to analyze a series of multitemporal stereo orthophotos from the Bavarian Forest National Park (BFNP) to create surface models. The final products can be subtracted from the available LiDAR-based DTMs to provide canopy height models (CHM) representing the actual tree heights. The final aim of this analysis is to enable a further generation of multitemporal forest structure data that will be sued for a change detection of forest structural attributes as affected by natural disturbances.
Actual evapotranspiration (ETact) is an essential component of the water balance and its determination for large areas is difficult on regional scale and can be explored within an innovation laboratory. The use of remote sensing data to determine ETact is particularly suitable to provide area based indicators for the evaluation of the efficiency and productivity of irrigation systems. Seasonal analysis of ETact is hampered by either spatial (MODIS) or temporal (Landsat) resolution. In order to provide a high-resolution (temporal and spatial) and dense remote sensing dataset Landsat and MODIS data will be fussed using the ESTARFM algorithm.