Remote sensing data is one of the largest sources of big data in the world, with an estimated 150 terabytes of new data generated every day, enabling seamless monitoring of ecosystems. In our research, we develop new remote sensing approaches to detect and understand changes in ecosystems in the wake of global climate and land use change. We focus intensively on forest ecosystems, with current emphasis on the quantification of disturbances (e.g. windthrow, bark beetle or fire) in Europe and the assessment of the recovery function after disturbances. Methodologically, our research is based on the evaluation of dense time series of optical satellite sensors, such as Landsat or Sentinel-2. To process the large number of individual images, we develop and use state-of-the-art approaches (so-called data cubes) on super-computers of the Leibniz Supercomputing Centre. Our remote sensing-based products are actively used to monitor changes in Europe's forests (LINK) and are used in a variety of research projects.

We use active laser scanning methods to create three-dimensional models of ecosystems. Terrestrial laser scanning enables the creation of highly detailed 3D models for local areas (e.g. a forest stand), while airborne laser scanning provides aerial data for landscape-level analyses. At our professorship we have access to a modern terrestrial laser scanner as well as a variety of existing airborne data sets. In our research, laser scanning plays an important role in the quantification of ecosystem structure (e.g. tree heights or gaps in the canopy), in the development of local geophysical models (e.g. for modeling local thermal characteristics), and in the large-scale analysis of biodiversity. As part of the global StrucNet network, we operate two long-term monitoring plots with permanent laser scanners, which provide continuous monitoring of canopy structural features in deciduous and coniferous forest systems.

One of the strengths of our research is the integration of remote sensing data with other spatial data (e.g. climatological reanalysis data) and existing models (e.g. Global Dynamic Vegetation Models). To this end, we cooperate with leading groups worldwide, but also develop our own approaches and methods. We use modern statistical methods (e.g. multi-level models, extreme value statistics) to quantify, explain and model changes in ecosystems. A particular focus is on gradual and abrupt changes in tree mortality and how these are related to climatic extremes (e.g. droughts and heat waves).

In our group, we have access to a High-Performance Cluster (HPC) for the efficient processing of large Earth observation datasets. The cluster is operated by the Leibniz Computing Center of the Bavarian Academy of Sciences, and our group has exclusive access to two nodes, each with 48 CPU cores and 786 GB of RAM. To efficiently handle the large volumes of data (several hundred terabytes, 1 TB = 1000 GB), all data is stored on a specialized storage solution (Data Science Storage). For students, we can offer temporary access to the cluster via virtual machines specifically set up for students as part of their thesis projects. This allows students to gain experience in handling large to very large Earth observation datasets during their studies.

In addition to an efficient computing infrastructure, our group maintains a range of devices for close-range and in-situ recordings:

• Riegl VZ400i terrestrial laser scanner
• APOGEE PS100 RGB/NIR spectroradiometer
• Emlid Reach RS3 RTK GNSS receiver (rover & base)
• LEAF MK3 permanent terrestrial laser scanner
• Various microclimate loggers and cameras

Students have the opportunity to learn how to use these devices in seminars and can borrow the equipment for their thesis projects.