Forest Dynamics

image
Image of a secondary tropical forest in Sabah, Malaysia, © Nikolai Knapp.

Forests cover roughly 30% of the earth’s land surface, but land-use and climatic change are modifying and threatening these ecosystems. To protect forests a better understanding of forest dynamics is crucial.

Therefore, we have developed a family of process-based forest models: FORMIND and FORMIX. The models are used to simulate the spatio-temporal dynamics of mixed-species forests in combination with disturbances (e.g. logging, drought, storms). Both models calculate tree growth, tree mortality, the carbon balance of individual trees, competition for light, water and space, and recruitment processes by explicit seed dispersal.

FORMIND Logo
Our forest models have been applied in Malaysia, French Guyana, Ecuador, Paraguay, Venezuela, Brasilia, Mexico, Chile, Panama, Sri Lanka, Madagascar, Tansania, USA and Germany. Large scale versions for forests of the Amazonian Region have been developed (simulating every tree in the Amazon). We apply complex spatial statistics on observed and simulated tree data in order to get new insights into forest structures and dynamics. We couple the forest models also with different types of remote sensing data (Lidar e.g GEDI, optical remote sensing e.g. Sentinel 2, radar e.g. Tandem-X).

In an ERC project we are analysing the coexistence of tree species in global forests. Therefore we are working in close cooperation with the forest megaplot network FORESTGEO (from the Smithsonian Institute). Together with TU Munic and DLR Oberpfaffenhofen we have also an own forest megaplot (Traunstein, 25 ha).
In an large AI project we use satellite data to estimate carbon fluxes and pools of global forests (3D-ABC Hemholtz Foundation Model Initiative).

A recently developed forest packing model FPM follows a novel static approach focussing on the structure of forests. Thereby, tree crowns are randomly packed within a forest only competing for space. The simple model has so far been applied to forests in Panama and Sri Lanka. The forest packing model predicts the tree size distribution of forests as well as the spatial heterogenity of local forest structure.

Virtual Forest
Visualization of a crown packing for a tropical forest in Panama. © Thorsten Wiegand /UFZ

If you are interested in using FORMIND, FORMIX or GRASSMIND, please contact us or visit www.formind.org for download requests.

Selected Publications

Contact

Links