Forest and Tree Mapping at mundialis

Modern technologies for sustainable forestry

Challenge

The development of effective monitoring systems is essential to ensure sustainable forest management. These systems must work consistent and reliable over long periods of time. One challenge is the early detection of pests, diseases and other stress factors, as many damages often only become visible late and require special technologies for their detection. In addition, the precise measurement of tree height, crown diameter and location requires a high degree of accuracy. Inaccuracies in this data can significantly affect the assessment of tree health and available forest resources.
mundialis is addressing these challenges and develops innovative solutions that enable comprehensive and accurate monitoring of our forests.

Services
  • Accurate single tree detection: Updating and expanding tree inventories through using advanced machine learning techniques.
  • Comprehensive forest monitoring: Employing remote sensing techniques to monitor forest areas and detect changes in vegetation.
  • Early warning systems for plant health: Use of innovative technologies such as Sun Induced Fluorescence (SIF) for early detection of plant stress and damage.
  • Supporting sustainable reforestation: Identifying and monitoring potential areas for reforestation projects through web portals and stakeholder networking.
Customer
Our methods for forest and tree mapping have already been used in various customer projects.

Story

Our expertise in analysing and processing spatial data opens up innovative possibilities for accurately recording and monitoring forests and trees. By using modern machine learning techniques, neural networks and automated processes, we offer customised solutions ranging from individual tree detection to the close monitoring of forest damage and afforestation areas.

Single tree detection

Updating and expanding tree cadastres is a crucial step for modern green space management and forestry. With our machine learning methods, we not only identify the position of a tree, but also its crown diameter, tree height and other parameters. By combining a wide range of data sources (orthophotos, point clouds, surface models), we are able to capture significantly more trees than conventional cadastres. We achieve excellent results, particularly with free-standing trees.

Forest monitoring

For certification of sustainable forestry and to comply with the EU regulation on deforestation-free supply chains (EUDR), we are evaluating the possibilities of remote sensing for large-scale monitoring of forest areas. These innovative approaches are intended to supplement the previous on-site inspections and enable automated processes in the long term to prevent deforestation in global supply chains.
In an ongoing project in collaboration with the Irish Forestry Service, we are automatically detecting clear-cutting in forest areas throughout Ireland. Due to the challenges posed by cloud cover, which currently only allows for annual analyses instead of quarterly ones, we are continuously working to improve our methods, for example by using Sentinel-1 data.

Frühwarnsysteme für Waldschäden

We are working on the innovative potential of sun-induced fluorescence (SIF). Plants that carry out photosynthesis release excess energy in the form of fluorescence, which can be detected by new satellite missions such as FLEX or EnMAP. This can be used to identify damage to vegetation much earlier than with conventional spectral indices. An AI-based method that enables the early detection of declining plant health would also be conceivable.

Reforestation potential 

On the web portals HERMOSA and ESMERALDA, which mundialis helped to develop, potential areas for reforestation and renaturation can be identified and monitored globally. In addition, these portals offer the opportunity to link local projects, global initiatives and donors. In cooperation with international partners, we thus create the basis for sustainable reforestation projects worldwide.