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 spatial data analysis and processing opens up innovative ways to accurately track and monitor forests and trees. Using advanced machine learning techniques, neural networks and automated processes, we offer customized solutions ranging from single tree detection to detailed monitoring of forest damage and reforestation areas.

Single Tree Detection

Updating and expanding tree inventories is a crucial step in modern green space management and forestry. With our machine learning methods we not only identify the location of a tree but also its crown diameter, height, and other parameters. By combining different data sources (orthophotos, point clouds, surface models), we can capture significantly more trees than traditional registers can represent. We achieve particularly outstanding results with isolated trees.

Forest Monitoring

To achieve certification for sustainable forestry and to comply with the EU Deforestation Regulation (EUDR) for deforestation-free supply chains, we evaluate remote sensing options for large-scale monitoring of forest areas. These innovative approaches aim to complement traditional on-site inspections and, in the long term, enable automated processes to prevent deforestation in global supply chains. In an ongoing project with the Irish Forest Service, we are automatically detecting clear-cutting across Ireland’s forested areas. We are continually improving our methods, including the use of Sentinel-1 data, as cloud cover currently limits analysis to an annual rather than quarterly basis.

Early Warning Systems for Forest Damage

We are exploring the innovative potential of Sun Induced Fluorescence (SIF). Plants undergoing photosynthesis emit excess energy in the form of fluorescence, which novel satellite missions like FLEX or EnMAP can detect. This capability can be used to identify vegetation damage much earlier than with conventional spectral indices. A potential AI-based method could enable the early detection of declining plant health.

Reforestation Potential

The Hermosa and ESMERALDA web portals, co-developed by mundialis, allow areas with reforestation and restoration potential to be identified and monitored globally. The portals also provide a platform for linking local projects, global initiatives and funding partners. Together with international partners, we are laying the foundations for sustainable reforestation projects around the world.