A tree is a data point

Trees as Infrastructure Assets

Trees are more than background greenery; they are essential infrastructure, much like fire hydrants, streetlights, or drains. However, unlike built infrastructure, which typically depreciates over time, trees, as living infrastructure, grow in value. Whether economic (e.g. property value), environmental (e.g. CO₂ capture), or health (e.g. psychological well-being)—their contribution to society grows exponentially over time (Rosen and Lane, 2019).

Amid intensifying climate challenges, this realisation of value has sparked a need to integrate, monitor, and maintain green infrastructure, with trees at the heart of such programs. From transnational policies like Europe’s Nature Restoration Law, to city-focused initiatives such as the Green Riyadh Project, trees are actively being intertwined into the urban landscape. Beyond government-led efforts, other non-profit and public–private collaborations—like the Amigos De Los Ríos – Emerald Necklace Project in California and the Amsterdam Birthday Forest—are reshaping how communities engage with urban nature.

And this tree revolution is not just limited to cities.

Beyond Cities: Large-Scale Tree Planting

Around the world, transformative tree-focused initiatives are reshaping our future. The Saudi Green Initiative aims to plant 10 billion trees and restore 40 million hectares of degraded land (a land mass 10x the size of the Netherlands). The program has already planted 43.9 million trees, and more than $180 billion has been earmarked for the project.

Across Africa, over 20 countries are committed to the Great Green Wall—a continent-wide endeavour to plant 8 billion trees across the Sahel region to halt the advance of the Sahara Desert. If successful, the project will restore 100 million hectares of degraded land and transform the continent.

A Tree is a Data Point

Driving these initiatives is technological advancement. Trees—whether lining a small street footpath in Jakarta or spanning the vast Amazon—are being watched. With the rising need for climate transparency, tree data has become as critical for success and reporting as engagement metrics are for a social media campaign.

From governments and NGOs to landowners and investors, stakeholders now want to know: How many trees are there? What species are they? Where are they located? Are they healthy?

And why? Because trees are no longer just part of the background—they're increasingly recognized as valuable assets for climate mitigation, biodiversity, and community wellbeing. Whether it’s reporting on emissions reductions, verifying nature-based solutions, or planning urban infrastructure, reliable tree data underpins credibility, funding, and impact. From carbon markets to ESG reporting, stakeholders can’t afford to act without it.

In this context, trees have evolved into measurable assets. They are tracked, analyzed, and sometimes even represented as digital twins. Their condition and health are monitored and often modelled to derive additional insights.

This is where geospatial technologies come in.

With geo-information tools, a single tree can be mapped as a point enriched with attributes such as height and canopy width. Nonetheless, recent advances in high-resolution satellite imagery and LiDAR datasets allow for even richer and more robust information to be gathered about individual trees and forest canopies, all at scale.

Tree monitoring experts

At Space4Good, we specialize in delivering value by watching trees from above. Since 2020, we’ve been working at the intersection of remote sensing, regenerative agriculture, forestry, and biodiversity—bringing space technology to nature-based solutions.

Our team of analysts combine geospatial science with ecological knowledge to turn petabytes of open-source and proprietary satellite data into actionable insights. Outputs include carbon sequestration estimates, biomass growth over time, and both visual and quantitative reports that summarize key tree asset statistics. We help partners like those below to develop and understand the full story of their tree data.

Monitor afforestation project success
In afforestation projects, besides basic tree and plot data (like “planting date”, “tree ID” or “tree location”), decision-makers want to know:

a) tree growth metrics such as “tree height” and “crown diameter” and

b) tree health metrics such as “leaf color” or signs of disease/ pests.

These analytics can be derived by remote sensing technologies. The advantage of Earth observation solutions is that you can monitor at scale, salvaging the costs of the expensive and time-consuming physical audits.

Space4Good developed a satellite-based methodology to assess afforestation and monitor tree health and mortality. This approach was co-designed with project developer Hommes et Terre in Burkina Faso and integrates validation protocols in line with the available in-situ data to evaluate tree growth over time intervals. Using sample plots and optical satellite data as the primary input, we employed a pixel-based approach to identify where afforestation efforts were proving most successful. To evaluate tree growth rates, we analysed vegetation index trendlines (flat=no growth, increasing=vegetation growth, decreasing=vegetation decrease). Further, we examined the distribution of vegetation index values across all pixels within a plot for each satellite observation. We found that the NDVI trendline, combined with the 75th percentile of pixel values—which highlights the healthier, better-performing parts of each plot—offered the most reliable representation of growth performance.

From there, pixel-level vegetation performance was categorized into three classes: poor, moderate, and good. Based on the proportion of pixels in each category across a site, an overall site classification was assigned, providing a clear, aggregated view of afforestation success at the plot level.
Figure 1. Selected Hommes et Terre patrimonium of sites classified into afforestation success (poor, moderate, good)
This technology saves time, reduces costs, and improves accuracy. Building on lessons learned, Space4Good partnered with Ecosia for a similar project and automated monitoring across 11,000+ afforestation sites globally - each ranging from 1–200 ha.), By replacing manual and resource-intensive tracking with our remote sensing solution, the partnership enabled consistent, transparent, and timely insights into global reforestation progress at an unprecedented scale.
Automate inventory via tree detection, identification and classification
Identification of a tree is essential, but sometimes difficult, particularly when project developers or forest managers haven’t planted the trees themselves i.e. natural-born trees don’t come with a planting date or a unique ID; sometimes past data of such plantings does not exist. Space4Good addresses this challenge.

Via a project named RetreevAIble Space4Good, LucidMinds and UNL developed a state-of-the-art urban tree monitoring and analytics platform for Amsterdam. The project employed a mix of high-quality resolution datasets to 1) identify trees at the individual tree level 2) classify its species and 3) assess each tree’s biomass.

The Actueel Hoogtebestand Nederland (AHN4) LiDAR dataset provided data in 3D point cloud format, which allowed us to locate tree tops, detect tree crowns and provide information about the tree’s structure (height, volume).

For species classification, we used high-resolution SuperView satellite imagery to capture vegetation indices and seasonal changes. Spring, the peak of the growing season, provided distinct features that helped our model differentiate between species. Autumn, when leaves start falling, thereby losing chlorophyll content, marks a characteristic moment for each tree species. Here, spectral signatures analysis further enabled us to complete their identification. Amsterdam's comprehensive tree database served as the foundation for training and later for validation.

Next, analysing LiDAR data enabled us to estimate tree allometric equations, including the above-ground biomass and other important physical metrics.

The result- a dataset of over 8,000 individual trees, with key species information and other parameters such as their health, biomass and sequestered carbon. These insights allow cities to link tree monitoring with climate goals, carbon accounting, and environmental reporting, going far beyond basic inventories to truly understand the value of this living infrastructure.

RetreevAIble offers a blueprint for other municipalities in urban tree monitoring to combine structural (LiDAR) and spectral (satellite) data for a far richer, more accurate profile of each tree. Further, the municipal ground-truth data boosts the overall accuracy and makes the outputs more reliable for city planning, and tree maintenance.
Figure 2. a) Predicted tree species mapping b) individual tree metrics
This solution works outside of urban contexts as well. Space4Good, in collaboration with NatureMetrics and the ReForester Consortium, was selected as a semi-finalist in the XPRIZE Rainforest Challenge. This brought the consortium to Singapore, where we tested our model in one of the world’s most complex and biodiverse ecosystems: the tropical rainforest. By integrating drone-based LiDAR, multispectral, and thermal data, Space4Good developed an AI-powered pipeline capable of accurately identifying and classifying individual tree species. In this manner, a total of 579 different tree tops across 10ha were identified in a completely automated manner. From this, individual species were classified and verified with the help of local botanists. We delivered high-precision tree species distribution maps that offer deeper insights into the structure and composition of tropical forests.
Figure 3. a) Tree Segmentation Result from the multispectral sensor b) Tree species distribution map example
Estimate carbon and biomass uptake with >90% accuracy
Insights on biomass and sequestered carbon in agroforestry systems for carbon markets solicit more advanced metrics and modelling. Still, individual trees serve as the starting point for these calculations.

CarboCatch is a pilot initiative between Space4Good and the Louis Bolk Institute, designed to support carbon farming through improved Monitoring, Reporting, and Verification (MRV) for agroforestry farmers and project developers in the Netherlands. CarboCatch addresses the major barrier in the carbon credit market: the time and cost attributed to the manual audits. By leveraging AI and geospatial technologies, Space4Good’s team has streamlined and automated the MRV process, ensuring that trees are watched with minimal human intervention, reduced verification costs, and faster credit issuance, unlocking greater accessibility and scalability for nature-based carbon projects. However, each tree is still accounted for. CarboCatch is focused on accuracy, which reduces uncertainty and therefore increases the value and reliability of carbon certificates.

The CarboCatch platform models and visualizes biomass growth and carbon sequestration for agroforestry systems. Using ground truth data, earth-observation data and advanced machine learning techniques, we accurately estimate above-ground biomass and calculate both current and potential carbon uptake per plot.
Figure 4. CarboCatch platform with farm biomass layer
The project first focused on walnut trees grown in row-based agroforestry systems across three Dutch provinces: Brabant, Flevoland, and Gelderland. These sites were categorized into either clay or sandy soil plots.

In-situ data collection included measurements such as diameter-at-breast height (DBH), tree height, canopy diameter, and root diameter, following biomass estimation standards outlined by SNK. On the other hand, high-resolution multispectral satellite imagery (1–3 m; RGB, NIR, and Red Edge bands) and AHN LiDAR point cloud data are incorporated, particularly for extracting accurate canopy height information. Users can explore platform outputs, including biomass per tree or plot and carbon sequestration estimates, calculated using a standardized formula based on biomass carbon fraction and the carbon-to-CO₂ equivalent conversion factor.

By taking a tree-level approach, Space4Good identified that clay plots, which are generally more uniform in tree age and density, led to narrower biomass prediction ranges. This enabled improved model performance as the reduced variability simplified model predictions. In contrast, sandy soil plots with varied tree conditions introduced more complexity, highlighting the challenges of modelling diverse agroforestry settings.

CarboCatch, therefore, represents a step forward in the carbon credit market by enabling tree-level MRV that is both scientifically robust and operationally scalable. Unlike traditional systems that rely heavily on costly, time-consuming field audits, CarboCatch automates the process using high-resolution remote sensing, LiDAR, and machine learning—without compromising accuracy. Its ability to model individual trees, account for site-specific differences like soil type, and adapt to complex agroforestry conditions brings unprecedented precision and replicability to carbon accounting. By delivering clear, accessible data outputs, the platform improves transparency and trust, ultimately increasing the value and credibility of carbon credits and empowering farmers and project developers to scale nature-based solutions with confidence.
Large-scale green infrastructure monitoring
Space4Good is also exploring tree-derived insights for large-scale green infrastructure projects. We have an ongoing collaboration with NEDAMCO Africa, providing EO services, including stratification and biomass modelling to support carbon credit estimations for the Great Green Wall afforestation projects in Ethiopia. The results of this work helped Nedamco conceptualise the carbon credit potential within the entirety of Ethiopia.

Finally, our involvement in the INNO4CIFs project tackles this from a regional perspective. INNO4CFI promotes carbon farming initiatives in more than 13 European regions, and Space4Good is providing the satellite-based solution for carbon assessment and monitoring of such initiatives.

Conclusion

Do you want to transform your trees or living infrastructure into data points that can be measured and monitored?

Our Space4Good team can help you harness geospatial intelligence to turn your trees into measurable, monitorable data points. Whether you're looking to enhance your sustainability reporting, meet regulatory requirements, or improve tree/forest management, we’re here to support your journey from space to canopy. Visit our website or contact us via hello@space4good.com.