Seeing nature from above: Using remote sensing data for nature-based decision making

(*This article is based on insights from our incredible NTC Now session featuring Luc Bettaieb and Greta C. of Versant ).

In today's fast-evolving landscape of conservation, technology is offering unprecedented ways to understand and protect nature. One of the most transformative tools in this effort is remote sensing—a method that allows us to monitor ecosystems and track environmental changes from above.

While fieldwork remains crucial, remote sensing technologies like satellites, drones, and advanced radar systems are revolutionizing how we gather data about the natural world. These tools capture insights at scales once thought impossible, from mapping deforestation across continents to detecting subtle shifts in biodiversity over time.

Organizations seeking to assess project sites, monitor supply chains, or validate nature-based investments now frequently rely on remote sensing to paint a comprehensive picture of their environmental impact. Whether it's identifying the most suitable land for restoration or ensuring that supply chain operations align with sustainability goals, the ability to analyze vast landscapes swiftly and accurately is key.

In a recent NTC Now session, the Nature Tech Collective was joined by Greta C. lead ecologist, and Luc Bettaieb , co-founder and CTO at Versant, to demystify the world of remote sensing. They explored how combining aerial imagery and sophisticated modeling techniques helps organizations make data-driven decisions for conservation and business strategies alike.

What are earth observation (EO) and remote sensing technologies?

Earth Observation (EO) and Remote Sensing Technologies refer to the tools and methods used to collect data about the Earth's surface, atmosphere, and ecosystems from a distance. These technologies enable us to monitor large-scale environmental changes, track human impact, and support data-driven decision-making for conservation, agriculture, business, and policy.

Often used interchangeably with “Earth Observation”, remote sensing gathers data from a distance using satellites, drones, or aircraft. Instead of manually tracking individual animals, researchers can analyze habitat changes, forest cover, and climate conditions over large areas.

Earth Observation (EO) technologies refer to the application of remote sensing technologies to study Earth's systems over time, including land, water, climate, and biodiversity. Typically, these technologies rely heavily on satellite data but tend to integrate aerial and ground-based sources too.

Types of remote sensing technologies and the nature insights they reveal

Remote sensing tools vary in their capabilities. By using them in combination, organizations can gain a comprehensive understanding of ecosystems, enabling more effective decision-making for conservation and restoration efforts.

Satellites

Satellites capture a variety of image data, often spanning different segments of the light spectrum. These images are categorized as multispectral or hyperspectral data, which means they detect and measure wavelengths of light that can give us valuable insights into vegetation health, water quality, and even changes in land cover.?

By analyzing how different surfaces on the Earth reflect light across these wavelengths, satellites can help monitor changes that aren’t visible to the human eye, such as early signs of drought, vegetation stress, or deforestation.

Drones & aerial imagery

Beyond satellite imagery, remote sensing is also applied using aerial imagery and drones, which offer higher-resolution data and allow for more detailed analysis of specific areas. Drones, for instance, can fly lower to the ground, capturing high-definition images that reveal finer details that satellites might miss.

Synthetic Aperture Radar (SAR)

Another important tool in remote sensing is Synthetic Aperture Radar (SAR).

SAR technology uses radar waves to capture detailed information about the Earth's surface, even in areas with dense cloud cover or at night. This technology is useful for mapping contours and topography, helping to identify features like soil moisture, land subsidence, or even vegetation structure in areas where optical imagery may be less effective.

LiDAR (Light Detection and Ranging)?

LiDAR is a powerful technique that sends laser pulses toward the ground and measures the time it takes for them to return. This technology allows for the creation of highly accurate 3D maps of the Earth's surface, including detailed measurements of elevation, topography, and canopy structure.

LiDAR is especially valuable in forests and wetlands, as it can help assess the density and height of vegetation, track changes in landforms, and even map the ground beneath dense foliage.

How to make sense of and visualize remote sensing data

Remote sensing generates vast amounts of data, often in the petabyte range, which must be processed and transformed into usable insights. This is where data products come in.

Data products convert raw remote sensing data, such as satellite images, into actionable information—like land cover maps that categorize areas as water, forest, or urban. These products simplify complex data, making it easier for non-experts to make informed decisions.

GIS is one of the primary ways to visualize and interpret location-based data, widely used for mapping, spatial analysis, and geographic data visualization. It helps organize and understand information tied to specific locations.

GIS platforms like ArcGIS, QGIS, and Google Earth Engine help visualize and analyze this data. Google Earth, for instance, provides a no-code tool that allows users to overlay satellite images, zoom into specific areas, and create interactive visualizations. These tools make remote sensing data more accessible and useful, even for professionals in conservation and restoration.

In recent months, Google has made a number of updates to Google Earth to make the tool more useful for professionals involved in restoration and conservation work

Web platforms like Google Maps and Mapbox offer interactive tools for terrain mapping and environmental monitoring, while programming languages like Python and R provide flexible analysis options.

In short, data products—such as land cover maps and vegetation indices—transform complex remote sensing data into usable insights. These products are visualized through tools like Google Earth and GIS platforms, enabling better decision-making for conservation and restoration efforts.

Why use remote sensing & data products to identify land plots for nature restoration?

Identifying suitable land for nature restoration requires understanding its impact on local biodiversity. For example, if a protected species like the large blue butterfly is present, projects must minimize harm and, if necessary, offset impacts by restoring nearby areas.

To ensure restoration efforts genuinely benefit the species, the land must be evaluated for additionality—meaning the project should create a clear, positive impact that wouldn’t have occurred otherwise.

Remote sensing and data products can play a crucial role in assessing additionality, helping to verify that restoration initiatives lead to meaningful ecological gains.

How Versant uses remote sensing data to identify the best land for restoration projects

Versant helps project developers identify the best land for restoration and compensation projects by using satellite data, spatial analysis, and land use filtering.?

When evaluating a piece of land, Versant looks at satellite images to gather information about how the land reflects light. This is done by examining specific light wavelengths (also called spectral bands) that can show details about the land’s health, like vegetation, soil, and moisture levels.?

By analyzing this data, Versant generates measurements that help determine the land’s overall condition and suitability for restoration.

Along with evaluating land health, Versant also considers land use data to ensure the land fits the specific needs of a particular project. For example, if a client is not allowed to restore land that is being used for farming, the system can automatically filter out agricultural areas, so the project focuses only on suitable sites.

What does this process look like for Versant's clients?

For Versant’s project developer clients, the process of identifying and selecting land for restoration projects is streamlined and data-driven:

Step 1: Receiving the impact report

The process starts when a client shares an impact report, which explains their need for land compensation. This could be for a specific species or a habitat, like a wetland. The report helps Versant begin its analysis.

Step 2: Species distribution modeling

Next, Versant uses species distribution modeling. This means looking at data about where the species has been seen and combining it with climate information to predict where the species is most likely to thrive. This helps identify land that would be best for supporting the species.

Step 3: Refining the search with custom parameters

Once potential areas are identified, Versant applies their client’s specific requirements to further narrow down options. For example, if a client cannot restore agricultural land due to regulations, those areas are filtered out.

Step 4: Assessing land health and restoration potential

The next step is for Versant to evaluate how much improvement can be made on the land through restoration. Versant uses its habitat model to check the condition of each plot. The type of ecosystem (such as wetland or forest) is considered, and metrics like plant growth for wetlands or biodiversity for forests are applied. Versant also measures how well the land is regenerating over time to determine its potential for restoration.

Step 5: Compare to baseline

To ensure the best land is selected, Versant compares each parcel to a nearby baseline—a similar, healthy ecosystem. This helps gauge how much potential the land has for restoration compared to a well-functioning ecosystem.

Step 6: Final report and decision support

After completing the analysis, Versant provides their clients with a final list of the most suitable parcels for restoration. These parcels are the ones with the highest potential to support the targeted species or habitat. The client can then use this report to inform next steps, whether that’s for internal planning or coordination with field teams who will conduct further assessments.

Important points to consider when using remote sensing data for nature-based decision making

1. Higher the data resolution, higher the cost

Higher resolution, like 50cm per pixel, offers a lot more detail but generates extremely large files that require significant storage and processing power. This can be expensive, especially for large-scale projects. For global or regional analysis, medium resolution data (around 1km per pixel) is often more practical and cost-effective, providing enough detail for broader insights without the high costs of storing and processing high-resolution data.

The level of granularity and resolution needed depends on the specific questions an organization is trying to answer: For broader assessments, like pre-diagnostics, companies can work with less precise data to quickly identify potential areas of interest.?In some industries, like agriculture, higher-resolution data might be necessary. For example, precision agriculture could use drones with hyperspectral or multispectral sensors to monitor crops and track specific agricultural practices over time. This type of data helps companies improve regenerative practices or track crop health with more precision.

2. Timeliness needs

For some use cases, up-to-date data isn’t crucial. For instance, in biodiversity monitoring, companies can work with less frequent data updates, while industries like wildfire tracking demand real-time or near-real-time data to respond to immediate threats.

3. The importance of verifying model predictions with ground truthing data

Remote sensing data and models, such as species distribution models, are useful tools, but they are based on certain inputs, like climate variables and land use patterns. Sometimes, these inputs can be oversimplified or miss critical nuances, such as the influence of human activity or unique local conditions.?

To ensure models are kept realistic and applicable to real-world situations, it must be validated with ground truth data - ground truthing involves collecting data directly from the field to verify and refine the model's predictions. This step ensures that the assumptions made during the modeling process align with actual conditions on the ground. It also helps identify any potential biases or oversights in the model and ensures that the model's predictions are reliable and actionable for decision-making.

In summary

Combining remote sensing and ground truthing offers a cost-effective, scalable solution for nature-based projects. Remote sensing enables large-scale monitoring of environmental changes, while ground truthing ensures the accuracy of predictions.

By using remote sensing to identify potential sites and ground truthing to verify those findings, organizations can streamline their workflows, reduce fieldwork costs, and make more informed, faster decisions. This synergy of technologies is helping shape a nature-positive future, driving impactful conservation and restoration efforts at scale.

Watch the full playback here