How AI and satellite data can improve scalability and transparency

As the level of ambition rises, so too does the need for effective methods to measure and document changes to the landscape across large geographical areas.

Traditionally, many assessments of natural values and restoration measures have been based on field surveys. In Sweden, the Swedish Board of Agriculture plays a central role in monitoring environmental and nature conservation measures in the agricultural landscape, where field-based surveys form an important part of the knowledge base. Such initiatives are extremely valuable and provide detailed knowledge that is difficult to replace. At the same time, they require time, resources and expert knowledge. When the aim is to monitor developments across thousands of agricultural plots or entire regions, there is therefore a natural need for complementary approaches. Nor is this need limited to public authorities. Stakeholders in the food sector, such as Arla Foods, also have a growing interest in being able to monitor, document and understand how natural values and restoration initiatives develop over time in the landscapes where their suppliers operate.

AI and satellite data open up exciting possibilities

Here, new technology is opening up exciting possibilities – which are currently being explored in a project led by RISE, with Arla, the Swedish Board of Agriculture and the Swedish Environmental Protection Agency as the stakeholders. Satellites are already collecting vast amounts of information about the Earth’s surface, and developments in artificial intelligence (AI) have made it possible to analyse these data sets on a scale that was previously unthinkable. By combining AI with satellite and airborne data, it is possible to automatically identify and track various landscape features over time, such as trees, shrubbery, small wooded areas, stone walls and pastureland.

When it comes to nature restoration, this is not about replacing people or ecological expertise. Rather, technology can act as a complement, helping us to direct resources to where they are most needed. If AI models can provide a general overview of large areas, it becomes easier to identify sites where more detailed surveys should be carried out. In this way, fieldwork can be used more strategically, whilst monitoring can cover significantly larger areas.

Comparability over time and a shared knowledge base

Another important aspect is transparency. When analyses are based on satellite data covering entire landscapes and on methods that can be repeated in the same way year after year, this creates better conditions for comparability. This can make it easier to track changes over time, document restoration measures and create a shared knowledge base for landowners, public authorities and other stakeholders.

At the same time, it is important to be realistic. AI is not a ready-made solution to all the challenges facing nature restoration. Certain landscape features are difficult to identify even in very high-resolution images, and smaller structures can be particularly challenging to detect in satellite data. The technology therefore needs to continue to be developed, whilst the results are quality-assured with the help of ecological expertise and field observations.

Despite these challenges, developments are progressing rapidly. The combination of AI and remote sensing has the potential to become a key tool for future nature restoration. At a time when both the level of ambition and the need for monitoring are increasing, this technology can help to make the work more scalable, more transparent and better adapted to the complex challenges facing our landscapes.