How to use AI in forest restoration and conservation

Image source: Pedro Abreu, MORFO
May 27, 2024

Technology is now indispensable for forest restoration and conservation, given the urgency and enormous proportions of forest degradation. Between 1990 and 2020, the equivalent of half of India was deforested worldwide. Of the 900 million hectares of degraded forest in the world, only 5% will be restored at the current rate of reforestation. That's why we need to rely on the best technology available to us.

In this article, find out how artificial intelligence is being used today in forest restoration and conservation projects, and the benefits this technology could bring to the world's forest ecosystems.

How can AI be used for forest restoration?

AI: a key tool for forest conservation

Artificial intelligence is increasingly used in forest restoration today. Thanks to it, long-used technology computer vision tools, such as satellite imagery, are being improved. AI can now measure canopy heights from satellite images, which have been used for decades to observe and analyze forests.

AI will undoubtedly be a revolutionary advance for the management and restoration of forest ecosystems, providing accurate data and facilitating informed decision-making. Conservation teams and researchers are increasingly using AI to monitor ecosystems, measure the impact of natural disasters and conserve natural resources. For example, the World Wildlife Fund (WWF) uses AI for its conservation projects.

A new article thus from Triple Pundit explains how WWF is using AI in various projects to advance conservation efforts. After the Australian bushfires of 2020, WWF collaborated with and local partners to deploy over 600 camera traps to track wildlife recovery. Images from these traps are analyzed using AI on the Wildlife Insights platform, enabling researchers to track species resurgence and engage a wider audience.

How MORFO uses AI for forest restoration

The same technology is now being implemented in forest vegetation restoration. At MORFO, a large-scale forest restoration company based in Brazil and France, we are also working on AI models capable of tracking species. MORFO is innovative in detecting not only mature trees, but also young shoots, thanks to our computer vision algorithms applied to our very high-resolution drone images. This is a major step forward for the restoration market, as it will enable us to detect the growth of different types of vegetation and project success rates at an early stage, facilitating less costly corrective action.

"Today, artificial intelligence is increasingly used to manage natural ecosystems, particularly to monitor their evolution on a large scale. However, applications vary between needs related to conservation and those related to ecosystem restoration." - Arnaud Lienhard, CTO of MORFO

What are the benefits of integrating AI into forest restoration projects?

The immediate benefits of integrating AI into restoration practices include rapid and accurate analysis of large areas. This is useful not only for monitoring the land after restoration, but above all for diagnosing the area prior to any planting. Analyzing the land before restoration is absolutely essential to guarantee effective long-term rehabilitation tailored to each project, as each ecosystem has its own differences and needs.

Precise, less costly monitoring

In the long term, AI will ensure sustainability by providing accurate and less costly monitoring. As a result, forest monitoring will be carried out more often, and will offer clearer, more accurate data.

The cost of monitoring forest evolution is one of the main barriers to restoration. Too high a cost means a reduction in the number of hectares restored. Too low a cost, on the other hand, is often synonymous with partial monitoring. A recent study by ETH Zurich reveals that, in Latin America, most monitoring efforts (74.3% for tree survival and 61.2% for growth) focus on tree planting, with less attention paid to biodiversity recovery, mainly by monitoring birds (37.5%), mammals (23.2%) and woody vegetation (24.3%).

Our results emphasize that, while a promising range of new forest restoration interventions are being implemented across Latin America, accessible and practical monitoring strategies [...] are urgently needed to support successful outcomes.Forest restoration in practice across Latin America (ETH Zürich, May 2024)

Predicting and reducing environmental risks with AI

AI will be extremely fast and more accurate at predicting environmental risks, enabling us to act more quickly when needed. Risks include situations such as:

  • In areas where vegetation is limited or grows too slowly, the integration of artificial intelligence into environmental management offers promising solutions. AI enhancement of carbon and biomass models will enable a better understanding of the factors influencing plant growth, leading to more precise interventions such as the supply of specific nutrients, irrigation regulation and selection of the most suitable species.
  • The recognition of insufficient biodiversity, signalling an ecosystem in danger, will be facilitated on a large scale thanks to the ability of AI models to count and identify individual plants and trees. These models can analyze aerial or satellite imagery data to precisely map plant species and estimate their numbers, enabling real-time monitoring of vegetation evolution.

Towards digital twins of forest ecosystems

But that's not all. Emerging innovations in AI promise to further increase the efficiency of forest restoration efforts. Digital twins are one example. A digital twin is a virtual replica of a physical object or system that can be used to simulate, monitor and analyze the behavior of its real-life counterpart. Already well-established in the building and industrial sectors, they are beginning to be applied to forestry. They will help with monitoring, management and prevention in forest ecosystem restoration projects, particularly in :

  • Simulation of forest growth taking environmental parameters into account: model seasonal variations, soil types and climatic conditions to predict how the forest will evolve in the future.
  • Monitoring the evolution of the site over time and in detail: measuring and analyzing canopy heights, carbon storage, biodiversity and other ecological indicators to assess the health and development of the forest ecosystem.
  • Virtual tours of a remote, hard-to-reach site: using technology virtual reality and drones to create immersive experiences, enabling researchers, project managers and other stakeholders to visit and assess the condition of the site without having to travel physically.

At MORFO, we work every day to research and develop our technology in order to restore native ecosystems more efficiently and rapidly. AI is one of the building blocks for successful large-scale, high-quality ecosystem restoration.

To find out more, read our articles on MORFO technology:

Chief Writer and Content Manager
Lorie Louque
- Paris, France
Share this article
contact us!
Our MORFO newsletter is sent out once a month. You'll find out about our plantations, team news and innovations we love.
Oops! A problem occurred while submitting the form.
2024 Morfo. All rights reserved.