Commercialising radar-based detection of deforestation and forest degradation

Keywords: carbon credits; classification; deforestation; forest degradation; radar; REDD+; satellite data; tropical forest

Deforestation is accelerating across tropical forests and woodlands. As much as 20% of man-made greenhouse gas emissions come from tropical forest loss, ten times the emissions from air travel. However some damage to forest is hard to detect or quantify: in particular small-scale deforestation and degradation. Examples of this include a slash-and-burn farmer clearing a small field, or
where some trees are taken from a forest but it remains forest - e.g. selective logging for high value trees. This makes it very difficult to accurately estimate the area of forest destroyed by people each year. It similarly makes it hard to measure the success of projects that aim to halt or reverse forest loss.

When looking to monitor deforestation and degradation, pretty much the only available data source used is optical satellite data. Effectively these sensors are advanced digital cameras collecting images every time they pass over an area. Images are mostly available for free at resolutions of up to about 30 metre pixel size, with an image at this resolution typically taken everywhere in the planet every few weeks. However, there are two major problems:

1. Much of the tropics is cloudy most of the time. For some tropical forest areas, good cloud-free images are only available once every few years.

2. Optical satellite data can only see the top of the canopy, and can confuse trees with grass and shrubs. This means small-scale deforestation and degradation can often be missed.

We have produced a potential solution using a different type of satellite data: radar data. Radar can 'see through' cloud cover and the top of the canopy to discover the three dimensional structure of forests, solving the above problems. We know that long wavelength radar data can be used to map changes, but before our innovation it was not known that this could also be done with short wavelength data. There is no guaranteed provision of long wavelength radar data, just a collection of one-off satellites with data policies that do not allow commercial use without a significant charge. However, short wavelength data is far more readily available: the European Union has just funded a satellite series called Sentinel-1, which commits to providing consistent short-wavelength radar data into the 2030's from a number of satellites, and with the data provided free of charge for commercial use. NERC research at the University of Edinburgh has led to an algorithm that can use this data successfully.

On completion of this project we will be able to create maps of deforestation and forest degradation every quarter at a 20 m resolution. There is no comparable product available from optical data, and we have submitted a patent for this radar-based technology, so we mean to start the only company able to sell the most reliable forest change product available.

We have commissioned market research suggesting there are a wide number of potential users. Our primary market would be project developers that are generating carbon credits by protecting forest areas that were threatened with destruction: managers need satellite products to demonstrate historical rates of forest loss and to monitor the success of their projects. Such developers are already spending millions of dollars a year on inferior products.

We would also aim to sell to large multinational companies who are trying to remove deforestation from their supply chains: large companies such as Nestle and Unilever have already committed to this, but need to provide evidence of this to their customers. Finally there are opportunities to sell the product to timber companies (who may need to prove the source of their timber is sustainable for certification purposes), and biofuel producers, including those providing wood pellets to an increasing number of power stations.

There are four major groups that we expect to benefit if we manage to successfully commercialise a radar-based deforestation and forest degradation detection product.

1. All projects that aim to sell carbon credits for reducing deforestation and forest degradation (REDD+) need reliable data on past and current deforestation and degradation, and currently spend about 15 % of the income they receive from carbon credits on these monitoring costs (and potentially far more). Our product will benefit this sector in four ways. Firstly it will provide consistent data at a lower cost than current products in the market, as our free data source, scale and consistent automated production means we will be able to undercut our competitors. This will leave more funds to go back to local communities and to fund more projects. Secondly, projects will no longer have data gaps due to cloud cover, as radar data can see through clouds: this increases confidence in projects and reduces their costs related to the collection of alternative monitoring data. Thirdly our product may increase the price project developers can sell their carbon credits for, as the increased accuracy and transparency of our product will increase confidence in the sale (reduced monitoring costs and increased revenues should increase the attractiveness of REDD+ against other less-sustainable land use options). Finally, our product can detect forest degradation, which is very difficult or impossible to detect with optical data. This will make projects that attempt to halt degradation more likely to be commercially viable, allowing greater REDD+ participation by forest-rich African nations where forest degradation is particularly prevalent.

2. Governments, companies, and non-governmental organisations that purchase carbon offsets from forest carbon projects or otherwise invest in projects that aim to avoid forest loss (e.g. government aid projects) would like to know if their investments are actually having an impact on deforestation and climate change. As such they will benefit from the existence of consistent, transparent, cloud-independent quarterly deforestation/degradation data. They will be able to purchase these data to confirm independently that their carbon credits are real or their projects are working in a way that is simply not possible with current products.

3. Multinational corporations, including fast moving consumer goods firms, timber firms and the biomass sector need to be able to confirm to themselves and their customers that their products are 'deforestation free' or 'sustainably harvested'. Proving this currently involves employing consultants to produce one-off reports, as there is no consistent, trusted datasets available for purchase. For example many palm oil producers have signed the Sustainable Palm Oil Manifesto, committing to avoid deforesting high carbon stock forests, but are currently struggling to implement this as there are no consistent independent datasets of deforestation and degradation available for the cloudy areas where most palm oil is produced, SE Asia and west Africa. Our product will assist these companies by providing a single, global solution for forest loss monitoring that they can easily use for Corporate & Social Responsibility or certification purposes.

4. The global population will benefit from the existence of this product, since our prosperity ultimately depends upon services derived from functioning ecosystems. By allowing forest loss to be monitored anywhere, regardless of cloud cover or the extent of loss, and reducing costs for any organization that wishes to monitor forest loss we can increase transparency in land-use and forest management. We believe this could facilitate the reduction of the global rate of deforestation and degradation, therefore reducing greenhouse gas emissions and preserving ecosystem/climate services in living forests.