Tropical deforestation contributes 12% of total anthropogenic carbon dioxide emissions globally1. Illegal logging is costing nations tens of billions of dollars each year. Although governments are making headway on agreements to stop this destruction, so far there is no coherent plan to monitor tropical forests on the scale or timescales necessary to do so.

An image from the UK-DMC2 satellite shows forest in the Brazilian Amazon in red and cleared areas in green (detail in inset). Credit: UK-DMC2/DMC INTERNATIONAL IMAGING

Incentives are being negotiated for states to implement the United Nations REDD+ framework: Reducing Emissions from Deforestation and Forest Degradation, extended to include conservation, sustainable management of forests and the enhancement of forest carbon stocks. The Intergovernmental Panel on Climate Change (IPCC) is also developing forest remote-sensing plans for consideration by the 19th Conference of the Parties (COP-19) to the United Nations Framework Convention on Climate Change (UNFCCC), which will be held in Warsaw this November.

Satellites provide the only means of viewing vast forest areas regularly — the tropics cover almost half of Earth's land area. But basic decisions have yet to be made on which Earth-observing systems should be used and how forest data should be monitored, reported and verified. In our view, the mapping strategies proposed so far are too sparse and slow, making it impossible to identify forest damage until at least a year later.

We believe that an early warning system is needed to allow authorities to stop illegal logging quickly. Two strategies are necessary to achieve this: first, a new set of tropical orbiting radar satellites that can 'see' through clouds to monitor global forests daily; second, a plan for existing satellites to assess forest carbon stocks several times a year, to account for seasonal variations.

The REDD+ working group, which meets in Bonn, Germany, from 29 April to 3 May, must agree on a comprehensive, rapid-response global monitoring system in time for it to be ratified at COP-19.

Real-time monitoring

More than one billion people depend on forests for their livelihood. The illegal logging trade is estimated to be worth between US$30 billion and $100 billion annually, with governments losing $10 billion each year in tax income2. Stolen wood is estimated to depress world timber prices by up to 16%.

As well as curbing these losses, implementing REDD+ and halving deforestation rates by 2050 would prevent the release of 50 petagrams of carbon in total into the atmosphere3,4, saving a potential warming of 0.14 °C. This is a small but crucial step towards potentially much larger carbon savings.

Implementing REDD+ will require many facets of the carbon cycle to be monitored, including tree cover and carbon emissions from felling. There is also a need to assess the state and health of forests by analysing colour changes due to chlorophyll loss. Biomass, dead organic matter and soils must be included in the carbon budget.

Because different types of forest vary in their capacity to store carbon, changes in land use must be recognized. For example, Indonesia's policy of converting its 91 million hectares of tropical forest to oil palm plantations reduces the land's value from a REDD+ perspective, because much less carbon is being stored. Around 72% of the frontier forest has already been lost.

Optical and radar satellites working in different parts of the electromagnetic spectrum can monitor all of these characteristics. But international bodies and governments have been slow to formulate and agree on common guidelines for so doing. Without the right technologies or globally accepted assessment standards, billions of dollars could be wasted on projects that do not deliver.

Illegal logging in Indonesia costs billions. Credit: SIGIT PAMUNGKAS/REUTERS

The European Union has focused on preventing illegal logging, rather than specifying how forestry information should be collected by countries signed up to its 2005 Forest Law Enforcement, Governance and Trade regulation. This has led to a proliferation of incomparable collection methods. Moreover, there is no plan for REDD+ carbon credits to be accepted in the European Union Emissions Trading Scheme.

Some groups have published guidelines for satellite monitoring for REDD+, including the UN Food and Agriculture Organization's Global Observation of Forest and Land Cover Dynamics programme and the Group on Earth Observations' Global Forest Observations Initiative. In our view, these reports lack ambition and an understanding of the potential of satellites. Another group, the Committee on Earth Observation Satellites, for example, foresees doing only one optical image survey a year — insufficient to track seasonal variations in forest carbon stocks.

To achieve good annual coverage, optical measurements should be taken at least every 1–2 weeks. To spot illegal logging, coverage needs to be even more frequent. Observations must be made at least daily, with data analysed much more quickly than at the current rate of months. If we are to protect tropical forests, we must go beyond mapping damage long after it has happened and instead offer early warnings.

Way forward

Governments and the IPCC need to accept that satellites are the only efficient and realistic way to provide monitoring for REDD+5,6. This must be enshrined in international law through the UNFCCC process and documented in detail. Governments must commit to securing and maintaining systems of Earth-observing satellites to perform the monitoring.

Remote-sensing satellites operate in two spectral regimes: optical and radar. Optical sensors in different colours are sensitive to vegetation greenness, fractional tree cover, forest type and vegetation density. They can cover millions of square kilometres in a single image while resolving detail to 20 metres, or down to 1 metre over a smaller area. Current Earth-observation efforts include NASA's Terra and Aqua satellites, the China–Brazil Earth Resources Satellite programme, the Disaster Monitoring Constellation (DMC) and France's SPOT satellite.

Radar systems, unlike optical systems, can penetrate cloud cover, which is common in the tropics. The microwave signals transmitted by Synthetic Aperture Radar (SAR) satellites, such as TerraSAR-X, are reflected back by leaves, branches and trunks to the satellite's receivers. Such satellites have historically been expensive, costing between £250 million (US$384 million) and £500 million each in orbit. Fortuitously, a new generation of lower-cost radar mini-satellites is due to be launched within the next year. These include the British NovaSAR-S range (£45 million launched and insured). The European Sentinel-1 mission and the Japanese ALOS-2 satellite each cost about ten times this.

Radar and optical systems are both needed. An early warning constellation of five tropical-orbiting radar satellites should provide daily scanning of tropical forests at 5–20-metre resolution, irrespective of weather, to monitor logging in real time. We estimate that this could be done for £200 million, plus a dedicated ground crew. Ideally, these data should be managed by an international consortium that includes developed and developing countries, and would be run under the auspices of the UNFCCC.

Low-resolution optical satellites can be constructed cheaply to monitor forests weekly or monthly throughout the year7. Existing satellite systems, including NASA's MODIS, DMC, SPOT or Landsat, could contribute, with much of the data made available for free.

But to provide regular ground data with which to calibrate all these satellite observations and reduce uncertainties in their carbon estimates, major efforts will also be required by those countries participating in REDD+.

As discussions continue this year, we urge policy-makers to back the right satellites and strategies to monitor and save the world's forests.