Satellite Data For a Better Carbon Offset Market

Research shows that the current carbon offset market is ineffective in realizing climate goals. This problem can be solved with satellite imagery to improve plan, monitor, and predict the outcomes of future carbon offset projects.

The carbon offset market is a voluntary initiative where companies that contribute to global heating can compensate for their carbon emissions.

For example, there are non-profit corporations that run carbon credit programs to protect rainforests using the money from polluting companies.

While scientists have been critical of how such market initiatives are set up, they also pointed out there were benefits, such as a measurable action to stop climate change that otherwise would not have happened. But today, science is not so sure about the effectiveness of such programs.

Why predicting the future is hard

For example, recent research seems to indicate that most carbon offsets from the world’s largest vendor of carbon credits, Verra, do not represent genuine carbon reductions, making them worthless.

Researchers pointed out that only a handful of projects showed evidence of deforestation reductions, while the threat to forests had been overstated significantly for their projects, meaning that forests “saved” were not necessarily in danger.

Most of all, the research showed that it’s hard to predict the future and decide on a set of universal methods for carbon credit programs to protect rainforests. What could help in this context is to make the use of satellite data a requirement in MRV (Measurement, Reporting and Verification) systems of voluntary carbon market initiatives. In addition to terrestrial lidar data, machine learning and AI can help to better understand the local situation to make better assumptions and predictions for potential deforestation dangers.

Mapping and predicting possible deforestation areas

Satellite data is especially helpful to understand where deforestation has happened in the past and where future deforestation might happen. For example, deforestation happens at a faster rate along major roads and rivers, so mapping and analyzing such areas using satellite data might help to select potential deforestation areas to protect through carbon credit programs.

Predicting deforestation using satellite imagery is already an existing application that uses machine learning to automate this analysis process, using large volumes of public, historical satellite imagery that have been collected over many decades.

Using satellite imagery, AI and mapping technology will enable the improvement of current carbon offset programs and make them more effective, so that companies that buy carbon credits contribute to the preservation of biodiversity and rainforests.

A better understanding of a local situation through a common picture provided by satellite imagery can help make better assumptions about what is needed, where, and how this can be achieved.

Setting and achieving more realistic climate goals

Additionally, they will also help to solve other potential problems that result from deforestation reductions, for example, monitoring of local communities that might be affected as the result of a new conservation program and forced to move as their habitat suddenly becomes a protected area where they no longer have access to. To avoid potential tensions between authorities and indigenous communities, a common picture might benefit all involved parties. Also, satellite imagery can help make better assumptions on the size of potential deforestation reductions it might achieve, and avoid double counts of forest areas for carbon credit which have happened in the past. Finally, because satellite imagery is captured almost daily, it lends itself to interim evaluations of a project that might result in more realistic project goals in the short term, or measure the effects of major events such as elections that might have major deforestation consequences in the case of the Brazilian rainforests.