In a previous blog, we talked about the need for carbon removal and the importance of natural climate solutions (NCS) to get it done. To ensure that carbon removal projects serve their objective they must fulfil certain quality criteria: Additionality and permanence.
Additionality means that the project item needs to store more carbon than it would under a business-as-usual scenario. Therefore, ensuring additionality requires accurate information on the baseline conditions (check our blog post on do’s and dont’s of the carbon market to see why it’s so crucial). Permanence, on the other hand, is the certainty that the carbon remains stored permanently and does not re-enter the carbon cycle. While the easiest and most effective way to achieve permanence is to avoid the extraction and incineration of fossil fuels altogether, in practice, this proves a tricky task since this option doesn’t exist for the removal of historic emissions.
Direct Air Carbon Capture and Storage (DACCS) and Bio-Energy with Carbon Capture and Storage (BECCS) can geologically store carbon for thousands of years, albeit being a resource-intensive undertaking (see the previous blog). The other possibility is through NCS by preventing degradation of current carbon sinks and restoring degraded carbon sinks: Natural, healthy ecosystems such as primary forests, peatlands, grasslands and mangroves are also capable of storing carbon for hundreds to thousands of years. And while NCS are typically more resilient to climatic factors than conventional land management, allowing the carbon to remain stored for long time periods, they still don’t imply permanence; the risk of leakage in this field remains significant. So what happens when this carbon stock does get leaked during forest fires or due to changes in climate or land use?
The voluntary carbon standard (VCS) addresses this problem in Agriculture, Forestry, and Other Land Use (AFOLU) projects by assigning a number of carbon credits to be deposited in the AFOLU pooled buffer account. This account holds non-tradable buffer credits in order to compensate for the non-permanence risk associated with AFOLU projects. Buffer credits are cancelled to cover carbon known, or believed, to be lost (also known as leakage). Effectively, this means that the sequestered carbon is higher than the number of carbon credits to make up for potential leakage. Be as it may, permanence requires a great deal of regular monitoring of the stored carbon.
Traditionally, creating baseline scenarios and monitoring carbon storage within the Monitoring, Reporting and Verification (MRV) process has been a long, tedious and costly undertaking, occupying up to ⅓ of project funds. The sampling had to be done over long periods of time just to create baseline scenarios in various locations, not to mention the monitoring necessary to verify additional carbon storage. Moreover, this monitoring process usually only takes place on a project scale. Hence, sampling results result in low data resolution which is available only for the specific areas monitored. This results in insufficient availability of reliable, consistent, and comparable, long-term data
Since the development of earth observation (EO) scientific research and technological advancement have made major leaps. What started as strictly military technology has now evolved to be one of the most important tools to monitor and analyze environmental data. However, until recently, satellite data has not been able to offer the reliability and high resolution necessary to enhance the monitoring processes of the strict voluntary carbon standard (VCS).
The development of new sensor technology, the commercialization of space exploration and EO, as well as the popularization of machine learning, promise to fill that gap. Measurements from space that were previously not possible can now be recorded accurately with 20 to 250 spectral channels, ranging from ultraviolet wavelengths to long-wave infrared. This technology delivers high-quality and highly useful data for environmental monitoring, while hyperspectral technology allows conclusions to be drawn on dynamic environmental influences. For instance, it is now possible to qualitatively assess the development of soils or vegetation. CarbonMapper, for example, has recently launched a satellite program that uses a hyperspectral satellite constellation to pinpoint, quantify and track methane and CO2 leakage.
Meanwhile, machine learning algorithms can analyze the overwhelming new spur of available information. Software as a service (SaaS) now promises to provide new advances in data collection for monitoring environmental data such as climate change mitigation. Machine learning techniques have been applied to historical remote sensing datasets of forests and deforestation with high-resolution data from recent years, leading to significant improvements in the proxies for the calculation of above-ground and below-ground biomass.
At SEQANA, we analyze EO data proxy measurements of soil properties using machine learning algorithms, calibrated with in situ measurements, to accurately predict the carbon content in the soil in a given area. These predictions can be used for baseline monitoring as well as the evaluation of additionality or leakage. More importantly, that data is available in high resolution, on large scales and over long time periods at a fraction of the cost. This greatly accelerates and streamlines the MRV process, allowing for more projects to enter the market at a faster pace in order to dissolve the bottleneck of carbon crediting through NCS.