Marco Albani and Alessandro Baccini met in the 1990s at the University of Florence, where both studied Forest Science. While Alessandro went on to develop state-of-the-art technologies to measure carbon dynamics in woody vegetation from space, Marco became a business leader on land use, climate change and natural capital. In 2021, Alessandro and Marco, together with Chloris co-founders Giulio Boccaletti and Mark Friedl, created Chloris to help scale nature-based solutions.

Chloris is a software company, operating at the intersection of space-tech and nature-tech. Our mission is to accelerate the global transition to net-zero and nature-positive economies with the most scalable, robust and reliable natural capital data, backed by best-in-class science.

Today, our software delivers 23 years of spatially explicit above-ground biomass stock and change data and insights (in tonnes of AGB and tonnes of CO2e). Our data shows the impact of all above-ground biomass gains and losses, including deforestation, degradation, growth and regrowth of live woody vegetation at 30m resolution and with pixel-level uncertainty. Our time series delivers annual data from the year 2000 onwards.

Today, many market participants rely on a combination of standard land use cover/ change approaches, field measurements and/ or expensive LiDAR campaigns to estimate the potential and impact of various nature-based solutions, from large scale forest conservation, to improved forest management, afforestation, reforestation, restoration and agroforestry projects implemented at the farm-level.

The Chloris software delivers direct estimates of above-ground biomass for any area of interest. The nature of our data and how we deliver it to customers unlock a number of key advantages for the design, development and monitoring of investments. It’s the combination of these differentiators that makes our data a true game changer for those organizations looking to bring more scale, speed, cost-effectiveness and integrity to nature-based solutions. Specifically, Chloris data has the following key differentiators.

Degradation and growth of vegetation
Our technology delivers spatially explicit data on vegetation dynamics. This means that we not only detect deforestation, but also more difficult to detect degradation, growth and regrowth of above-ground biomass - for every 30m pixel on the planet, no matter if it is defined as forest or not.

Quantified uncertainty at the pixel-level
Our data is spatially explicit and delivers quantified uncertainty for every 30m pixel for both biomass stock and change, at a 95% CI.

Scalability and Consistency in Space and Time
Our models use data from spaceborne sensors collected over long time periods at continental scale and capture the full range of variation in carbon density in space and time. Our software can be tasked at any scale, from individual plots, to entire jurisdictions or large numbers of individual small plots. Our time series approach ensures consistent data going back to the year 2000. The combination of these factors deliver a degree of scalability and consistency in space and time which is unrivaled in this sector.

Speed
The efficiency of our cloud-based software infrastructure and automated Quality Assurance/ Quality Control process allow us to deliver data and insights at the speed of business and at the click of a button.

Cost-effectiveness
The input data and model design, as well as scalability of our software solution unlock considerable cost-effectiveness to our customers seeking high-quality solutions that work at scale.

The Chloris technology is built with state-of-the-art machine learning algorithms that fuse data from multiple Earth observation satellites to provide wall-to-wall, annual measurements of the carbon stocks and change in woody vegetation (forests, shrubs, and mangroves), with pixel-level uncertainty going back to the year 2000.

The ability of the Chloris Platform to provide accurate data and insights is the result of multiple machine-learning and AI advances made by our science and engineering team. To ensure quality predictions, our machine learning models filter and pre-process input data for both quality and representativeness and create novel predictive features, which provide the foundation for our cutting-edge mapping algorithms.

Our time series approach is implemented in two main steps: First, we create wall-to-wall maps of above-ground carbon stocks at annual time steps with a spatial resolution of 30 m. Second, we process the annual estimates of carbon stock at each 30 m grid cell using a Bayesian time series approach that removes noise in the time series and identifies statistically significant changes and trends in carbon stocks across time.

Our models are trained and calibrated with spaceborne-LiDAR data. We only retain the highest quality input data for deriving predictive features and training our machine-learning models. In addition to the continental-scale training and LiDAR-based calibration, we can also apply local calibration with high-quality field data.

Learn more about our technology in our Technology White Paper.

The Chloris data is used by project developers and other market participants in the Voluntary Carbon Market. Main use cases and applications include:

Screening and feasibility: Rapid assessment of historic annual biomass stock and change dynamics since 2000, for efficient identification of high-potential project opportunities

Project development and origination: Improved project design and quality, including improved and more cost- and time-efficient field campaigns, reliable degradation and removals accounting, representative reference areas and dynamic baselines

Project and portfolio monitoring: Annual updates on dynamics within project area, reference area and leakage belts for improved risk management and stakeholder engagement.

Bespoke solutions: from digital twins, assessment of selection bias, performance benchmarks, forest carbon potential, and other solutions, our team stands ready to explore bespoke solutions with you.

Companies in the Food, Land-Use and Agriculture (FLAG) sector invest in the conservation and restoration of nature as part of their journey to reaching climate goals and regulatory requirements. Main use cases and applications of our data for FLAG companies include:

Emissions and removals accounting: accounting for all emissions related to above-ground biomass losses (deforestation and degradation), as well as accounting of removals realized by growth of vegetation (e.g. as a result of reforestation, restoration and agroforestry interventions or natural growth).

Insetting & supplier engagement: development of additional carbon insetting projects going beyond any compliance obligations for improved supplier engagement

Ongoing monitoring at scale: cost-effective, ongoing farm-by-farm monitoring, including more difficult to distinguish on-farm replanting vs actual deforestation.

Meet voluntary and compliance reporting obligations: enhanced footprinting, target setting, due diligence and reporting as per SBTI FLAG guidance, GHG Protocol, TNFD, EUDR and other major reporting frameworks.

The Chloris Platform is built for scale and designed to support jurisdictions in issuing high quality carbon credits. It is the only service in the market that can immediately and accurately provide 23 years of carbon stock and change data for entire jurisdictions, at a price point that is exponentially less than what it would cost to do so with other instruments.

Because the system is fully scalable from operational (project) to full jurisdictional scale, it ensures consistent accounting for nesting approaches.

The Chloris technology can also be adapted to specific jurisdictional needs, such as by incorporating jurisdiction-specific allometric equations or by integrating a jurisdiction’s field data in calibration and validation.

Our scalable cloud-based platform can also be licensed to jurisdictions to create their own instance of the platform, and customized to their analytical and reporting needs.

The Chloris Platform leverages 20 years of paradigm-shifting research in coupling Earth observation data with machine learning, conducted at Boston University and the Woodwell Climate Research Center by our Co-Founder and Chief Scientist, Dr. Alessandro Baccini. As a result, our technology has been extensively stress-tested and vetted for over a decade via peer-review in both the remote sensing and carbon science communities. In January 2023, VERRA recognised Chloris as an approved service provider for the collection of activity data at the jurisdictional scale for allocation to projects.

We deliver spatially explicit pixel level uncertainty estimates as well as aggregated at the site level, taking into account the spatial autocorrelation of the pixel-level errors. For each 30 m pixel, we provide a confidence interval with the estimate on the uncertainty on both stock and change, and give customers the chance to choose the confidence interval they would like to apply. The quantified uncertainty at the pixel level allows for more reliable quantification of the uncertainty at the polygon/ project level, the relevant scale for reporting uncertainty.

In addition to standard machine-learning validation approaches (testing the predictions against 20% of the training data set that the algorithm did not see previously), we also conducted an extensive validation effort to compare our AGB stock and change estimates against high-quality airborne LiDAR estimates (rather than conducting plot-to-pixel comparison). The validation results show that our approach is a reliable, scalable method capable of detecting and quantifying carbon stocks and their changes operationally.

Read the full Validation paper.

Reach out to info@chloris.earth to discuss partnership opportunities that are most valuable to your company.

Answers to our frequently asked questions

How remote sensing is fueling nature-positive approaches

In defense of nature-based carbon markets