Tech companies hone in on carbon measurement
A keystone of carbon markets and how they work is the measurement of soil data. Today, there are nearly as many technology companies claiming to deliver solutions to measure carbon as there are programs for farmers to choose.
- READ MORE: Begin carbon farming by reducing erosion
The businesses run the spectrum from start-ups with prototypes or pilot programs to established firms working with partners in the industry and farmers in the fields. The technologies they offer run the gamut from algorithms to soil cores, artificial intelligence to satellites, or combinations of all these.
Carbon measurement infrastructure, which has been lacking in ag tech, can be classified in four distinct buckets. They all have very different approaches, says Chris Tolles, chief executive officer and cofounder of start-up Yard Stick PBC.
The four buckets are proximal, where measurement is taken on the field and in the soil; modeling, in which complex if/then statements are applied; remote sensing via satellites and other imagery solutions; and the status quo of conventional soil sampling.
Yard Stick believes proximal sensing will be the core of all soil carbon measurement for the foreseeable future, hence its innovation in that domain. The company was founded in 2021 by Tolles and Kevin Meissner alongside Cristine Morgan, chief scientific officer of the Soil Health Institute.
“We started Yard Stick because soil carbon removal has massive climate impact potential, and we want to get it from potential to reality,” Tolles says.
The Yard Stick team believes that ground truthing and touching the soil are the best ways to understand its makeup. Based on prior academic research of Morgan and others, they developed a soil probe that harnesses in situ spectroscopy to identify the spectral signature of soil carbon.
- READ MORE: Great carbon expectations
The probe is constructed with fiber optics and a sapphire lens that spins when drilled into the soil. It records a spectral movie along the 45 centimeters of the probe. At the hilt of the probe, torque and force sensors enable measurement of bulk density. Another sensor captures depth, which helps characterize the soil profile.
Data is collected on site without the need to send soil samples to labs for analysis. While Yard Stick’s process does rely upon teams of people on the ground using the probes, eliminating the steep cost of lab work is the primary cost advantage. Also, the simplicity of the device allows teams to cover more sample sites per day.
In addition to hardware, Yard Stick has built a software platform that serves the carbon project developers that are their customers. Tolles and his team are ensuring the data is high-quality, auditable, and useful.
“Yard Stick is unapologetically focused on the premium end of the soil carbon removal market,” he says. “Understandably, folks want measurement to be low-cost, but they also want price to be high per ton. You pay to get quality. We are focused on customers committed to quality and climate legitimacy who know they need to show their work.”
- READ MORE: Navigating the carbon markets
Yard Stick is currently running pilot programs to refine the soil probe. The company will be working on more than 100,000 acres this fall and expects to publish peer-reviewed data on the measurement technology next year.
“A big part of our culture is simultaneously being appropriately critical of the gaps that do exist in carbon measurement and being hope-filled,” Tolles says. “I think our technology can be something truly transformative to ensure that agricultural lands provide real climate benefits and that land managers are appropriately rewarded for their work combating climate change.”
Aadith Moorthy, chief executive officer of Boomitra, pivoted the company and technology three years after its founding to meet the demand of measuring carbon in the soil.
Boomitra uses proprietary artificial intelligence (AI) and remote sensing capabilities for monitoring, recording, and verifying carbon content, nutrients, and moisture levels in the soil.
“The function is supervised learning, which means to take input and produce some output,” Moorthy explains.
The input is real-time data from dozens of satellites and includes the electromagnetic spectrum, the normalized difference vegetation index (NDVI), and more. Boomitra also calibrates the AI per geographic region with local ground-truth data to ensure accuracy. An output is soil organic carbon measurements.
- READ MORE: Capturing carbon credits from grassland
“We measure soil carbon at absolute points in time,” Moorthy says. “The difference between two points in time is the CO2 absorbed by the soil due to a change in carbon sequestration by the farmer.”
Boomitra’s process eliminates conventional soil sampling and labs. Moorthy says the cost savings is upward of $100 per sample every 20 to 40 acres.
Boomitra functions across the entire chain of carbon farming. In addition to measuring carbon, the company delivers data for third-party certification and audits, sells the carbon credits, and returns the vast majority of each carbon credit to the farmer.
“The Boomitra technology enables us to evaluate carbon sequestration at the 10s of feet of scale on the ground. This means we can work with farms that have only a few acres. And we’ve built AI models in different parts of the world as we try to reach more farmers,” he says.
Boomitra serves farmers around the globe but is seeking to expand into grassland and cropland in the United States. The company continues to demonstrate traction along the life cycle of carbon measurement and payment.
Instead of assigning certain practices for farmers to implement that sequester carbon, Boomitra pays for performance. Moorthy recognizes that the farmer, the local agronomists, scientists, and other partners have the expertise on what crop and land management is successful and viable, and that as a technology company, Boomitra’s role is to support producers and increase their access to the carbon credit markets.
|Consider Components of Carbon Programs|
Travis Kraft, director of U.S. row crops with Locus Ag, recommends familiarizing yourself with the following components of carbon programs and their contracts when selecting what works for you.
“Educate yourself and ask questions,” Kraft says. “Make sure you know every single part of a carbon program’s system so that you are confident in its legitimacy and the long-term opportunities you have.”
Metaspectral has developed a platform that makes it easy for anyone with or without a technical background to use Deep Learning and AI to analyze hyperspectral data. The analysis, once performed, can be used in agriculture to detect greenhouse gases, quantify chlorophyll and nutrient levels, and detect early disease outbreaks in real time.
“The James Webb Space Telescope is a hyperspectral sensor,” Tissera says. “NASA has been working with this technology for over 20 years and it has come a long way. This technology has also become cheaper so it is feasible to use this data for more applications, including here on Earth.”
Hyperspectral cameras capture many frequencies of light; Up to 200x more than a typical color image that has only channels of red, green, and blue. They also collect data at a rate of gigabit per second. Companies working with this type of data need highly efficient compression and high-throughput analysis solutions due to the vast quantities of raw data that are produced, which is what Metaspectral can do.
Metaspectral has developed a platform that makes it easy for anyone to use Deep Learning and AI to analyze hyperspectral data. The analysis can be used in agriculture to detect greenhouse gases, quantify chlorophyll and nutrient levels, and detect early disease outbreaks in real time.
The advancement of computing power, the boom of artificial intelligence, and the accessibility of hyperspectral cameras enables a company like Metaspectral to perform data analysis for applications in agriculture, defense and security, space, manufacturing, and others. The company has a particular focus on implementations in sustainability.
“We operate at the convergence of high performance computing, deep learning (AI), and hyperspectral sensing. We feel like we’re ahead of the game on the computing and AI software front, and now that the hardware has also advanced significantly, we can apply our software to perform real-time analysis,” Tissera says. The platform is technology agnostic in terms of sensor hardware, i.e. data from any hyperspectral sensor can be analyzed on the platform.
As a part of a funding agreement with the Canadian Space Agency, Metaspectral’s spectral analysis platform was applied this year for greenhouse gas modeling (CO2 and Methane), and Tissera sees a bright future for the technology as they grow the business. This greenhouse gas monitoring software is already capable of predicting CO2 levels at surface of the Earth with only a 3% margin of error, using satellite hyperspectral data.
In agriculture, they partner with carbon project developers. Tissera believes we may see drones capturing hyper-localized data for carbon measurement, and the use of hyperspectral satellites for mapping of national levels of agriculture data in addition to data at field levels.
Tissera says, “Reaching our climate goals will require access to the best available data on emissions levels.”
Data-driven decisions, which are integral to carbon farming and markets, must be based upon quality data at every step in the chain.
Headquarters: Vancouver, Canada
Partners: SCALE Aerospace Accelerator, Canadian Space Agency, and more.
|Carbon Measurement Companies to Watch|
In addition to Yard Stick, Boomitra, and Metaspectral, other companies are providing tech solutions to measure soil data. Here are some to put on your radar: