Harnessing satellites, flux towers, AI, machine learning and MIR-spectroscopy to build a better picture of soil carbon stores
Soil carbon sequestration remains an important part of Australia’s emissions reduction policy but measuring soil carbon has been expensive and time-consuming.
It was argued that increasing soil carbon could offset greenhouse gas emissions by at least four per cent.
Now the race is on to develop faster, cheaper ways to monitor small changes in soil carbon concentration over vast distances. Such advances will support global efforts to reduce levels of carbon dioxide in the atmosphere, while also offering farmers an extra revenue stream.
Enormous potential for carbon storage
The 2015 United Nations Climate Change Conference recognised the potential for carbon storage as the world’s soils contain two to three times more carbon than the atmosphere.
It called on all countries to work together to increase the level of carbon stored in the soil, arguing that increasing the concentration of carbon in the top 30-40 centimetres by just 0.4% a year would soak up most of the annual increase of CO2 in the atmosphere.
That same year, early adopter Corporate Carbon (now AgriProve) registered Australia’s first soil carbon project at Niels Olsen’s property in West Gippsland, Victoria.
After setting the baseline, the project team was able to record increases in soil carbon which followed changes in farming practices.
In 2019, the farm generated Australia’s first soil carbon credits in the Emissions Reduction Fund.
AgriProve founder and Managing Director Matthew Warnken says Australia remains the only place in the world where soil carbon credits count towards the nation’s Paris target for emissions reduction. Other countries have unregulated, voluntary markets.
“Australia has more than 80 per cent of all the soil carbon projects on the measurement registries in the world,” Warnken says.
“So that is a testament to the farming community and their willingness to engage and be part of these solutions.”
AgriProve says it has become one of Australia’s leading soil carbon project developers, with more than 250 registered projects. It’s also the only company to have soil carbon credits issued through the Emissions Reduction Fund.
But there are many other companies now working with farmers to establish a baseline measurement of soil carbon now, in order to capitalise on changes in the future.
Using our eyes in the sky
In the US, two remote sensing scientists founded the company Perennial (formerly Cloud Agronomics) to “unlock soil as the world’s largest carbon sink”.
Leveraging their experiences at NASA, co-founder David Schurman and chief scientist Dr Jim Kellner have begun building a soil carbon measurement, reporting and verification platform which will ultimately use remote sensing data in place of physical sampling.
“We can take advantage of lessons learned during the same transition in forests over the last few decades, where remote sensing has become the international gold standard that replaced on-the-ground measurements,” Kellner says.
“Our technology reduces or eliminates the need for physical soil sampling. Reducing the sampling burden will drive down the cost of generating carbon offsets in agricultural land.”
Last year Perennial created the highest resolution map of soil carbon across the US.
But Perennial Australia Managing Director Oli Madgett says mapping relies on more than satellite data alone. A lot of effort is going into “ground truthing” that data.
The US also has a huge independent soil carbon dataset, which includes extensive laboratory analysis of soil cores from sources such as the US Department of Agriculture.
While Australia has world-leading soil carbon measurement methodology, the dataset is a long way behind that of the US, because there is no central repository, yet (the federal government has set aside $8 million to develop a national soil carbon dataset as part of the National Soil Strategy).
Across Australia, Perennial is testing remotely sensed data against the physical data from soil cores on “blind validation farms”, to further develop and refine their model.
“We take physical soil cores down to 30 centimetres although increasingly to a metre — and then we marry that data from known points on the Earth, with the remotely sensed imagery from space,” Madgett says.
“Then we also bring into our model other geospatial data layers that have a relationship to how carbon is distributed across the landscape, so things like topography or rainfall or other covariants which have an impact on carbon.
“And we’re using machine learning to help model out soil carbon.”
Satellites can only see the soil when it’s bare and then, only what’s at the surface, but Madgett says the model can also make inferences from different forms of vegetation: “if we don’t have visibility then we can use reflectance from biomass as an alternative input into the model”.
Towers of power in gas exchange
At the Queensland University of Technology, Global Change Professor Peter Grace and Associate Professor David Rowlings are using “eddy covariance flux towers” and models to assess changes in soil organic carbon at scale on grazing land.
It’s part of Meat and Livestock Australia’s plan to become carbon neutral by 2030. The flux towers are aligned with the Terrestrial Ecosystem Research Network.
Grace says the towers offer an inexpensive alternative to capture the temporal change and spatial variability in soil carbon across a landscape.
“If you use a flux tower, you can actually look at carbon dioxide exchange between soil, plants and the atmosphere. You’re getting this data in real time,” Grace says.
“It’s collecting changes in the CO2 concentration of the atmosphere and that’s a product of plant uptake from photosynthesis, soil decomposition, and CO2 from root respiration.”
“We then use ecosystem simulation models to accurately derive changes in soil carbon”.
The towers stand three to nine metres tall, so they can capture data about the change in CO2 “over many, many hectares”, up to 50 or even 100 hectares if the tower is high enough.
The flux towers form part of a regional reference network and the information will feed into more sophisticated models to enable the grazing industry to make use of soil carbon credits at scale.
“Small changes over large areas add up to being very significant,” Grace says.
A National Innovation Challenge
On 1 December 2021, then Minister for Industry, Energy and Emissions Reduction Angus Taylor released a new Emissions Reduction Fund method for soil carbon, “allowing modelled estimates of changes in soil carbon for the first time”.
It was argued that increasing soil carbon could offset emissions by 4–16 per cent, but measuring soil carbon abatement was a major barrier to project uptake.
Reducing the cost of soil carbon measurement to less than $3 per hectare per year (a ten-fold reduction or more from the going rate) was deemed a priority.
The government issued a $50 million National Soil Carbon Innovation Challenge to determine the feasibility of lower-cost, accurate technical solutions for measuring soil organic carbon.
In the first funding round, 17 projects were successful with total funding of just over $1 million.
The national science agency CSIRO (Commonwealth Scientific and Industrial Research Organisation) received the largest grant of $100,000 for in-situ (on site) infrared sensing for low-cost, accurate detection of soil carbon.
CSIRO will also work on Australia’s national soil spectral library, “empowering rapid soil organic carbon measurement”, with a grant of $50,000.
Proximal soil sensing is where a spectrometer is held close to, or in direct contact with, the soil. These scans use the mid-infrared (MIR) and/or near-infrared (NIR) part of the electromagnetic spectrum of (non-visible) light to infer the chemical composition of a soil, including the carbon content, depending on the wavelengths absorbed or reflected back.
Some of that CSIRO technology has been patented and commercialised. The company Ziltek, based in Thebarton, South Australia, sells a hand-held MIR device called RemScan which is mainly used in the assessment and remediation of oil and gas contamination.
RemScan Chief Executive Officer Dr Sean Manning says “the next killer app” for the technology is “inexpensive soil carbon measurement.”
It will also aid in the determination of sampling points under the ERF method by providing maps of surface soil carbon variation.
Another homegrown company, Hone Carbon, is involved in several grants to develop their NIR device Hone Lab Red, which is handy for soil carbon as well as grain and leaf testing.
An announcement about larger Development and Demonstration grants, expected in April had been delayed indefinitely.
In response to questions from Cosmos, a spokesperson for the Department of Climate Change, Energy, the Environment and Water says: “Remaining funding for the program will be considered by Government.”
Co-benefits, trading and trade-offs
Carbon Storage Partnership Director at the University of Tasmania, Associate Professor Matthew Harrison, is a strong advocate for building soil carbon, because there are so many benefits.
But he says it’s important to recognise there are limits – four to five per cent soil organic carbon is about the maximum – and potential pitfalls. He advises against selling credits and encourages farmers to work towards becoming a carbon neutral business instead.
“Building up soil carbon is just good practice, it results in improved productivity, improved water holding capacity, improved drought resilience, it adsorbs more nutrients and so therefore, your soils are more fertile … So that’s all good, that’s fine and that actually results in improved productivity,” Harrison says.
“So as a practice, I would fully advocate it, but in terms of trading soil carbon for money, I probably wouldn’t go there, because soil carbon is ephemeral. It can go up for a short time for five or 10 years, but then it will probably go down again. In the long term, if you look at soil carbon, it follows climate, it follows rainfall.”
Clare Peddie joined Cosmos in 2022 after 15 years as Science and Environment Writer at The Advertiser in Adelaide. She previously worked at CSIRO Land and Water, UniSA, and the Cooperative Research Centre for Vaccine Technology. A passionate science communicator with an extensive and ever-expanding network of contacts, she thrives on exploring new ideas with talented people. Follow @ClarePeddie