Our science
Locally specific, globally scalable digital soil measurement
We take a multidisciplinary approach, combining landscape-scale analysis, remote sensing and targeted ground truthing.
By fusing environmental and climate data with machine learning, and our patented digital-twin methodology, we infer soil organic carbon at high resolution, consistently from 2017 to today.
Our science
Locally specific, globally scalable digital soil measurement
We take a multidisciplinary approach, combining landscape-scale analysis, remote sensing and targeted ground truthing.
By fusing environmental and climate data with machine learning, and our patented digital-twin methodology, we infer soil organic carbon at high resolution, consistently from 2017 to today.
Our science
Locally specific, globally scalable digital soil measurement
We take a multidisciplinary approach, combining landscape-scale analysis, remote sensing and targeted ground truthing.
By fusing environmental and climate data with machine learning, and our patented digital-twin methodology, we infer soil organic carbon at high resolution, consistently from 2017 to today.
10m x 10m
resolution
10m x 10m
resolution
10m x 10m
resolution
10-day
measurement cycle
10-day
measurement cycle
10-day
measurement cycle
9 years
of continuous records
9 years
of continuous records
9 years
of continuous records
800k
calibration samples
800k
calibration samples
800k
calibration samples
Aligned and validated to global standards
Aligned and validated to global standards
Aligned and validated to global standards
Land Sector and Removals Standard
Land Sector and Removals Standard
Land Sector and Removals Standard
ISO 14064-1:2019
ISO 14064-1:2019
ISO 14064-1:2019
ISO 14064-2:2019
ISO 14064-2:2019
ISO 14064-2:2019
Science-based Targets Initiative (FLAG)
Science-based Targets Initiative (FLAG)
Science-based Targets Initiative (FLAG)
Oxford Offsetting Principles
Oxford Offsetting Principles
Oxford Offsetting Principles
ICVCM Core Carbon Principles
ICVCM Core Carbon Principles
ICVCM Core Carbon Principles
audited by
audited by
audited by
Société Générale de Surveillance
Société Générale de Surveillance
Société Générale de Surveillance
Det Norske Veritas
Det Norske Veritas
Det Norske Veritas
Methodology
How we measure soil carbon
Methodology
How we measure soil carbon
Methodology
How we measure soil carbon
1
Functional classification
With machine learning, we identify land statistically similar to your area of interest across topography, climate and environment. This functional class is what makes each assessment locally specific.
1
Functional classification
With machine learning, we identify land statistically similar to your area of interest across topography, climate and environment. This functional class is what makes each assessment locally specific.
1
Functional classification
With machine learning, we identify land statistically similar to your area of interest across topography, climate and environment. This functional class is what makes each assessment locally specific.
2
Earth observation
We gather satellite data sensitive to soil organic carbon – across multiple spectral bands, every 10 metres, every 10 days, back to 2017.
2
Earth observation
We gather satellite data sensitive to soil organic carbon – across multiple spectral bands, every 10 metres, every 10 days, back to 2017.
2
Earth observation
We gather satellite data sensitive to soil organic carbon – across multiple spectral bands, every 10 metres, every 10 days, back to 2017.
3
Empirical modelling
We build the mathematical relationship between the satellite data and soil carbon data from the functional class, then apply it to your land.
3
Empirical modelling
We build the mathematical relationship between the satellite data and soil carbon data from the functional class, then apply it to your land.
3
Empirical modelling
We build the mathematical relationship between the satellite data and soil carbon data from the functional class, then apply it to your land.
4
Quality assurance
Every assessment is tested against internal accuracy criteria, and only released once it meets them.
4
Quality assurance
Every assessment is tested against internal accuracy criteria, and only released once it meets them.
4
Quality assurance
Every assessment is tested against internal accuracy criteria, and only released once it meets them.
our perspective
For the first time we can measure the living state of the ground itself – everywhere, consistently, over time. And once you can measure something, you can finally value it.
Professor Jacqueline McGlade
Founder & Chief Science Officer, Downforce
our perspective
For the first time we can measure the living state of the ground itself – everywhere, consistently, over time. And once you can measure something, you can finally value it.
Professor Jacqueline McGlade
Founder & Chief Science Officer, Downforce
our perspective
For the first time we can measure the living state of the ground itself – everywhere, consistently, over time. And once you can measure something, you can finally value it.
Professor Jacqueline McGlade
Founder & Chief Science Officer, Downforce
Our approach
What makes our science effective
Our approach
What makes our science effective
Our approach
What makes our science effective
01
Patented functional classification
Soil organic carbon can't be read from a satellite signal alone, the same reflectance means different things across a landscape. Functional classification resolves this by grouping land with similar physical and climatic behaviour, so carbon is inferred in context rather than in isolation.
01
Patented functional classification
Soil organic carbon can't be read from a satellite signal alone, the same reflectance means different things across a landscape. Functional classification resolves this by grouping land with similar physical and climatic behaviour, so carbon is inferred in context rather than in isolation.
01
Patented functional classification
Soil organic carbon can't be read from a satellite signal alone, the same reflectance means different things across a landscape. Functional classification resolves this by grouping land with similar physical and climatic behaviour, so carbon is inferred in context rather than in isolation.
02
Scalable measurement across time & space
Our method scales because its inputs are already global. Multispectral satellite data gives continuous coverage back to 2017, and established soil carbon datasets span enough of the world to build the model in any functional class, without requiring physical sampling at each site.
02
Scalable measurement across time & space
We check it, run exclusions for roads, buildings and woodland, and the science team quality assures every assessment before releasing it. What you see has already been verified against what the farm actually looks like. There is no six-month wait for someone to book a site visit.
02
Scalable measurement across time & space
Our method scales because its inputs are already global. Multispectral satellite data gives continuous coverage back to 2017, and established soil carbon datasets span enough of the world to build the model in any functional class, without requiring physical sampling at each site.
03
Embracing uncertainty
Variability is real in any natural system. We quantify it, account for it, and report conservatively using three year moving average data.
03
Embracing uncertainty
Variability is real in any natural system so, we quantify it, account for it, and report conservatively.
03
Embracing uncertainty
Variability is real in any natural system. We quantify it, account for it, and report conservatively using three year moving average data.
04
Calibrated & validated using global data
Calibration draws on soil carbon data from scientific datasets worldwide. Each carries different depths, methods and dates, so rather than discard what doesn't fit, we characterise its context and correct for it. Our calibration base widens with every assessment.
04
Calibrated & validated using global data
Calibration draws on soil carbon data from scientific datasets worldwide. Each carries different depths, methods and dates, so rather than discard what doesn't fit, we characterise its context and correct for it. Our calibration base widens with every assessment.
04
Calibrated & validated using global data
Calibration draws on soil carbon data from scientific datasets worldwide. Each carries different depths, methods and dates, so rather than discard what doesn't fit, we characterise its context and correct for it. Our calibration base widens with every assessment.