Soil carbon measurement is one of the hardest things to get right in carbon accounting and one of the easiest to get wrong. Unlike combustion emissions, which follow predictable formulas, soil carbon shifts over time, varies from one patch of ground to the next, and hides at different depths. Getting a number you can actually trust takes more than a methodology. It takes a measurement system built to survive real field conditions and real audit scrutiny.
New to how emissions sources and sinks get quantified in general? How carbon emissions are measured is a good place to start before diving into the soil-specific complexity here.
Why Soil Carbon Is Harder to Measure Than Other Emission Sources and Sinks
What makes soil carbon different from combustion emissions
When a facility burns fuel, the emissions follow a known chemical reaction, immediate, measurable, predictable. Soil carbon doesn’t play by those rules.
It builds up and breaks down slowly, responding to tillage, cover crops, weather, and microbial activity that unfold over years, not minutes. That slow pace doesn’t make soil carbon harder to measure on its own — it just means you have to wait longer between measurements to see a real change. The actual challenge is getting an accurate number in the first place: solid field collection and consistent lab work, done right, every time.
How soil organic carbon variability affects measurement accuracy
Carbon levels can swing dramatically within the same field, partly because of texture, drainage, topography, or land-use history, and partly for no explainable reason at all. Every field has some natural randomness in it that no map or model can fully predict.
That randomness is exactly what a thin sampling plan misses. Sample too few points and your average will look fine on paper while hiding real risk underneath, the kind of risk that surfaces the moment an auditor starts asking questions. It’s a bigger threat to a project’s credibility than most teams expect (Seqana has a good breakdown of just how much small-scale variability alone can throw off a project’s numbers).
Why depth matters more than most projects assume
Most projects focus on the topsoil, roughly the first 30 centimeters. But carbon lives throughout the soil profile, and how it’s distributed depends heavily on soil type and management history.
Measure only the surface and you risk over- or under-stating what’s really there and missing carbon that has moved deeper, which matters for permanence claims. Sampling deeper isn’t automatically better either: dig too far down and you start diluting the carbon signal with soil that never held much to begin with. Depth is a balance to get right, not a box to check.
The Main Methods Used for Soil Carbon Measurement
Laboratory analysis: the traditional standard
Dry combustion is the accepted standard for measuring soil organic carbon today. A sample goes to a lab, gets burned at high temperature, and the carbon released is measured precisely.
What it gives you is a carbon percentage. Turning that into the actual carbon stock a project needs to report takes one more step most people don’t think about: factoring in bulk density, essentially how dense the soil itself is. Skip that step, or eyeball it, and your final number can be off before you’ve even started.
The real limitation isn’t accuracy, it’s scale. Lab results take days to weeks, which is rarely the bottleneck for a carbon project’s timeline, but every sample costs money. That cost is what really limits how many samples a large project can afford to collect, and fewer samples means less confidence in the final number.
Field sampling protocols: what makes or breaks a sampling plan
The quality of any soil carbon measurement comes down to the sampling plan behind it. A strong plan accounts for variability across the field through stratification, grouping similar areas together instead of sampling blindly — picks a sensible depth instead of just meeting the bare minimum, and balances how many samples you take against what you can afford.
Skimp on any of that and the results might look fine on paper but won’t hold up under review, one of the most common reasons soil carbon projects run into trouble at carbon verification. Even small mistakes in the field, like losing soil during extraction, can throw off the final number more than people expect. More often than not, it’s the sampling design, not the lab, that’s the weak link.
Proximal sensing soil carbon: what changes with direct contact technology
Proximal sensing means measuring soil carbon up close or in direct contact with the soil, instead of from a distance. Optical spectroscopy, reading how soil reflects light across different wavelengths, from visible to infrared, is one of the most widely used ways to do this, translating an optical signal into a carbon estimate.
The real dividing line is between remote sensing, which works from satellites or drones, and direct contact sensing, which physically touches the ground. Touching the soil gives a cleaner, more precise signal, since it skips the interference from air, moisture, and surface conditions that remote methods have to work around. That precision is what makes the difference when a project needs numbers strong enough to stand up to registry review.
Why Models Alone Are Not Enough
What soil carbon models actually do
Biogeochemical models like RothC, Century, and DNDC simulate how soil carbon changes over time using inputs like climate, soil properties, and management practices. Done well, they help projects cover more ground with fewer direct measurements — but only when they’re properly grounded in real field data. On their own, without that anchor, their output isn’t something you can defend under audit.
Where models break down without strong field data
A model is only as good as the data behind it. Without enough real measurements to calibrate and check it against, a model can spit out numbers that look precise but don’t hold up under scrutiny and even a well-calibrated model has limits on how confident it can really make you.
Registries take this seriously: under Verra’s rules, biogeochemical models need to be validated by an Independent Modelling Expert, not just any auditor. And leaning too heavily on a model with little real field data behind it, tends to push uncertainty up and the credits a project can claim down.
The Hybrid Approach: Where Rigor Meets Scale
How field data and modeling work together
The strongest soil carbon measurement systems don’t pick a side, they combine direct field measurement with modeling. Field measurements are the anchor: the ground truth that gives a model something real to work from. Regular re-measurement in the field is what keeps that relationship honest as conditions change over time.
What a defensible hybrid system looks like in practice
A defensible hybrid system has a few things in common: sampling driven by where the variability actually is, not just convenience; enough field measurement to keep uncertainty in check (more uncertainty simply means fewer claimable credits, not an automatic fail); a model calibrated on local data rather than borrowed from somewhere else; and a data trail, from field to final number, that’s fully documented.
This is the standard that serious soil carbon MRV frameworks are built around. Projects that build it in from day one reach verification without surprises. Projects that treat measurement as a cost to cut tend to find out too late that the savings weren’t worth it.
What Auditors Expect From Soil Carbon Measurement
Verification bodies hold soil carbon to a stricter standard than most other project types, simply because the underlying variability is higher. Under Verra’s VM0042, projects need to show that uncertainty stays within bounds, that sampling data is traceable from field to final output, and that any model involved was calibrated on real, local data.
Gaps in traceability and thin sampling are the two most common reasons soil carbon projects get delayed or docked credits at audit. For more on what that process actually involves, see carbon verification process.
How Direct Contact Proximal Sensing Changes the Equation
This is the tension ChrysaLabs was built to resolve: getting enough high-density field data, at a cost and speed that actually makes large-scale projects work.
Our direct contact soil sensing probe goes straight into the soil and scans the full profile down to one meter, every six inches, in about 30 seconds — capturing carbon where it actually lives, not just at the surface, the way remote sensing has to. That full-profile view strengthens permanence claims and makes long-term carbon numbers more reliable.
Probe data feeds into a validated model built on laboratory analysis to produce registry-aligned carbon estimates. The whole pipeline is automated, traceable, and built to withstand an audit. Curious how it works under the hood? See the Technology page.
ChrysaLabs is the first direct contact proximal sensing system validated by an independent auditor under Verra’s VM0042, with nearly 7 million tonnes of soil carbon quantified across more than 6 million acres. For regenerative agriculture programs and project developers working at scale, that track record means stronger credit volumes and more buyer confidence.
Rigor and scale usually pull in opposite directions. This is what it looks like when a measurement system delivers both — at a cost-to-accuracy ratio that makes bankable carbon assets achievable for projects of any size.
Soil Carbon Measurement Is Where Carbon Project Credibility Is Won or Lost
Every carbon project team knows soil is complex. What they underestimate is how much the measurement system decides the outcome. An otherwise well-designed project with weak measurement won’t survive verification. One built on defensible, traceable, registry-aligned data will.
The measurement approach isn’t a technical footnote — it’s one of the highest-leverage decisions a project makes, and it shapes everything downstream: sampling costs, credit volumes, buyer confidence. Getting soil carbon measurement right, from the first core sample to the final audit, is what turns a good project into a bankable one — and direct contact proximal sensing is quickly becoming the benchmark everything else gets measured against.
