Stratification and Sampling Planning

How can you trust someone to plan a project for you if they’ve never set foot on your farm?

We start where it matters—in the field. We walk the paddocks, dig into the soil, read the landscape, and talk to the people who know it best. That’s how we ground-truth our work before a single map is drawn.

Then we layer in high-resolution data—terrain, vegetation, soils, and satellite imagery—carefully screened for what matters on your farm.

Because no two farms are the same—and carbon doesn’t follow generic rules.

Our first line of defence against soil carbon variation is smart CEA design.

CEAs aren’t just administrative zones, they’re tailored to reflect how you actually use your the land and how it naturally functions. We consider management nuances like paddock resting, crop rotations, and fence lines, alongside landscape features like slope, water flow, and soil type.

Why more CEAs? Because often, less assumption means better accuracy. By capturing this complexity up front, we reduce variability before we even start stratifying.

With CEAs defined, we focus on what’s happening inside them—identifying meaningful differences in soil carbon drivers using high-resolution spatial data and targeted analysis.

Each stratum is designed to group land that behaves similarly with respect to soil carbon. Our goal is to minimise variation within strata and maximise it between them—a core principle for building efficient, defensible sampling plans.

And no, our strata aren’t always neat little blocks. Soil doesn’t organise itself in tidy polygons—so why should your strata?

Our non-contiguous stratification reflects how carbon actually varies across your farm—irregular, patchy, complex.

Once strata are in place, we build a sampling plan that’s statistically rigorous and regulator-ready.

Sampling locations are selected using a pseudo-random number generator, seeded with publicly available data—like the ASX 200 index—to ensure complete transparency and reproducibility. This method ensures we can’t choose the points, and neither can anyone else.

For every stratum, we assign a full set of primary and reserve locations in advance. This avoids mid-sampling guesswork and keeps every decision clean, documented, and defendable.

It’s not just random—it’s verifiable. And that’s what makes it credible. This planning layer ensures the sampling itself is unbiased before the first shovel hits the ground.

In some regions—especially across WA—changes in soil carbon can be subtle, sometimes near lab detection limits.

That’s why we oversample at baseline, capturing real variation and building a strong foundation for future measurements. Cutting costs by sampling less increases the risk of missing genuine change — or falling short at audit.

We also carefully plan initial sampling depth. Because if you later disturb soil deeper than your baseline sampling depth, you may not be able to claim credits. The rules require your baseline to extend at least 10 cm deeper than any future disturbance—so we build in that flexibility from day one.

If you don’t measure your baseline properly, you can’t prove results—or get paid for them—with confidence.

When it’s time to sample, we use independent field teams—not employed by us, not involved in project design, and with no stake in the outcome.

Sampling follows a double-blind protocol: technicians don’t know which points are primary or reserve, and no changes can be made in the field. Every point is geo-verified, photographed, and documented. Any deviation is recorded with evidence.

Samples are sent to an independent lab under full chain-of-custody, ensuring results are tamper-proof and audit-ready.

This isn’t guesswork or gut feel—it’s controlled, traceable, and tamper-resistant.

We hold ourselves to the highest standards because soil carbon data must stand up to audit, regulation, and scientific scrutiny.