The true-up is a periodic check to ensure that a process-based model continues to reflect reality. At set intervals (for example, every five years), project developers must remeasure soil carbon stocks at the same locations sampled at project start. These new measurements are then used to confirm that the model remains unbiased. The exact requirements for how this is done are defined by the project’s applicable methodology, for instance Verra’s VM0042 together with the validation guidance in VMD0053.

During a true-up, the goal is to demonstrate that your model is still reliable. Under Verra’s VM0042, this requires showing two things: first, that the model explains soil carbon changes better than chance (R² > 0); and second, that it does not systematically overestimate sequestration in a way that could lead to over-crediting (i.e., Bias ≤ 0).

Want to dive deeper?
In Verra’s VM0042 (using VMD0053 for validation), passing the true-up typically requires:

1. Goodness of fit: The model must show a positive explanatory power (R² > 0).
2. Bias check: A statistical test (Welch’s test at 95% power) must demonstrate that the model’s bias in the project scenario is no greater than the model bias in the baseline scenario. Baseline bias can be estimated from either dedicated control sites (optional) or from the VMD0053 baseline dataset in case baseline control sites are not available.

The VMD0053 validation criteria of “90% of observed values within a 90% prediction interval” are not required at true-up, since they are impractical in SOC projects. Instead, the true-up focuses on goodness of fit and bias.

Different standards may set different rules, but Verra’s VM0042 and VMD0053 currently provide the most concrete example.

Note: While not part of the criteria for passing the true-up, it is important to note that the model prediction error and resulting uncertainty deduction must be recalculated using new project data.

If you take too few soil samples, you risk failing verification at true-up even if your model worked. If you take too many samples, the project becomes costly and inefficient. 

Want to dive deeper?

Under methodologies such as Verra’s VM0042, the true-up bias test must be passed with 95% statistical power. There is a minimum number of samples required to achieve this power for your project. Based on your specific inputs,the calculator helps estimate the minimum number of remeasurements needed, to give your project the best chance of passing if the model performs well in your area.

True-up sample sizes are often smaller than those needed for a full measure and re-measure design.

Want to dive deeper?
Measure and remeasure (e.g. as outlined in Verra’s Sampling Handbook) often requires a larger sample size to directly detect SOC change between two points in time with high confidence. By contrast, the true-up builds on the model validation framework, meaning fewer samples are often needed. But this efficiency comes with risk: if the model performs poorly in your project area and does not pass the true-up, you may find yourself without enough data to switch to measure and re-measure as a fallback.

Paired sampling compares two measurements from the same locations over time (e.g., SOC in the same geolocation at t₀ and t₁). It requires:

• Plot-level continuity (e.g., GPS-referenced sample sites),
  
  
• The ability to calculate change within each location,
  
  
• An estimate of the variance of change to calculate sample size requirements.
  
  

However, in SOC projects, paired sampling is often impractical because:

• Soil sampling is destructive, so it is impossible to re-sample the exact same soil,
  
  
• Re-sampling the exact same location is practically unfeasible. Due to GPS and operational limitations, resampling tends to happen within a 10m radius of the initial sample. Within a 10m radius, the spatial SOC stock variability tends to be higher than the expected change, which leads to mixing up spatial variability with actual change (see Myth 3 in our Variance blogpost here).
  
  
• An estimate of the variance of change is required to calculate the sample size. There is no scientifically robust way to estimate the variance of change ex-ante. 

Independent sampling compares two separately sampled groups, which allows you to:

• Sample SOC at t₀ vs. SOC at t₁, independently. In other words, no need to resample the exact same location
  
  
• Use the variance of SOC stock snapshot to calculate sample sizes.

In most SOC projects, independent sampling is more practical and defensible. That’s why we typically recommend using the 1-sided independent-sample MDD test design. It tends to require higher sample sizes though, so practical considerations have to be balanced with budget constraints and project economics. 

If you're unsure about the use of paired- vs independent sampling and what they mean for your sample size and/or project economics, reach out to Seqana for advice.

If the process based model (PBM) underperforms or fails in your project area, a sample size  derived solely for the PBM true-up will likely be insufficient to pivot to measure and re-measure (M&RM). To preserve that fallback and avoid delays, Seqana recommends determining the initial sample size based on the Economic Optimum Number of Samples (EONS) for M&RM until the PBM has passed its first in-project true-up. 

Want to dive deeper?
The sample size calculated for the true-up is based on ex-ante assumptions of the PBM performance from external validation data. These may not transfer perfectly to your project site. If the true-up later shows that the model is biased or inaccurate, and you don’t have enough samples for measure and re-measure, you cannot simply pivot your quantification approach to measure and re-measure. You would need to go back into the field and collect additional samples in the next verification period, delaying issuance and causing additional costs.

This is why it is strongly recommended to first calculate your sample size using the EONS (Economic Optimum Number of Samples) until you have passed the first true-up. After the first round of remeasurements, when you have evidence of how the model performs on your specific site, you can calculate sample size based on true-up requirements.

EONS = Economic Optimum Number of Samples. It balances the cost of sampling against the benefit of reducing uncertainty deductions and optimizing projects for economic performance. 

Diving deeper:
Before your project has in-project validation data, sample size planning should be based on economic and statistical trade-offs (EONS or MDD). This ensures you collect enough samples to:

1. validate the process based model in your area and
2. maintain the option of switching to measure-and-remeasure if needed. Once the model has been validated with project-specific data, you can use the true-up sample size calculator to fine-tune your sample size going forward.

Most methodologies require soil carbon stocks to be re-measured at least once every five years for true-up purposes. For a model true-up, these measurements must be taken at the same sampling units used at project start (i.e: paired sampling). The new data is then used to run the true-up tests and update the model’s performance assessment.

The inputs should be provided by the developer of your process-based model (PBM). This could be an internal modeling team or an external service provider. The inputs must  also be reported in Model Validation Reports (MVRs). For example, Verra’s VMD0053 (from version 3 onwards) requires all inputs needed for true-up sample size calculations to be presented in a standardized format within the MVR template.

If you’re unsure where to find a particular input, contact your PBM provider, or reach out to Seqana for guidance.

‍

Yes. If you’re not sure how to extract variances, interpret sample size results, or integrate fallback options into your monitoring plan, please reach out to Seqana. We’re happy to help!

‍