The credibility and financial viability of Soil Organic Carbon (SOC) projects hinge on accurate Measurement, Reporting, and Verification (MRV). Project developers need robust SOC estimates that leverage the novel scientific and technological innovations. The goal is clear: improve the precision of SOC stock estimates and use that increased precision to reduce required sample sizes and/or uncertainty deductions, thereby boosting overall project Return on Investment (ROI).

While Digital Soil Maps (DSMs) are one path to improving estimation, they are not the only, or always even the best option. Model-assisted Estimations (MAE) offer a powerful, scientifically rigorous alternative.

Accepted within leading standards like the Gold Standard and Australian Clean Energy Regulator, MAE is a long-standing approach, which enables projects to significantly improve their economic viability. By reducing uncertainty deductions and/or decreasing sampling costs while not requiring model validation or an Independent Modeling Expert (IME) Review, MAE adds scientific rigor without adding unnecessary project complexity. Leveraging MAE, project developers can increase their project’s return on investment while maintaining scientific integrity in the claimed carbon credits.

‍

What is Model-assisted Estimation (MAE)?

MAE is a well-established statistical framework which reduces uncertainties in SOC stock (change) estimates. By pairing the predictions from a “working model” with design-based principles using the sampling design of a probability sample, the results remain unbiased while improving precision of the estimates of SOC Stock (change). 

What makes model-assisted estimations appealing is that their scientific integrity is maintained regardless of whether or not the assumptions of the model are met. The scientific rigor of model-assisted estimations depends on the correct implementation of the sampling design, not the performance of the model. As long as project proponents adhere to standard compliant  sampling design, it is not necessary to validate the model empirically before using it in the project context. At the same time, the higher the explanatory power of the working model, the greater the precision gains achieved through MAE. 

A key feature of model-assisted estimators is their unique decomposition into two components:

1. Model prediction: A prediction of SOC stock(change) based on the working model that explains part of the population variance of SOC Stock (change).
2. Bias-correction term using the sampling design: A crucial element estimated from the probability sample, which captures and corrects for any systematic prediction errors of the working model.

MAE leads to design unbiased results, making it an excellent approach for project developers that want to leverage precision gains without adding complex model validation and IME review requirements. 

We explore the technical details of applying Model-assisted Estimations to SOC projects in our whitepaper, which you can explore further here.

‍

How MAE Benefits Your Project Economics

Leveraging Model-assisted Estimation offers project developers three critical project improvements:

1. Reduce Uncertainty Deductions

Uncertainty deductions are an often underestimated factor that greatly affects the final project economics. There are effective, compliant strategies that reduce uncertainty deductions through greater precision of SOC stock estimations. MAE is one of those powerful approaches. 

MAE improves the precision of SOC stock estimations by helping to explain the variance in your SOC stock estimates. Project areas with a high SOC Stock (change) variance are not the issue; the problem is the unexplained or residual variance, variance not captured by the model, which leads to higher uncertainty deductions. By leveraging MAE, project developers can explain a greater portion of their SOC Stock (or stock change) variance, directly reducing the uncertainty deductions applied to the project’s carbon outcomes.

Want to dive deeper into how variance affects your project? Read our blog post: Debunking Variance Myths in Soil Carbon Projects

‍

2. Reduce Sample Size

MAE’s increased precision can also be leveraged to reduce upfront project costs. Project developers with a fixed targeted uncertainty deduction can achieve that uncertainty deduction with a smaller sample size when applying MAE compared to basing their SOC Stock (change) estimates on soil samples alone. 

‍

3. Avoid Model Validation and Lengthy IME Review Timelines

MAE is a tool that allows project developers to utilize precision improvements without model validation and without requiring an Independent Modeling Expert (IME) to review their results.

This benefit stems from the fact that MAE estimators are inherently design unbiased, provided the project developer adheres to a scientifically rigorous probability sampling design. The scientific integrity rests on the correct implementation of the sampling design, not on the empirical validation of the model itself before use. Our white paper describes the technical details on why MAE can be used safely without model validation and IME review. 

In methodologies that accept MAE, no IME review is currently required, streamlining the verification process and enabling project developers to benefit from the added precision without adding a large amount of project complexity to their project timelines.

‍

How does MAE compare to Digital Soil Maps (DSM)?

While both MAE and DSMs act as a precision layer to extract more value from each soil sample, they differ fundamentally in how they work and the type of value they provide. 

The magnitude of precision improvement from either approach is highly project-dependent: DSMs often deliver higher gains, but not always. MAE is design-unbiased, meaning it is technically impossible for the SOC stock (change) estimate to be biased, when sampling design is executed properly. As a result, MAEs do not require model validation or IME review, making their application simpler and faster. DSMs, by contrast, do require both model validation and IME review as safeguards to ensure that the resulting SOC stock (change) estimates are not biased. 

It’s critical to remember that choosing between MAE and DSMs is highly project specific. Both tools have the ability to offer precision gains, but each have benefits and drawbacks, which can vary depending on the project criteria. 

Implementing Model-assisted Estimations in your project

MAEs are one tool that can help project developers improve the precision of the SOC stock estimates of their project, without vastly altering their existing project workflows. MAE can be implemented at several points throughout the project lifecycle, enabling both new projects and projects that have already executed sampling campaigns to apply MAE and benefit from the outcomes of precision gains. 

‍

If you are curious to dive deeper into how Model-assisted Estimations work, you can find a comprehensive, technical overview of the methods and their application to SOC in our recent whitepaper.

Download the Whitepaper

‍

Want to dive deeper into how to leverage MAE or DSMs for your project? 

Seqana can help guide you through the process. 

‍