Effective Monitoring, Reporting, and Verification (MRV) is the cornerstone of high-integrity carbon farming projects. A well-known approach to quantify the effect of a carbon farming project is to measure sequestered soil organic carbon (SOC) in the project area and compare it to sequestered SOC in designated baseline control sites (“Measure & Re-Measure” in Verra’s VM0042). The required number of soil samples in the project area and in baseline control sites is a key driver of MRV costs and determines the uncertainty deductions enforced by carbon methodologies.

This presentation addresses a critical discrepancy between expected and actual sample size requirements when following the “Measure & Re-Measure” approach in a classical design-based estimation (DBE) framework. Sample size equations commonly found in the literature and in methodologies focus on temporal change rather than effect size (against a baseline). This leads to significantly lower sample sizes than required to achieve the desired statistical power. Apart from deriving more suitable equations, we analyze the role of sample allocation between project area and baseline control sites. Our research shows that the optimal sample allocation ratio can rarely be attained, as it demands practically infeasible sampling densities in the baseline control sites. At the same time, deviations from optimal allocation can inflate total sample sizes by a factor of three or more, directly threatening project economics.

We propose a two-fold remedy to this inefficiency. First, instead of applying a purely statistical criterion for sample size determination, we propose using the Economic Optimum Number of Samples (EONS). EONS is a holistic approach that integrates sampling costs, uncertainty deductions, and carbon credit prices. The economic optimum is attained where the cost of taking an additional soil sample equals the marginal revenue gained from lower uncertainty deductions. Second, we demonstrate how shifting from classical DBE to a Digital Soil Mapping (DSM) approach drastically improves project economics by lowering required sample sizes. We provide a method to determine EONS for DBE and DSM-based approaches, and argue that DSMs should be the preferred option wherever applicable.