Farm-scale soil carbon auditing

Abstract

A novel method for soil carbon auditing at farm scale based on data value is presented. Using a map of carbon content with associated uncertainty, it optimizes stratified random sampling: number of strata, stratum boundaries, total sample size and sample sizes within strata. The optimization maximizes the expected profit for the farmer on the basis of sequestered carbon price, sampling costs, and a trading parameter that balances farmer's and buyer's risks due to uncertainty of the estimated amount of sequestered carbon. The stratification is optimized by a novel method (Ospats), an iterative procedure that re-allocates grid points to strata on the basis of pairwise differences between predictions and covariances of prediction errors. Optimal sample sizes are calculated from variance predictions by Ospats. An application on an Australian farm has shown that soil carbon changes across farms and regions can be audited effectively using the proposed method. It is concluded that sample bulking and returning to the same sites in subsequent sampling rounds are not recommendable.