Elsevier

Agricultural Systems

Volume 169, February 2019, Pages 24-30
Agricultural Systems

Using model predictions of soil carbon in farm-scale auditing - A software tool

https://doi.org/10.1016/j.agsy.2018.11.007Get rights and content
Under a Creative Commons license
open access

Highlights

  • Introduction of a software package for optimised sampling design in the context of farmscale soil carbon auditing, using a grid of predictions with associated error.

  • Generates a stratified random sample from the optimised stratification and sample size.

  • Maximises the expected profit for the farmer on the basis of sequestered carbon price and sampling costs.

Abstract

We introduce a software tool for optimal sampling design in the context of farm-scale soil carbon auditing, where the amount of sequestered soil carbon will be estimated from a random sample. Existing tools do not use available ancillary information, or do not have the functionality needed for farm-scale soil carbon auditing.

Using a grid of predicted carbon content with associated uncertainty, the software optimises a stratified random sampling design, such that the profit is maximised 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.

As the algorithm is computationally intensive, the package is written in Julia for speed. From a case study we conclude that our software is an effective tool for farm-scale soil carbon auditing, and that it outperforms the existing tools in terms of efficiency and functionality.

Keywords

Soil carbon auditing
Stratified random sampling
Prediction error
Map uncertainty
Value of information
Julia

Cited by (0)