SOC

This project maps soil organic carbon (SOC) and, where possible, carbon stocks to support soil health assessment and climate mitigation. We integrate SOC measurements with covariates capturing terrain, climate, land cover/management, geology, and remote sensing signals (e.g., SWIR, time series, bare-soil composites). Machine learning models capture nonlinear drivers of SOC variability and are evaluated using spatially informed validation. Deliverables include SOC/stock maps and reliability layers that guide monitoring, reporting, and targeted carbon-smart practices.

Main focus

• High-resolution SOC (and stock) mapping for climate and soil health insights.
• Combine multi-source covariates to capture controls on SOC variability.
• Provide confidence information for robust reporting and decision support.

Objectives

1. Build SOC prediction datasets from soil measurements and environmental drivers.
2. Train and validate ML/ensemble models with spatially robust evaluation.
3. Produce SOC/stock maps (and confidence layers) for climate-smart management.

Graphical abstract

:::: {.row .justify-content-sm-center} ::: {.col-sm-10 .mt-3 .mt-md-0}

::: ::::

::: caption Graphical abstract summarizing the workflow and key results. :::

Papers (selected)

Paper 1 — Multi-Depth SOC Mapping

DOI

Paper 2 — SOC Mapping Using Machine Learning

DOI

Paper 3 — 3D Soil Organic Matter Prediction

DOI