Soil Salinity

Maps salinity risk and trends using ML with terrain, climate, land use, and remote sensing to guide mitigation.

This project maps soil salinization as a land degradation process using machine learning and multi-source environmental data. We link measured salinity indicators to drivers and symptoms captured by terrain/hydrology, climate, land use/irrigation proxies, geology, and remote sensing signals. Models are evaluated with spatially robust validation to ensure realistic generalization. Outputs include salinity maps and, when feasible, risk/confidence layers to guide monitoring and mitigation.

Main focus

  • Data-driven salinity mapping to support land degradation mitigation.
  • Integrate drivers (climate/terrain/land use) with remote sensing evidence.
  • Communicate risk and reliability to prioritize actions and sampling.

Objectives

  1. Compile salinity observations and relevant driver covariates for modelling.
  2. Train ML/ensemble models with spatial validation for realistic performance.
  3. Deliver salinity + risk/confidence layers for targeted monitoring and intervention.

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 — Global Soil Salinity Estimation

DOI




Paper 2 — Wavelet-Transformed SVR for Soil Salinity Prediction

DOI




Paper 3 — EM38 Inversion Mapping of Soil Salinity

DOI



References