Land Suitability
Learns suitability patterns from soil and environmental constraints, producing spatial suitability maps and key limiting factors.
This project produces land suitability maps using machine learning to capture complex interactions among soil, terrain, and climate constraints. We combine soil and environmental covariates (e.g., texture/soil limitations, SOC, salinity risk, pH/carbonates, slope, climate variables) to predict suitability classes or continuous scores. Spatially informed validation ensures realistic generalization for mapping. Outputs include suitability maps, key limiting-factor summaries, and optional confidence layers for planning and management.
Main focus
- Data-driven land suitability mapping beyond fixed threshold/rule methods.
- Identify limiting factors that drive low suitability and management needs.
- Produce decision-ready suitability maps with realistic validation.
Objectives
- Compile suitability targets and constraint covariates for modelling.
- Train ML models and evaluate with spatially robust validation.
- Deliver suitability maps + limiting-factor summaries for land-use decisions.
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 — ML for Land Suitability and Sustainable Agricultural Production
Paper 2 — Multi-Criteria Decision Models for Wheat Suitability Assessment
Paper 3 — Regional-Scale Soil Management Zone Mapping Using ML
