Module 4 · Soil spectroscopy for digital soil mapping

Using MIR spectra to predict soil properties: pre-processing, calibration models and their use in digital soil mapping.

Format: 4 parts · Level: Advanced · Runs in: Google Colab (R) · Material: SoilFER Training Manual + Training Resources

Read the manual chapter

What you will learn

• load and explore MIR spectral libraries
• pre-process spectra (prospectr)
• calibrate and validate spectral prediction models
• use spectral predictions as input to mapping

Notebooks

Each notebook opens in Google Colab with the R runtime. Run the Setup cell first: it downloads the training project, installs the R packages and gets the course data. Then run the cells in order.

Notebook
Part 1 · Read and pre-process spectra	Open in Colab.ipynb
Part 2 · Spectroscopy calibration models	Open in Colab.ipynb
Part 3 · Digital soil mapping with spectral predictions	Open in Colab.ipynb
Part 4 · Additional exercises	Open in Colab.ipynb

Parts 2–4 continue from the previous part. In a fresh Colab session, run the Fresh runtime? cell at the top of the notebook first.

First time in Colab? Read Getting started.