kadyb / geomorph_clustering

High-resolution geomorphological clustering of Poland

Gomorphological clustering

This repository contains the code for “Toward geomorphometry of plains - Country-level unsupervised classification of low-relief areas (Poland)” article.

Reproduction

1. Open the geomorph_clustering.Rproj project file in RStudio.
2. Run 01_extract.R to generate a random sample from the rasters.
3. Run 02_clustering.R to train the data transformer and Gaussian mixture models (GMM). Finally, the model with the highest BIC value is selected.
4. Run 03A_predict_lowres.R to perform spatial clustering on low resolution (downsampled) rasters. The result will be maps with clusters (geomorphological units) and uncertainty.
5. Run 03B_predict_highres.R to perform high resolution spatial clustering. This operation is performed for all data in blocks of the specified size.

Dataset

The final high-resolution maps with geomorphological units and uncertainty, Gaussian mixture model and data transformation model, and low-resolution rasters (500 m) with geomorphometric variables are available in the Zenodo repository: https://zenodo.org/record/6415362. Note that in the project, rasters with 30 m resolution were used as input data. The "input_rasters" folder in the archive should be renamed to "rasters" to work with the shared code.