Soil Erosion Prediction in Python

Contents

This project consists of two scripts:

1. magic.py creates a dataset of crops around the coordinate indicated in the name of satellite images coming from the Sentinel-2 source (Copernicus). The coordinates come from the LUCAS dataset, which pairs a C-factor with a x-y coordinate on the globe.
2. train.py trains a ConvNet-based model to solve a regression task consisting in predicting the C-factor from the crops (image crop not plan crop) extracted in step 1.

N.B.: C-factor accounts for how land cover, crops and crop management cause soil loss to vary from those losses occurring in bare fallow areas.

Install

Set up your Python environment as follows (ordering is important):

pip install --upgrade pip
conda install -c conda-forge gdal
pip install rasterio opencv-python tqdm numpy scikit-learn wandb tmuxp tabulate pyright ruff-lsp
pip install torch torchvision torchaudio
pip install fire