Rice Crop Biophysical Parameters Retrieval from Sentinel-2 Imagery Using Parsimonious Multioutput Machine Learning Techniques

This dataset provides a Jupyter notebook used in training novel parsimonious multioutput machine learning models for simultaneous retrieval of the parameters from Sentinel-2 imagery, ~200 in-situ rice LAI and LCC measurements, study area shapefiles, prediction maps and SHAP feature importance plots output. Field campaigns were conducted in 2023 at Ahero Irrigation Scheme, Kenya, as part of Afri4CAst project.