This is a script that reads in Landsat-8 data, Esri Sentinel-2 10m land cover time series data and train a random forest classification algorithm to estimate fractional built cover at 30m scale. The trained model can be used to produce fractional land cover for other regions.

The script calculates 3 indices for WorldView2 data. The NDVI, the NDWI and the Built-Up Index.