Explanation of folder structure and contained data
script_data.zip:
- CLF_AC: empty folder for writing the classification results with AC-RF-model using the python script
- CLF_ACWC: empty folder for writing the classification results with ACWC-RF-model using the python script
- habitat_calval: contains two Geopackage-files (point data) with Sentinel-2 Rrs (ACWC) and Rb (AC) values and corresponding class IDs
- kd: - kd_estimated.csv = table with estimated Kd values used in the paper; 
      - LiDAR_Rrs_Kdestimation.pkl: table with LiDAR water depths and Sentinel-2 Rrs values to estimate Kd
- MODELS: the two random forest models as used in the paper exported as pkl-file, one for AC and one for ACWC
- python_code: contains the jupyter notebook file with all steps and further comments 
- R_deep: contains a txt-file with Sentinel-2 Rrs median values extracted in the deep water polygon using SNAP
- Rb: contains a geotiff with calculated bottom reflectances for each band as used in the paper
- Rrs: contains a geotiff with ACOLITE rertrieved Rrs [-] reflectances for each band as used in the paper
- z_model: - z_model_caldata.csv = LiDAR water depths and corresponding Sentinel-2 Rrs [-] values of band 665 nm to train the z-model
		   - z_model_valdata.csv = LiDAR water depths and corresponding modelled water depths (z_modelled)
shallow_water.tif = Geotiff with the shallow water area based on the 5 m depth contour and coastline

result_tifs.zipzip: contains the water depth raster and classification result rasters as used in the paper