Machine learning-driven characterization and prescription of aerosol optical properties for atmospheric models - Version 1.1

Code for Machine learning-driven characterization and prescription of aerosol optical properties for atmospheric models The codes provides information to process: 1) AERONET sky inversion dataset from Iberian Peninsula, using cluster analysis based on K-means method, in order to identify and characterize the typical aerosol optical regimes influencing the region. 3) MERRA-2 reanalysis dataset of aerosol fields (AOD at 550 nm and columnar mass density of aerosol types) to train and to test the prediction of the identified aerosol optical regimes using the Random Forest technique. Code 01 - produce the figure 2Code 02 - produce the figures 3,4,5Code 03(a,b) - produce figures 6,7,8Code 04 - produce figures 9,10Code 05 - produce figures 11,12Code 06(a,b) - produce figures 13,14,15 The code provided herein should provide enough guidance to reproduce all the figures of the manuscript.