Modis land surface temperature image downscaling using NDVI as a predictor with random forest regression.
This was the subject of my undergraduate thesis during my study in NKUOA (National and Kapodistrian University of Athens) in the department of Physics in the Atmospheric and Environmental Physics sector. The purpose of this code is to downscale land surface temperature products availlable in 1000m resolution from the MODIS collection to 250m resolution by generating NDVI products from the same collection and using them as "predictors" for the training of a random forest regression algorithm, wich is used to produce simulated land surface temperature products in 250m resolution.