Best ways to detect ideal areas for reforestation with sat data?

Hi HN! Is it possible to use only sat data available today to come up with ideal areas on which to try effective reforestation?

I wonder if currently such areas are decided by checking local geo data or physical surveys only instead of some sick machine learning and sat data. I know arable lands are monitored and can be prospected via sat data but I am looking for how to detect the best areas to try reforestation with a good degree of certainty that the reforestion is doable by sat data, before actually experimenting with the reforastion process in those places. I suppose this should involve a combo of terrain composition + land chemicals + rainfall data + wind patterns + actual green areas nearby + human presence + local tree species characteristics + historical tree coverage?