I have seen UHF frequencies being used very effectively for terrain classification including easily seeing extreme differences in polarimetric response between areas with established trees versus newly wooded areas. I would assume this is largely due to the longer wavelength than would be present, for example, at X band.