Currently, the redd QRF model contains 5 covariates, but 4 of them are also GAAs. Should we re-do the extrapolation model to extrapolate the 5th covariates, percentage gravel, across all master sample sites, and then make QRF predictions directly at all master samples sites? At the least, some kind of comparison between the two approaches seems warranted.
Currently, the redd QRF model contains 5 covariates, but 4 of them are also GAAs. Should we re-do the extrapolation model to extrapolate the 5th covariates, percentage gravel, across all master sample sites, and then make QRF predictions directly at all master samples sites? At the least, some kind of comparison between the two approaches seems warranted.