If you are using single base rtk (no RTN, no static) for site surveys and boundary surveys, assuming bearings, and using the fugarwe button what geoid model do you use? Think... cowboy RTK (not best practice I know). I know the accuracy of the derived heights is dependent upon the accuracy of your base location. My questions are: Which geoid model would you use and why? Would you conduct any validity testing? How? I am Trying to convince the boss to at least load geoid12a. Is it worth doing given the above scenario or should I be content with geoid03? Advice?