...a while since i last looked at that, tho I know we did this in the Castillo et al. (2012) work. We introduced the pft percentages through a clm.r file. I do not remember whether we started the run as "startup," "branch," or "restart," and I do not remember what other changes we made for the model to run without crashing.Looking at the code that I wrote many yrs ago in surfrdMod.F90, it seems unlikely to me that changing noveg alone would do anything. Subroutine surfrd_wtxy_veg_dgvm in surfrdMod.F90 is where pft weights get set to zero, so this may be another starting point for what you hope to do. Still, I'd be surprised if changes to this piece of code did not lead to inconsistencies that made the model crash.Bottom line: this can be done, I do not remember the details for how, and you will likely find that it's not the most straightfwd.Sam