Currently, if 15D_DEPTH_ZCUT = FALSE, and vturb_add is used to define additional micorturbulence for line broadening, then each successive batch of columns has vturb_add continuously added to atmos->vturb because this is never freed in readAtmos_hdf5. This happens even for the first set of columns. So, if vturb_add = 10 km/s, then atmos->vturb = 20 km/s for the first batch of columns, atmos->vturb = 30 km/s for the next batch, etc.,
This is not a problem if 15D_DEPTH_ZCUT = TRUE, since atmos-vturb is freed in realloc_ndep, called by setTcut.
I have added a statement to free atmos->vturb in readAtmos_hdf5 inside the conditional statements checking if setTcut should be called or not, which seems to fix the issue.
Currently, if 15D_DEPTH_ZCUT = FALSE, and vturb_add is used to define additional micorturbulence for line broadening, then each successive batch of columns has vturb_add continuously added to atmos->vturb because this is never freed in readAtmos_hdf5. This happens even for the first set of columns. So, if vturb_add = 10 km/s, then atmos->vturb = 20 km/s for the first batch of columns, atmos->vturb = 30 km/s for the next batch, etc.,
This is not a problem if 15D_DEPTH_ZCUT = TRUE, since atmos-vturb is freed in realloc_ndep, called by setTcut.
I have added a statement to free atmos->vturb in readAtmos_hdf5 inside the conditional statements checking if setTcut should be called or not, which seems to fix the issue.