Closed AndyHoggANU closed 2 years ago
The anu-tub-sigma
case seemed to be overly warm at the bottom -- @pedrocol pointed out that this was likely due to poor resolution at the bottom. We need to refine the grid at the bottom as well as near the top.
Another option is to use Pedro's trick to vanish layers when they get too thin. This will require a bit of recoding, but is worth a trying.
In addition, we have tested whether we can start the sigma case from a zstar restart file -- and just allow the regridding on timestep 1 to take care of everything. This is tested for 10 years in anu-tub-sig-hotstart
and seems to work just fine.
Altered the structure of the vertical grid to have high resolution near the base as well as the surface. See notebook here: https://github.com/AndyHoggANU/anu-tub/blob/master/input/anu-tub/vertical-grids.ipynb Seems to be running and has done 30 years. Will close now.
Started a new case. Uses exactly the same config as
anu-tub-n3
except with SIGMA instead of ZSTAR as the vertical coordinate (andNOLEAP
instead ofjulian
calendar).Calling this
anu-tub-sigma
.