We have been thinking that the best initial condition for the panantarctic run might be to revert to WOA.
As part of the testing for the initial condition, and the dependence of southern ocean upwelling and abyssal overturning, we have run three cases:
global-01-v1 - the first attempt at a global 0.1° z* simulation, starting from ACCESS-OM2 spunup state.global-01-v2 - as above, but starting from WOAglobal-01-v3 - as for v2 but with the HYCOM1 vertical coordinate.
I've now run a quick analysis on the first ~30 years of these runs:
https://github.com/AndyHoggANU/mom6-diagnostics/blob/main/global-01/ComparisonWithPanant.ipynb
In short, we seem to have a problem with the creation of dense water -- it is pretty strong when we start from ACCESS-OM2 but seems to disappear after about 20 years of running from WOA. This problem is even worse in the HYCOM1 coordinate than in z*...
This needs more investigation, but I think the takeaway message is that, for now, we need to investigate more before switching to HYCOM1. And we may want to carefully consider our initial condition.
We have been thinking that the best initial condition for the panantarctic run might be to revert to WOA.
As part of the testing for the initial condition, and the dependence of southern ocean upwelling and abyssal overturning, we have run three cases:
global-01-v1- the first attempt at a global 0.1° z* simulation, starting from ACCESS-OM2 spunup state.global-01-v2- as above, but starting from WOAglobal-01-v3- as for v2 but with the HYCOM1 vertical coordinate.I've now run a quick analysis on the first ~30 years of these runs:
https://github.com/AndyHoggANU/mom6-diagnostics/blob/main/global-01/ComparisonWithPanant.ipynb
In short, we seem to have a problem with the creation of dense water -- it is pretty strong when we start from ACCESS-OM2 but seems to disappear after about 20 years of running from WOA. This problem is even worse in the HYCOM1 coordinate than in z*...
This needs more investigation, but I think the takeaway message is that, for now, we need to investigate more before switching to HYCOM1. And we may want to carefully consider our initial condition.