Add relaxed slab ocean mode #6798
Conversation
whannah1
force-pushed
the
whannah/docn/add-relaxed-slab-ocean
branch
from
8f1d9b3 to
5a79da2
Compare
|
| @@ -0,0 +1,60 @@ | |||
| # Data Ocean - Idealized | |||
There was a problem hiding this comment.
I suggest putting the data ocean docs in one file instead of splitting them up. the TOC mkdocs generates on the rhs will make it easier to navigate the material.
There was a problem hiding this comment.
I thought it would be too much material for a single page. I was thinking that it would be good to add some plots of my RSO tests to show what it does to the model in a longer run. And for the others I was thinking we might want to eventually add some instructions to generate new SST datasets. Maybe I'm suffering from premature optimization.
There was a problem hiding this comment.
I reorganized the data ocean docs to put it all on a single page.
|
I wanted to get a sense of the long-term impact of using this new RSO mode - both to check that the SST would stay close to the target values and what sort of systematic response could be seen in other fields and over land. The plot below shows 5-yr F2010 simulations (EAMv3) and differences between the RSO case and control. They both use the same data ocean data and the RSO is configured to use a globally uniform mixed layer depth. There is a systematic "el-Nino-like" difference in the SST... But obviously this is nothing like a real ENSO mode, and we can see warmer SST along most of the equatorial region, so I think we're just seeing the effect of warming where the insolation is the strongest. This sort of thing should disappear in a "traditional" SOM simulation because there would be strong cooling by Q-fluxes along the equator to represent the equatorial upwelling - which is not represented in the RSO mode. Similarly, there's a notable cooling along the western boundary currents and warming in regions of strong coastal upwelling that would be counteracted by Q-fluxes in a traditional SOM. The changes over land are reminiscent of maps of surface temperature regressed on an ENSO index, but I couldn't find a good example from a quick literature search. If my intuition/memory is right on this, then perhaps the land temperature changes are not that surprising given the pattern of SST changes. I think it's important for us to be aware of these systematic changes when using the RSO mode. For short TC hindcasts it likely won't be a problem, but I would not want to use this for longer simulations due to the biases that it will introduce.
|
whannah1
force-pushed
the
whannah/docn/add-relaxed-slab-ocean
branch
from
244c6b2 to
99ab6ba
Compare
whannah1
changed the title
|
@whannah1 any suggestions for who else could review this? |
|
Perhaps @singhbalwinder could take a look? |
| RSO_Tdeep = 271.00 ! deep water temperature [K] | ||
| RSO_dT_o = 27.0 ! scaling temperature gradient | ||
| RSO_h_o = 30.0 ! scaling mixed layer depth |
There was a problem hiding this comment.
Thanks for fixing some of these. If you do not want to add _r8, the numeric literals should be assigned to variables as integers to maintain precision. Assigning variables with numbers with decimals (e.g., 30.0) will reduce their precision to single precision.
| RSO_Tdeep = 271.00 ! deep water temperature [K] | |
| RSO_dT_o = 27.0 ! scaling temperature gradient | |
| RSO_h_o = 30.0 ! scaling mixed layer depth | |
| RSO_Tdeep = 271 ! deep water temperature [K] | |
| RSO_dT_o = 27 ! scaling temperature gradient | |
| RSO_h_o = 30 ! scaling mixed layer depth |
Being explicit and adding _r8 (e.g., 30.0_r8) is still a better approach, as it will make it clear to the reader that these are real numbers. Declaring them as parameters is also preferable if these variables will stay constant during time-stepping.
There was a problem hiding this comment.
ok, I like those suggestions.
The relaxed slab ocean (RSO) is implemented as a simpler alternative to the traditional slab ocean mode that requires externally generated heat flux data that represents the effects of ocean transport on ocean surface temperature. In RSO mode we want to relax to a target SST field using only a specified mixed layer depth to allow things like cold wakes following a tropical cyclone. [BFB]
|
Noting that this creates NML diffs in all data ocean cases. |
Second merge of this branch. Fix runtime error with EAM SCM cases.
The relaxed slab ocean (RSO) is implemented as a simpler alternative to the traditional slab ocean
mode that requires externally generated heat flux data that represents the effects of ocean transport
on ocean surface temperature. In RSO mode we want to relax to a target SST field using only a
specified mixed layer depth to allow things like cold wakes following a tropical cyclone.
Also add documentation for this mode of the data ocean model.
[BFB]