Some revisitation of turbulent fuxes #80
Conversation
…maOcean.jl into glw-ss/ice-ocean-model
…maOcean.jl into glw-ss/ice-ocean-model
… into ss/inital-conditions
… into ss/inital-conditions
| end | ||
|
|
||
| Base.summary(::PrescribedAtmosphere) = "PrescribedAtmosphere" | ||
| Base.show(io::IO, pa::PrescribedAtmosphere) = print(io, summary(pa)) | ||
|
|
||
| """ | ||
| PrescribedAtmosphere(times; | ||
| reference_height, | ||
| measurement_height, |
There was a problem hiding this comment.
go back to reference height
(this doesn't refer to any measurements, at least not for JRA55 which is our only example)
There was a problem hiding this comment.
can we call it something different? In COARE 3.6 reference height is something different. Probably field_height?
There was a problem hiding this comment.
What is the reference height in COARE 3.6?
There was a problem hiding this comment.
I think this like many other parts should have another PR and ideally an issue so we can discuss separately and come up with something well thought out
Co-authored-by: Gregory L. Wagner <wagner.greg@gmail.com>
…ean.jl into ss/omip-simulation
|
It is mostly reorganizing code. We just came out of a PR that added new 8800 lines of code (#49), so there is a lot of fine tuning we can do (for example just splitting the gigantic file The actual code contribution of this PR is not that large if you remove all the restructuring of the code (which is literally just moving code in new files), the new examples and the new Manifests/Project.toml files. |
Then I think the way I can review this PR is not to look at the changes / diff but rather to sit down together and read the code from top to bottom |
| # JRA55 atmosphere is adjusted to formulae without this last term so we exclude it | ||
| @inline simplified_bulk_coefficient(ψ, h, ℓ, L) = log(h / ℓ) - ψ(h / L) # + ψ(ℓ / L) | ||
| @inline simplified_bulk_coefficients(ψ, h, ℓ, L) = log(h / ℓ) - ψ(h / L) # + ψ(ℓ / L) |
There was a problem hiding this comment.
Not sure I agree with this logic. JRA55 is also nudged strongly to observations. I think we want to treat it like observations --- the objective is not to set up a simulation that is somehow "consistent" with the coupled system that JRA55 was originally run with.
There was a problem hiding this comment.
It would be interesting to see the difference between the two formulations (maybe it is not that large). These "similiarity functions" are all parameterizations that include empirical elements or relations that are extrapolated from data, so rigorous mathematical consistency does not necessarily equate to results that are closer to observations.
There was a problem hiding this comment.
That's not a bad idea, in that case something like
inline profile(ψ::SimplifiedProfile, h, ℓ, L) = ...might work.
PS this is the profile, not the bulk coeffcient right? It's what we get when we integrate the similarity gradient.
rigorous mathematical consistency does not necessarily equate to results that are closer to observations.
The purpose of including the term is not to be "rigorous". It's to avoid "random" approximations that rely on hidden assumptions. I think the same argument is actually the real reason to include it. If including the extra term means that this expression generalizes to cases with larger ℓ or small L --- but does not affect the results for typical values of these parameters --- that is exactly why we should include it. Approximations need to have a purpose. The purpose can't be to drop a term in a mathematical expression. It would actually have to make a difference in computational performance or somethign like that to justify this.
src/OceanSeaIceModels/CrossRealmFluxes/similarity_theory_turbulent_fluxes.jl
Outdated
Show resolved
Hide resolved
|
|
||
| Adapt.adapt_structure(to, α::LatitudeDependentAlbedo) = | ||
| LatitudeDependentAlbedo(Adapt.adapt(to, α.direct), | ||
| Adapt.adapt(to, α.diffuse)) |
This PR revisits surface fluxes to have them produce similar results to COARE 3.6.
In addition there are some OMIP-specific infrastructural changes
an ongoing PR subjected to changes