Some revisitation of turbulent fluxes (#80)

* oncean only timestepping

* replace with zero

* small change

* fill ecco halo regions

* first commit

* working

* let's see

* try

* uff

* try like this?

* Try it on CPU

* better

* hmmm

* rename file

* try it now?

* more changes

* bugfix

* other bugfix

* better inpaint

* add comment

* remove vestigial code

* better

* let's see

* another fix

* show method

* let's use a barrier

* let's try like this?

* this should work?

* infer load on CPU

* show stuff

* try it out

* try it like this

* test

* test it out

* let's see

* let's go

* now it will work!

* let's go here!

* let's go!

* better!

* bugfix

* now it will work

* fixes

* maximum size

* performance measure

* test it out

* not working?

* another test

* gogogo

* let's see

* try it again

* try again

* another try?

* ok, now it works, remove workings

* let's go

* better progress

* hmmmm

* now it will work?

* allow interpolation on distributed grids

* let's go!

* small test

* bugfix

* try it like this

* try like this?

* try it now

* another bugfix

* add drag

* add bottom drag

* small correction

* bugfix

* now this should work!

* small correction

* ocean only model

* small fixes

* try it out

* try like this

* try this

* try it like this

* small bugfix

* add docstring

* bugfix

* try again

* bugfix

* bugfix again

* cmon!!

* typo

* tidying up a bit

* bugfix

* another typo

* limit _all_ the fluxes (also radiative ones)

* better explanation

* working now

* remove duplicate definition

* more changes

* omip branch

* let's go

* progress

* we're ready to go!

* change name

* bugfix

* bugfix

* add new stuff

* bugfix

* convert correctly

* give it another try

* all the stuff

* changes

* go ahead

* tests

* comparison to coare

* adding main

* some changes

* chnages

* maybe now it works?

* try it like this?

* hail mary!

* go now!

* try with forcing instead of immersed boundary

* set connected regions to zero

* remove small lakes?

* let's go with connected regions!

* Let's go!

* should be ready to run?

* this should work!

* small bugfix

* using Adapt

* changing the parameters for TabulatedAlbedo

* small bugfix

* try like this?

* cleanup a bit

* give it a try

* back to previous formulation

* this should work now!

* some changes

* updating

* 10 meters

* bugfix

* correct one

* add a maximum

* remove show

* this works

* fix ocean simulations

* change manifests

* guess this is the closest we get!

* Finally I got it!

* small fixes

* try it now

* bugfix

* remove vestigial files

* remove some vestigial files

* some small changes

* some small bugfixes

* more bugfixes

* removing useless interpolation

* some correcting

* some changes

* a comment

* more comments

* changing folder

* arch in radiation

* better default

* some changes

* name change

* another bugfix

* Update src/OceanSimulations/OceanSimulations.jl

Co-authored-by: Gregory L. Wagner <[email protected]>

* convert_to_native_frame

* correct reference name

* some changes

* check what happens with atmospheric velocity instead of du, dv

* add similarity theory to the model

* small typo

* change bulk coefficients function

* better comment

* bugfixes here and there

* add links

* correct line reference

* correct reference

* correct comment

* Some documentation of the radiation

* pretty massive bug in the iterative solver. Now onto checking the radiative fluxes

* removing some useless code

* remove the WarpedGrid from the omip

* typo of function name

* Update pipeline.yml

* add a couple comments

* fix the example

* near global example

* prototype on latlong

* should it work with the examples like this?

* remove docs manifest

* bugfixes in docs

* single column simulation works

* running

* bugfixes

* some restructuring

* add single column to docs

* changed filename

* this should work, now start with documenting

* updating Oceananigans

* small bugfix

* bugfix

* start without near global omip

* fix ecco test

* another bugfix

* Update src/OceanSeaIceModels/CrossRealmFluxes/roughness_lengths.jl

Co-authored-by: Gregory L. Wagner <[email protected]>

* Update src/OceanSeaIceModels/CrossRealmFluxes/tabulated_albedo.jl

Co-authored-by: Gregory L. Wagner <[email protected]>

* Update src/OceanSeaIceModels/CrossRealmFluxes/tabulated_albedo.jl

Co-authored-by: Gregory L. Wagner <[email protected]>

* correct typo

* Update src/OceanSeaIceModels/ocean_only_model.jl

Co-authored-by: Gregory L. Wagner <[email protected]>

* TemperatureDependentAirViscosity and ReynoldsScalingFunction

* update ScalarRoughnessLength

* correct FT in albedo

* bugfix

* bugfix

* FT for tabulated albedo

* some initial comments

* exposing al parameters

* better stability functions

* bugfixes

* bugfixxed

* some refactoring + adding flux examples

* some grammar improvements

* bugfix in docs

* some improvements

* bugfix

* revert to chloride

* changed pipeline

* use github documenter for the moment

* bugfix

* change measurement to reference

* conversion depending on k

* change to the

* change name

* bugfix

* namechange

* add surface flux tests

* change name to "gusty" velocity difference

* typo

* bugfix

* correctly include runoff in forcing

* bugfix

* speed up generate fluxes

* bugfix in the bathymetry

---------

Co-authored-by: Andre Nogueira Souza <[email protected]>
Co-authored-by: Gregory L. Wagner <[email protected]>

.buildkite/pipeline.yml

@@ -5,6 +5,8 @@ agents:
 
 env:
   JULIA_LOAD_PATH: "${JULIA_LOAD_PATH}:${BUILDKITE_BUILD_CHECKOUT_PATH}/.buildkite"
+  OPENBLAS_NUM_THREADS: 1
+  OMPI_MCA_opal_warn_on_missing_libcuda: 0
 
 steps:
   - label: "initialize"
@@ -39,14 +41,24 @@ steps:
     commands:
       - "julia --project -e 'using Pkg; Pkg.test()'"
 
-  - label: "documentation"
+  - label: "Run surface flux tests"
+    key: "tests_fluxes"
     env:
       CUDA_VISIBLE_DEVICES: "-1"
-      JULIA_DEBUG: "Documenter"
-      # TMPDIR: "$TARTARUS_HOME/tmp"
+      TEST_GROUP: "fluxes"
     commands:
-      - "julia --color=yes --project=docs/ -e 'using Pkg; Pkg.develop(PackageSpec(path=pwd())); Pkg.instantiate()'"
-      - "julia --color=yes --project=docs/ docs/make.jl"
-
+      - "julia --project -e 'using Pkg; Pkg.test()'"
+
+  # - label: "documentation"
+  #   env:
+  #     JULIA_DEBUG: "Documenter"
+  #   commands:
+  #     - "julia --color=yes --project=docs/ -e 'using Pkg; Pkg.develop(PackageSpec(path=pwd())); Pkg.instantiate()'"
+  #     - "julia --color=yes --project=docs/ docs/make.jl"
+  #   agents:
+  #     slurm_mem: 120G
+  #     slurm_ntasks: 1
+  #     slurm_gpus_per_task: 1
+
   - wait: ~
     continue_on_failure: true

.gitignore

@@ -18,17 +18,19 @@ docs/build/
 docs/site/
 docs/src/literated/
 
+# Output files 
+*.nc
+*.jld2
+*.mp4
+*.png
+*.svg
+*.gif
+
 # File generated by Pkg, the package manager, based on a corresponding Project.toml
 # It records a fixed state of all packages used by the project. As such, it should not be
 # committed for packages, but should be committed for applications that require a static
 # environment.
 # Manifest.toml
-
-*.nc
-*.jld2
-*.mp4
 *.DS_Store
 *.swp
-*.svg
-*.gif
 *.zip