-
Notifications
You must be signed in to change notification settings - Fork 9
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Clean up one degree simulation + docstrings + bug w
ECCORestoring (#…
…222) * Clean up one degree simulation * Clean up internals and progress with one degree simulation * fix docstrings * (default: ...) -> Default: ... * remove stray spaces * more cleanup * cleanup + fixes in docs * remove $ * fix ECCOrestoring test * fix comment * fix bug with ECCORestoring * uncomment debug * fix bug * cleanup * fixes * proper arch * add units --------- Co-authored-by: Navid C. Constantinou <[email protected]>
- Loading branch information
Showing
22 changed files
with
486 additions
and
589 deletions.
There are no files selected for viewing
302 changes: 76 additions & 226 deletions
302
experiments/one_degree_simulation/one_degree_simulation.jl
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,245 +1,95 @@ | ||
| # # Near-global ocean simulation | ||
| # | ||
| # This Julia script sets up and runs a near-global ocean simulation using the Oceananigans.jl and ClimaOcean.jl packages. | ||
| # The simulation covers latitudes from 75°S to 75°N with a horizontal resolution of 1/4 degree and 40 vertical levels. | ||
| # | ||
| # The simulation runs for one year, and the results are visualized using the CairoMakie.jl package. | ||
| # | ||
| # ## Initial setup with package imports | ||
| # | ||
| # The script begins by importing the necessary Julia packages for visualization (CairoMakie), | ||
| # ocean modeling (Oceananigans, ClimaOcean), and handling dates and times (CFTime, Dates). | ||
| # These packages provide the foundational tools for setting up the simulation environment, | ||
| # including grid setup, physical processes modeling, and data visualization. | ||
|
|
||
| using Printf | ||
| using ClimaOcean | ||
| using ClimaOcean.ECCO: ECCO4Monthly | ||
| using OrthogonalSphericalShellGrids | ||
| using Oceananigans | ||
| using Oceananigans.Units | ||
| using ClimaOcean | ||
| using CairoMakie | ||
| using CFTime | ||
| using Dates | ||
| using Printf | ||
|
|
||
| arch = CPU() | ||
| z = exponential_z_faces(Nz=30, depth=6000) | ||
| Nx = 120 | ||
| Ny = 60 | ||
| Nz = length(z) - 1 | ||
|
|
||
| grid = TripolarGrid(arch; z, size = (Nx, Ny, Nz), north_poles_latitude=55, first_pole_longitude=70) | ||
|
|
||
| @show grid | ||
|
|
||
| # ### Grid configuration | ||
| # | ||
| # We define a global grid with a horizontal resolution of 1/4 degree and 40 vertical levels. | ||
| # The grid is a `LatitudeLongitudeGrid` capped at 75°S to 75°N. | ||
| # We use an exponential vertical spacing to better resolve the upper ocean layers. The total depth of the domain is set to 6000 meters. | ||
| # Finally, we specify the architecture for the simulation, which in this case is a GPU. | ||
|
|
||
| arch = GPU() | ||
| z_faces = exponential_z_faces(Nz=40, depth=6000) | ||
| Nx = 360 | ||
| Ny = 180 | ||
| Nz = length(z_faces) - 1 | ||
| grid = TripolarGrid(arch; | ||
| size = (Nx, Ny, Nz), | ||
| halo = (7, 7, 7), | ||
| z = z_faces, | ||
| north_poles_latitude = 55, | ||
| first_pole_longitude = 70) | ||
|
|
||
| # ### Bathymetry and immersed boundary | ||
| # | ||
| # We retrieve the bathymetry from the ETOPO1 data, ensuring a minimum depth of 10 meters | ||
| # (depths shallower than this are considered land). The `interpolation_passes` parameter | ||
| # specifies the number of passes to interpolate the bathymetry data. A larger number | ||
| # results in a smoother bathymetry. We also remove all connected regions (such as inland | ||
| # lakes) from the bathymetry data by specifying `connected_regions_allowed = 3` (Mediterrean | ||
| # sea an North sea in addition to the ocean). | ||
| bottom_height = regrid_bathymetry(grid; | ||
| minimum_depth = 10, | ||
| interpolation_passes = 5, | ||
| major_basins = 3) | ||
| # For plotting | ||
| bottom_height[bottom_height .>= 0] .= NaN | ||
| fig = Figure(size = (1200, 400)) | ||
| ax = Axis(fig[1, 1]) | ||
| hm = heatmap!(ax, bottom_height, colormap = :deep, colorrange = (-6000, 0)) | ||
| cb = Colorbar(fig[0, 1], hm, label = "Bottom height (m)", vertical = false) | ||
| hidedecorations!(ax) | ||
| save("bathymetry.png", fig) | ||
| nothing #hide | ||
| #  | ||
| bottom_height[isnan.(bottom_height)] .= 0 | ||
| grid = ImmersedBoundaryGrid(grid, GridFittedBottom(bottom_height)) | ||
| # ### Ocean model configuration | ||
| # | ||
| # To configure the ocean simulation, we use the `ocean_simulation` function from ClimaOcean.jl. This function allows us to build | ||
| # an ocean simulation with default parameters and numerics. The defaults include: | ||
| # - CATKE turbulence closure for vertical mixing | ||
| # - WENO-based advection scheme for momentum in the vector invariant form | ||
| # - WENO-based advection scheme for tracers | ||
| # - `SplitExplicitFreeSurfaceSolver` with 75 substeps | ||
| # - TEOS-10 equation of state, see [`TEOS10EquationOfState`](https://clima.github.io/SeawaterPolynomials.jl/dev/#The-TEOS-10-standard) | ||
| # - Quadratic bottom drag with a drag coefficient of 0.003 | ||
| # | ||
| # The ocean model is then initialized with the ECCO2 temperature and salinity fields for January 1, 1993. | ||
| closure = (Oceananigans.TurbulenceClosures.IsopycnalSkewSymmetricDiffusivity(κ_skew = 1000.0, κ_symmetric = 1000.0), | ||
| ClimaOcean.OceanSimulations.default_ocean_closure()) | ||
| momentum_advection = WENOVectorInvariant(vorticity_order = 5) | ||
| tracer_advection = WENO(; order = 5) | ||
| ocean = ocean_simulation(grid; momentum_advection, tracer_advection, closure) | ||
| model = ocean.model | ||
| date = DateTimeProlepticGregorian(1993, 1, 1) | ||
| set!(model, T=ECCOMetadata(:temperature; date), S=ECCOMetadata(:salinity; date)) | ||
| # ### Prescribed atmosphere and radiation | ||
| # | ||
| # The atmospheric data is prescribed using the JRA55 dataset, which is loaded | ||
| # into memory in 4 snapshots at a time. The JRA55 dataset provides atmospheric | ||
| # data such as temperature, humidity, and wind fields to calculate turbulent fluxes | ||
| # using bulk formulae, see [`CrossRealmFluxes`](@ref). | ||
| # | ||
| # The radiation model specifies an ocean albedo emissivity to compute the net radiative | ||
| # fluxes. The default ocean albedo is based on Payne (1982) and depends on cloud cover | ||
| # (calculated from the ratio of maximum possible incident solar radiation to actual | ||
| # incident solar radiation) and latitude. The ocean emissivity is set to 0.97. | ||
| backend = JRA55NetCDFBackend(41) | ||
| atmosphere = JRA55_prescribed_atmosphere(arch; backend) | ||
| radiation = Radiation(arch) | ||
| nothing #hide | ||
| # ### Sea ice model | ||
| # | ||
| # This simulation includes a simplified representation of ice cover where the | ||
| # air-sea fluxes are shut down whenever the sea surface temperature is below | ||
| # the freezing point. Only heating fluxes are allowed. This is not a full | ||
| # sea ice model, but it prevents the temperature from dropping excessively | ||
| # low by including atmosphere-ocean fluxes. | ||
|
|
||
| # Closure | ||
| gm = Oceananigans.TurbulenceClosures.IsopycnalSkewSymmetricDiffusivity(κ_skew=1000, κ_symmetric=1000) | ||
| catke = Oceananigans.TurbulenceClosures.CATKEVerticalDiffusivity() | ||
| closure = (gm, catke) | ||
|
|
||
| # Polar restoring | ||
| @inline function restoring_mask(λ, φ, z, t=0) | ||
| ϵN = (φ - 75) / 5 | ||
| ϵN = clamp(ϵN, zero(ϵN), one(ϵN)) | ||
| ϵS = - (φ + 75) / 5 | ||
| ϵS = clamp(ϵS, zero(ϵS), one(ϵS)) | ||
| return ϵN + ϵS | ||
| end | ||
|
|
||
| restoring_rate = 1 / 1days | ||
|
|
||
| restoring_mask_field = CenterField(grid) | ||
| set!(restoring_mask_field, restoring_mask) | ||
|
|
||
| @inline sponge_layer(λ, φ, z, t, c, ω) = - restoring_mask(λ, φ, z, t) * ω * c | ||
| Fu = Forcing(sponge_layer, field_dependencies=:u, parameters=restoring_rate) | ||
| Fv = Forcing(sponge_layer, field_dependencies=:v, parameters=restoring_rate) | ||
|
|
||
| dates = DateTimeProlepticGregorian(1993, 1, 1) : Month(1) : DateTimeProlepticGregorian(1994, 1, 1) | ||
| temperature = ECCOMetadata(:temperature, dates, ECCO4Monthly()) | ||
| salinity = ECCOMetadata(:salinity, dates, ECCO4Monthly()) | ||
|
|
||
| FT = ECCORestoring(arch, temperature; grid, mask=restoring_mask_field, rate=restoring_rate) | ||
| FS = ECCORestoring(arch, salinity; grid, mask=restoring_mask_field, rate=restoring_rate) | ||
| forcing = (T=FT, S=FS, u=Fu, v=Fv) | ||
|
|
||
| momentum_advection = VectorInvariant() | ||
| tracer_advection = Centered(order=2) | ||
| ocean = ocean_simulation(grid; momentum_advection, tracer_advection, | ||
| closure, forcing, | ||
| tracers = (:T, :S, :e)) | ||
|
|
||
| set!(ocean.model, | ||
| T = ECCOMetadata(:temperature; dates=first(dates)), | ||
| S = ECCOMetadata(:salinity; dates=first(dates))) | ||
|
|
||
| radiation = Radiation(arch) | ||
| atmosphere = JRA55_prescribed_atmosphere(arch; backend=JRA55NetCDFBackend(41)) | ||
| sea_ice = ClimaOcean.OceanSeaIceModels.MinimumTemperatureSeaIce() | ||
| nothing #hide | ||
| # ## The coupled simulation | ||
| # | ||
| # Finally, we define the coupled model, which includes the ocean, atmosphere, | ||
| # and radiation parameters. The model is constructed using the `OceanSeaIceModel` | ||
| # constructor. | ||
| # | ||
| # We then create a coupled simulation, starting with a time step of 10 seconds | ||
| # and running the simulation for 10 days. | ||
| # We will eventually increase the time step size and end time as the simulation | ||
| # progresses and initialization shocks dissipate. | ||
| # | ||
| # We also define a callback function to monitor the simulation's progress. | ||
| # This function prints the current time, iteration, time step, | ||
| # as well as the maximum velocities and tracers in the domain. The wall time | ||
| # is also printed to monitor the time taken for each iteration. | ||
| coupled_model = OceanSeaIceModel(ocean, sea_ice; atmosphere, radiation) | ||
| coupled_simulation = Simulation(coupled_model; Δt=10, stop_time = 10days) | ||
| wall_time = [time_ns()] | ||
| coupled_model = OceanSeaIceModel(ocean, sea_ice; atmosphere, radiation) | ||
|
|
||
| simulation = Simulation(coupled_model; Δt=1, stop_iteration=10) | ||
|
|
||
| wall_time = Ref(time_ns()) | ||
|
|
||
| function progress(sim) | ||
| ocean = sim.model.ocean | ||
| u, v, w = ocean.model.velocities | ||
| T = ocean.model.tracers.T | ||
| Tmax = maximum(T) | ||
| Tmin = minimum(T) | ||
| umax = maximum(abs, u), maximum(abs, v), maximum(abs, w) | ||
| step_time = 1e-9 * (time_ns() - wall_time[1]) | ||
| @info @sprintf("Time: %s, Iteration %d, Δt %s, max(vel): (%.2e, %.2e, %.2e), max(T): %.2f, min(T): %.2f, wtime: %s \n", | ||
| prettytime(ocean.model.clock.time), | ||
| ocean.model.clock.iteration, | ||
| prettytime(ocean.Δt), | ||
| step_time = 1e-9 * (time_ns() - wall_time[]) | ||
|
|
||
| @info @sprintf("Time: %s, n: %d, Δt: %s, max|u|: (%.2e, %.2e, %.2e) m s⁻¹, extrema(T): (%.2f, %.2f) ᵒC, wall time: %s \n", | ||
| prettytime(sim), iteration(sim), prettytime(sim.Δt), | ||
| umax..., Tmax, Tmin, prettytime(step_time)) | ||
| wall_time[1] = time_ns() | ||
| end | ||
| coupled_simulation.callbacks[:progress] = Callback(progress, IterationInterval(1000)) | ||
| nothing #hide | ||
| # ### Set up output writers | ||
| # | ||
| # We define output writers to save the simulation data at regular intervals. | ||
| # In this case, we save the surface fluxes and surface fields at a relatively high frequency (every day). | ||
| # The `indices` keyword argument allows us to save down a slice at the surface, which is located at `k = grid.Nz` | ||
| ocean.output_writers[:surface] = JLD2OutputWriter(model, merge(model.tracers, model.velocities); | ||
| schedule = TimeInterval(1days), | ||
| filename = "surface", | ||
| indices = (:, :, grid.Nz), | ||
| overwrite_existing = true, | ||
| array_type = Array{Float32}) | ||
| nothing #hide | ||
| # ### Spinning up the simulation | ||
| # | ||
| # As an initial condition, we have interpolated ECCO tracer fields onto our custom grid. | ||
| # The bathymetry of the original ECCO data may differ from our grid, so the initialization of the velocity | ||
| # field might cause shocks if a large time step is used. | ||
| # | ||
| # Therefore, we spin up the simulation with a small time step to ensure that the interpolated initial | ||
| # conditions adapt to the model numerics and parameterization without causing instability. A 10-day | ||
| # integration with a maximum time step of 1.5 minutes should be sufficient to dissipate spurious | ||
| # initialization shocks. | ||
| # We use an adaptive time step that maintains the [CFL condition](https://en.wikipedia.org/wiki/Courant%E2%80%93Friedrichs%E2%80%93Lewy_condition) equal to 0.1. | ||
| # For this scope, we use the Oceananigans utility `conjure_time_step_wizard!` (see Oceanigans's documentation). | ||
| ocean.stop_time = 10days | ||
| conjure_time_step_wizard!(ocean; cfl = 0.1, max_Δt = 90, max_change = 1.1) | ||
| run!(coupled_simulation) | ||
| nothing #hide | ||
| # ### Running the simulation | ||
| # | ||
| # Now that the simulation has spun up, we can run it for the full 100 days. | ||
| # We increase the maximum time step size to 10 minutes and let the simulation run for 100 days. | ||
| # This time, we set the CFL in the time_step_wizard to be 0.25 as this is the maximum recommended CFL to be | ||
| # used in conjunction with Oceananigans' hydrostatic time-stepping algorithm ([two step Adams-Bashfort](https://en.wikipedia.org/wiki/Linear_multistep_method)) | ||
| ocean.stop_time = 100days | ||
| coupled_simulation.stop_time = 100days | ||
| conjure_time_step_wizard!(ocean; cfl = 0.25, max_Δt = 10minutes, max_change = 1.1) | ||
| run!(coupled_simulation) | ||
| nothing #hide | ||
| # ## Visualizing the results | ||
| # | ||
| # The simulation has finished, let's visualize the results. | ||
| # In this section we pull up the saved data and create visualizations using the CairoMakie.jl package. | ||
| # In particular, we generate an animation of the evolution of surface fields: | ||
| # surface speed (s), surface temperature (T), and turbulent kinetic energy (e). | ||
| u = FieldTimeSeries("surface.jld2", "u"; backend = OnDisk()) | ||
| v = FieldTimeSeries("surface.jld2", "v"; backend = OnDisk()) | ||
| T = FieldTimeSeries("surface.jld2", "T"; backend = OnDisk()) | ||
| e = FieldTimeSeries("surface.jld2", "e"; backend = OnDisk()) | ||
| # Set land values to NaN using the hacky method of deducing where T = 0 | ||
| times = u.times | ||
| Nt = length(u) | ||
| T1 = interior(T[1], :, :, 1) | ||
| land = T1 .== 0 | ||
| for n = 1:Nt | ||
| ui = interior(u[n], :, :, 1) | ||
| vi = interior(v[n], :, :, 1) | ||
| Ti = interior(T[n], :, :, 1) | ||
| ei = interior(e[n], :, :, 1) | ||
| ui[land] .= NaN | ||
| vi[land] .= NaN | ||
| Ti[land] .= NaN | ||
| ei[land] .= NaN | ||
| end | ||
| iter = Observable(Nt) | ||
| Ti = @lift T[$iter] | ||
| ei = @lift e[$iter] | ||
| si = @lift begin | ||
| s = Field(sqrt(u[$iter]^2 + v[$iter]^2)) | ||
| compute!(s) | ||
| end | ||
| fig = Figure(size = (800, 400)) | ||
| ax = Axis(fig[1, 1]) | ||
| hm = heatmap!(ax, si, colorrange = (0, 0.5), colormap = :deep) | ||
| cb = Colorbar(fig[0, 1], hm, vertical = false, label = "Surface speed (ms⁻¹)") | ||
| hidedecorations!(ax) | ||
| CairoMakie.record(fig, "near_global_ocean_surface_s.mp4", 1:Nt, framerate = 8) do i | ||
| iter[] = i | ||
| end | ||
| nothing #hide | ||
| #  | ||
| fig = Figure(size = (800, 400)) | ||
| ax = Axis(fig[1, 1]) | ||
| hm = heatmap!(ax, Ti, colorrange = (-1, 30), colormap = :magma) | ||
| cb = Colorbar(fig[0, 1], hm, vertical = false, label = "Surface Temperature (Cᵒ)") | ||
| hidedecorations!(ax) | ||
| CairoMakie.record(fig, "near_global_ocean_surface_T.mp4", 1:Nt, framerate = 8) do i | ||
| iter[] = i | ||
| end | ||
| nothing #hide | ||
| #  | ||
| fig = Figure(size = (800, 400)) | ||
| ax = Axis(fig[1, 1]) | ||
| hm = heatmap!(ax, ei, colorrange = (0, 1e-3), colormap = :solar) | ||
| cb = Colorbar(fig[0, 1], hm, vertical = false, label = "Turbulent Kinetic Energy (m²s⁻²)") | ||
| hidedecorations!(ax) | ||
| CairoMakie.record(fig, "near_global_ocean_surface_e.mp4", 1:Nt, framerate = 8) do i | ||
| iter[] = i | ||
|
|
||
| wall_time[] = time_ns() | ||
|
|
||
| return nothing | ||
| end | ||
| nothing #hide | ||
| #  | ||
|
|
||
| add_callback!(simulation, progress, IterationInterval(1)) | ||
|
|
||
| run!(simulation) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.