🌎 A framework for realistic ocean-only and coupled ocean + sea-ice simulations driven by prescribed atmospheres and based on Oceananigans and ClimaSeaIce.
ClimaOcean is a registered package. To install from a Julia REPL:
julia> using Pkg
julia> Pkg.add("ClimaOcean")
julia> Pkg.instantiate()Use Pkg.add("url=https://github.com/CliMA/ClimaOcean.jl.git", rev="main") to install the latest version of ClimaOcean.
For more information, see the documentation for Pkg.jl.
Why? What's the difference between ClimaOcean and Oceananigans?
Oceananigans is a general-purpose library for ocean-flavored fluid dynamics. ClimaOcean implements a framework for driving realistic Oceananigans simulations with prescribed atmospheres, and coupling them to prognostic sea ice simulations.
Our system for realistic modeling is anchored by ClimaOcean.OceanSeaIceModel, which encapsulates the ocean simulation, sea ice simulation, prescribed atmospheric state, and specifies how the three communicate.
To illustrate how OceanSeaIceModel works we set up a simulation on a grid with 10 vertical levels and 1/4-degree horizontal resolution:
using Oceananigans
using Oceananigans.Units
using Dates, CFTime
import ClimaOcean
arch = GPU()
grid = LatitudeLongitudeGrid(arch,
size = (1440, 560, 10),
halo = (7, 7, 7),
longitude = (0, 360),
latitude = (-70, 70),
z = (-3000, 0))
bathymetry = ClimaOcean.regrid_bathymetry(grid) # builds gridded bathymetry based on ETOPO1
grid = ImmersedBoundaryGrid(grid, GridFittedBottom(bathymetry))
# Build an ocean simulation initialized to the ECCO state estimate on Jan 1, 1993
ocean = ClimaOcean.ocean_simulation(grid)
date = DateTimeProlepticGregorian(1993, 1, 1)
set!(ocean.model, T = ClimaOcean.ECCOMetadata(:temperature; date),
S = ClimaOcean.ECCOMetadata(:salinity; date))
# Build and run an OceanSeaIceModel (with no sea ice component) forced by JRA55 reanalysis
atmosphere = ClimaOcean.JRA55_prescribed_atmosphere(arch)
coupled_model = ClimaOcean.OceanSeaIceModel(ocean; atmosphere)
simulation = Simulation(coupled_model, Δt=5minutes, stop_time=30days)
run!(simulation)The simulation above achieves approximately 8 simulated years per day of wall time on an Nvidia H100 GPU.
Since ocean.model is an Oceananigans.HydrostaticFreeSurfaceModel, we can leverage Oceananigans features in our scripts.
For example, to plot the surface speed at the end of the simulation we write
u, v, w = ocean.model.velocities
speed = Field(sqrt(u^2 + v^2))
compute!(speed)
using GLMakie
heatmap(view(speed, :, :, ocean.model.grid.Nz), colorrange=(0, 0.5), colormap=:magma, nan_color=:lightgray)which produces
A second core abstraction in ClimaOcean is ocean_simulation. ocean_simulation configures an Oceananigans model for realistic simulations including temperature and salinity, the TEOS-10 equation of state, boundary conditions to store computed air-sea fluxes, the automatically-calibrated turbulence closure CATKEVerticalDiffusivity, and the WENOVectorInvariant advection scheme for mesoscale-turbulence-resolving simulations.
ClimaOcean also provides convenience features for wrangling datasets of bathymetry, ocean temperature, salinity, ocean velocity fields, and prescribed atmospheric states.
ClimaOcean is built on top of Oceananigans and ClimaSeaIce, so it's important that ClimaOcean users become proficient with Oceananigans.
Note that though ClimaOcean is currently focused on hydrostatic modeling with Oceananigans.HydrostaticFreeSurfaceModel, realistic nonhydrostatic modeling is also within the scope of this package.