Skip to content

Latest commit

 

History

History
84 lines (72 loc) · 8.06 KB

CHANGELOG.md

File metadata and controls

84 lines (72 loc) · 8.06 KB

Changelog

Unreleased

  • Interfaces for interpolation of AbstractGridArray #671
  • Test folder sorted into subfolders #671
  • Land model modularised + land netCDF output #671

v0.14

  • CompatHelper: Allow for Makie.jl v0.22 #663
  • Prescribed ocean/land heat fluxes modular #667
  • NetCDF output for tracers defined, precipitation rate output at initialize! again #657
  • Precipitation rate from large-scale condensation/convection for coupling #669
  • Fixed a bug, so that sum now works for LowerTraingularArrays #668
  • Require only AbstractVectors in the interpolation #665
  • Intitialize! simulation restructured #666
  • First time steps in main time loop #659
  • Generalise OneBandRadiation to NBandRadiation #633
  • New output variables definition simplified #653
  • Bug: get_vertices for full grids had southpole at 90˚N #654
  • Group initialize!, timestep!, finalize! of main time loop #652
  • Coordinates defined in degrees lond, latd, not colat, lon #647
  • Spectral Gradients are now differentiable, with correctness check in extended CI #638
  • set! for time step #650
  • Bug: ZonalWind had sqrt of negative for uncommon resolutions #649
  • Tracer advection #579
  • Buildkite CI with dummy pipeline #646
  • ConvectiveHeating implemented #639
  • Number format flexibility with set! #634
  • Forcing/drag for primitive models #635
  • RingGrids indexing with leading Colon should now always return another RingGrid instance #637
  • Roll back GPUArrays upgrade to ensure CUDA compatibility #636
  • Change default timestep to 40min at T31 #623

v0.13

  • AbstractSurfacePerturbation introduced #631
  • AbstractStochasticPhysics and SPPT implemented #629
  • Introduced seperate extended tests that are not run on every commit #628
  • One-band longwave radiation #624
  • compat entry for FiniteDifferences.jl #620
  • Slab ocean and net surface fluxes #613
  • added compat entries for FiniteDifferences.jl and Enzyme.jl #620 #622
  • update to GPUArrays v11, JLArrays v0.2 and remove CUDA from tests #590
  • Power spectrum for n-dim LowerTriangularArrays #618
  • Added custom EnzymeRules for the SpectralTransform and an extension for compatibility with FiniteDifferences.jl #589 #625 #627
  • Optical depth introduced and array-agnostic ColumnVariables #606
  • Include large-scale condensation tests #615
  • bugfix: large-scale condensation also at <100% #609

v0.12.1

  • ConstantLandTemperature implemented #612
  • set! for more boundary conditions #611
  • SpectralFilter for horizontal diffusion #601
  • GeoMakie weak dependency, globe function for 3D data visualisation #600
  • Zonal mean for AbstractGridArray #603
  • Rossby-Haurwitz wave with initial conditions for interface displacement for shallow water models#604

v0.12.0

  • OctaminimalGaussianArray/Grid to start with 4 points around the poles #595
  • Output both accumulated and precipitation rate as netCDF #596
  • Random processes for random pattern generation #592
  • Also allow SpectralGrid as positional argument to model constructors #593
  • De-interweave SpectralTransform #587
  • Rossby-Haurwitz wave initial conditions #591
  • Haversine formula and AbstractSphericalDistance #588
  • Array-agnostic SpectralTransform #583
  • Move CUDA dependency into extension #586
  • Stop supporting Julia v1.9 #585
  • feedback.verbose (de/activate the progressbar) is now set to isinteractive() to disable automatically for documentation #582

v0.11.0

  • Extend set! with orography keyword argument to set the orography with set! #578
  • Added a new benchmark suite for the dynamics functions #577
  • Introduced a new set! function that allows to set PrognosticVariables to new values with keyword arguments #563
  • Restructured dynamical core with prognostic/diagnostic variables array-agnostic and 3-dimensional #525
  • Modularised NetCDF output #573
  • Fixed a bug in RingGrids, now broadcasts are defined even when the dimensions mismatch in some cases #568
  • RingGrids: To wrap an Array with the horizontal dimension in matrix shape into a full grid, one has to use e.g. FullGaussianGrid(map, input_as=Matrix) now. #572
  • CompatHelper: Allow for JLD2.jl v0.5