A baseline model was designed on CliMT. The fork of CliMT used can be found at https://github.com/Ai33L/climt/tree/heatwave_sampling
Dynamical core
Parameterisations -
- Gray radiation with Frierson longwave optical depths (default surface values, but linear with pressure)
- Slab surface - relative humidity scaling of 0.7 over land
- Moist convection (Emanuel scheme)
- Diffusive planetary boundary layer
Model resolution is T42 (64 x 128), with 28 vertical levels. The timestep of integration used is 20 minutes.
The incoming solar flux is set to be highest at 10$\degree$N, with the profile with latitude
In this model, we have implemented an idealised land configuration, with land extending from 20° to 60° in both hemispheres.
The depth of the land slab is set to 1 m, while the depth of the ocean slab is set to 2 m.
The model data is stored using the Zarr storage format. The daily averaged model data is saved in the xarray dataset format.
Common data -
- longitude
- latitude
- area_type
These are saved before the model begins in the common directory.
Model data -
- air_pressure
- air_temperature
- boundary_layer_height
- specific_humidity
- northward_wind
- eastward_wind
- surface_air_pressure
- surface_temperature
- surface_upward_latent_heat_flux
- surface_upward_sensible_heat_flux
- boundary_layer_height
- divergence_of_wind
- surface_upwelling_longwave_flux_in_air
- surface_downwelling_longwave_flux_in_air
The daily averages of these are saved along the time axis in nested directories in xarray dataset format.
The model is spun up for 3 years. The model is then run for a period of 30 years, with model states saved every 10 days.
The 30 year climatology is computed from this run. The model states saved will be used for the GKLT algorithm later.