icon4py.model.atmosphere.dycore package

Subpackages

Submodules

icon4py.model.atmosphere.dycore.dycore_states module

class DivergenceDampingOrder(value, names=<not given>, *values, module=None, qualname=None, type=None, start=1, boundary=None)

Bases: int32, Enum

COMBINED = np.int32(24)

combined 2nd and 4th orders divergence damping and enhanced vertical wind off - centering during initial spinup phase

FOURTH_ORDER = np.int32(4)

4th order divergence damping

SECOND_ORDER = np.int32(2)

2nd order divergence damping

class DivergenceDampingType(value, names=<not given>, *values, module=None, qualname=None, type=None, start=1, boundary=None)

Bases: IntEnum

COMBINED = 32

combination of 3D div.damping in the troposphere with transition to 2D div. damping in the stratosphere

THREE_DIMENSIONAL = 3

divergence damping acting on 3D divergence

TWO_DIMENSIONAL = 2

divergence damping acting on 2D divergence

class HorizontalPressureDiscretizationType(value, names=<not given>, *values, module=None, qualname=None, type=None, start=1, boundary=None)

Bases: int32, Enum

Parameter called igradp_method in ICON namelist.

CONVENTIONAL = np.int32(1)

conventional discretization with metric correction term

POLYNOMIAL = np.int32(4)

Cubic / quadratic polynomial interpolation for pressure reconstruction

POLYNOMIAL_HYDRO = np.int32(5)

Same as igradp_method_polynomial, but hydrostatic approximation for downward extrapolation over steep slopes

TAYLOR = np.int32(2)

Taylor-expansion-based reconstruction of pressure

TAYLOR_HYDRO = np.int32(3)

Similar discretization as igradp_method_taylor, but uses hydrostatic approximation for downward extrapolation over steep slopes

class InterpolationState(e_bln_c_s: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='C2E', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], rbf_coeff_1: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Vertex', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='V2E', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], rbf_coeff_2: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Vertex', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='V2E', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], geofac_div: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='C2E', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], geofac_n2s: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='C2E2CO', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], geofac_grg_x: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='C2E2CO', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], geofac_grg_y: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='C2E2CO', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], nudgecoeff_e: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>)], ~numpy.float64], c_lin_e: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='E2C', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], geofac_grdiv: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='E2C2EO', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], rbf_vec_coeff_e: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='E2C2E', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], c_intp: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Vertex', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='V2C', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], geofac_rot: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Vertex', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='V2E', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], pos_on_tplane_e_1: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='E2C', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], pos_on_tplane_e_2: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='E2C', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64], e_flx_avg: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='E2C2EO', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64])

Bases: object

Represents the ICON interpolation state used in the dynamical core (SolveNonhydro).

c_intp: float64]
c_lin_e: float64]
e_bln_c_s: float64]
e_flx_avg: float64]
geofac_div: float64]
geofac_grdiv: float64]
geofac_grg_x: float64]
geofac_grg_y: float64]
geofac_n2s: float64]
geofac_rot: float64]
nudgecoeff_e: float64]
pos_on_tplane_e_1: float64]
pos_on_tplane_e_2: float64]
rbf_coeff_1: float64]
rbf_coeff_2: float64]
rbf_vec_coeff_e: float64]
class MetricStateNonHydro(mask_prog_halo_c: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], bool], rayleigh_w: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='_StaggeredK', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], wgtfac_c: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='_StaggeredK', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], wgtfacq_c: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], wgtfac_e: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='_StaggeredK', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], wgtfacq_e: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], time_extrapolation_parameter_for_exner: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], reference_exner_at_cells_on_model_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], reference_rho_at_cells_on_model_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], reference_theta_at_cells_on_model_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], reference_rho_at_edges_on_model_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], reference_theta_at_edges_on_model_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], reference_theta_at_cells_on_half_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='_StaggeredK', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], ddz_of_reference_exner_at_cells_on_half_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='_StaggeredK', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], ddqz_z_half: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='_StaggeredK', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], d2dexdz2_fac1_mc: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], d2dexdz2_fac2_mc: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], ddxn_z_full: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], ddqz_z_full_e: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], ddxt_z_full: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], inv_ddqz_z_full: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], vertoffset_gradp: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='E2C', kind=<DimensionKind.LOCAL: 'local'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.int32], zdiff_gradp: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='E2C', kind=<DimensionKind.LOCAL: 'local'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], nflat_gradp: ~numpy.int32, pg_exdist: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], exner_w_explicit_weight_parameter: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>)], ~numpy.float64], exner_w_implicit_weight_parameter: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>)], ~numpy.float64], horizontal_mask_for_3d_divdamp: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>)], ~numpy.float64], scaling_factor_for_3d_divdamp: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], coeff1_dwdz: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], coeff2_dwdz: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], coeff_gradekin: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='E2C', kind=<DimensionKind.LOCAL: 'local'>)], ~numpy.float64])

Bases: object

Dataclass containing metric fields needed in dynamical core (SolveNonhydro).

coeff1_dwdz: float64]
coeff2_dwdz: float64]
coeff_gradekin: float64]
d2dexdz2_fac1_mc: float64]
d2dexdz2_fac2_mc: float64]
ddqz_z_full_e: float64]
ddqz_z_half: float64]
ddxn_z_full: float64]
ddxt_z_full: float64]
ddz_of_reference_exner_at_cells_on_half_levels: float64]

Declared as d_exner_dz_ref_ic in ICON.

exner_w_explicit_weight_parameter: float64]

Declared as vwind_expl_wgt in ICON. The explicitness parameter for exner and w in the vertically implicit dycore solver. exner_w_explicit_weight_parameter = 1 - exner_w_implicit_weight_parameter

exner_w_implicit_weight_parameter: float64]

Declared as vwind_impl_wgt in ICON. The implicitness parameter for exner and w in the vertically implicit dycore solver. It is denoted as eta below eq. 3.20 in ICON tutorial 2023. However, it is only vwind_offctr that can be set via namelist. The actual computation of exner_w_implicit_weight_parameter is not shown in the tutorial.

horizontal_mask_for_3d_divdamp: float64]

Declared as hmask_dd3d in ICON. A horizontal mask where 3D divergence is computed for the divergence damping. 3D divergence is defined as divergence of horizontal wind plus vertical derivative of vertical wind (dw/dz).

inv_ddqz_z_full: float64]
mask_prog_halo_c: VERTICAL: 'vertical'>)], bool]
nflat_gradp: int32

The minimum height index at which the height of the center of an edge lies within two neighboring cells so that horizontal pressure gradient can be computed by first order discretization scheme.

pg_exdist: float64]

Extrapolation distance needed for HorizontalPressureDiscretizationType.TAYLOR_HYDRO.

rayleigh_w: float64]
reference_exner_at_cells_on_model_levels: float64]

Declared as exner_ref_mc in ICON.

reference_rho_at_cells_on_model_levels: float64]

Declared as rho_ref_mc in ICON.

reference_rho_at_edges_on_model_levels: float64]

Declared as rho_ref_me in ICON.

reference_theta_at_cells_on_half_levels: float64]

Declared as theta_ref_ic in ICON.

reference_theta_at_cells_on_model_levels: float64]

Declared as theta_ref_mc in ICON.

reference_theta_at_edges_on_model_levels: float64]

Declared as theta_ref_me in ICON.

scaling_factor_for_3d_divdamp: float64]

Declared as scalfac_dd3d in ICON. A scaling factor in vertical dimension for 3D divergence damping.

time_extrapolation_parameter_for_exner: float64]

Declared as exner_exfac in ICON.

vertoffset_gradp: int32]
wgtfac_c: float64]
wgtfac_e: float64]
wgtfacq_c: float64]
wgtfacq_e: float64]
zdiff_gradp: float64]
class PrepAdvection(vn_traj: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], mass_flx_me: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], dynamical_vertical_mass_flux_at_cells_on_half_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='_StaggeredK', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], dynamical_vertical_volumetric_flux_at_cells_on_half_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='_StaggeredK', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64])

Bases: object

Dataclass used in SolveNonHydro that pre-calculates fields during the dynamical substepping that are later needed in tracer advection.

dynamical_vertical_mass_flux_at_cells_on_half_levels: float64]

Declared as mass_flx_ic in ICON.

dynamical_vertical_volumetric_flux_at_cells_on_half_levels: float64]

Declared as vol_flx_ic in ICON.

mass_flx_me: float64]
vn_traj: float64]
class RhoThetaAdvectionType(value, names=<not given>, *values, module=None, qualname=None, type=None, start=1, boundary=None)

Bases: int32, Enum

Parameter called iadv_rhotheta in ICON namelist.

MIURA = np.int32(2)

2nd order Miura horizontal

SIMPLE = np.int32(1)

simple 2nd order upwind-biased scheme

class TimeSteppingScheme(value, names=<not given>, *values, module=None, qualname=None, type=None, start=1, boundary=None)

Bases: IntEnum

Parameter called itime_scheme in ICON namelist.

EXPENSIVE = 6

As STABLE, but velocity tendencies are also computed in both substeps (no benefit, but more expensive)

MOST_EFFICIENT = 4

Contravariant vertical velocity is computed in the predictor step only, velocity tendencies are computed in the corrector step only

STABLE = 5

Contravariant vertical velocity is computed in both substeps (beneficial for numerical stability in very-high resolution setups with extremely steep slopes)

initialize_prep_advection(grid: grid_base.Grid, allocator: gtx_typing.Allocator) → PrepAdvection

icon4py.model.atmosphere.dycore.dycore_utils module

icon4py.model.atmosphere.dycore.solve_nonhydro module

class IntermediateFields(horizontal_pressure_gradient: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], anyfloat], rho_at_edges_on_model_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], theta_v_at_edges_on_model_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], ~numpy.float64], horizontal_kinetic_energy_at_edges_on_model_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], anyfloat], tangential_wind_on_half_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='_StaggeredK', kind=<DimensionKind.VERTICAL: 'vertical'>)], anyfloat], horizontal_gradient_of_normal_wind_divergence: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Edge', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], anyfloat], dwdz_at_cells_on_model_levels: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>), Dimension(value='K', kind=<DimensionKind.VERTICAL: 'vertical'>)], anyfloat])

Bases: object

Encapsulate internal fields of SolveNonHydro that contain shared state over predictor and corrector step.

Encapsulates internal fields used in SolveNonHydro. Fields (and the class!) follow the naming convention of ICON to prepend local fields of a module with z_. Contrary to other such z_ fields inside SolveNonHydro the fields in this dataclass contain state that is built up over the predictor and corrector part in a timestep.

classmethod allocate(grid: Grid, allocator: ArrayNamespace | FieldBufferAllocatorProtocol[DeviceTypeT] | FieldBufferAllocatorFactoryProtocol[DeviceTypeT] | None) → IntermediateFields
dwdz_at_cells_on_model_levels: VERTICAL: 'vertical'>)], anyfloat]

Declared as z_dwdz_dd in ICON.

horizontal_gradient_of_normal_wind_divergence: VERTICAL: 'vertical'>)], anyfloat]

Declared as z_graddiv_vn in ICON.

horizontal_kinetic_energy_at_edges_on_model_levels: VERTICAL: 'vertical'>)], anyfloat]

Declared as z_kin_hor_e in ICON.

horizontal_pressure_gradient: VERTICAL: 'vertical'>)], anyfloat]

Declared as z_gradh_exner in ICON.

rho_at_edges_on_model_levels: float64]

Declared as z_rho_e in ICON.

tangential_wind_on_half_levels: VERTICAL: 'vertical'>)], anyfloat]

it is only used for the horizontal advection of w, and w[nlevp1-1] is diagnostic.

Type:

Declared as z_vt_ie in ICON. Tangential wind at edge on k-half levels. The bottom half level is never computed

theta_v_at_edges_on_model_levels: float64]

Declared as z_theta_v_e in ICON.

class NonHydrostaticConfig(*, itime_scheme: TimeSteppingScheme = TimeSteppingScheme.MOST_EFFICIENT, iadv_rhotheta: RhoThetaAdvectionType = RhoThetaAdvectionType.MIURA, igradp_method: HorizontalPressureDiscretizationType = HorizontalPressureDiscretizationType.TAYLOR_HYDRO, rayleigh_type: RayleighType = RayleighType.KLEMP, divdamp_order: DivergenceDampingOrder = DivergenceDampingOrder.COMBINED, divdamp_type: DivergenceDampingType = DivergenceDampingType.THREE_DIMENSIONAL, l_vert_nested: bool = False, deepatmos_mode: bool = False, iau_init: bool = False, extra_diffu: bool = True, rhotheta_offctr: float = -0.1, veladv_offctr: float = 0.25, fourth_order_divdamp_factor: float = 0.0025, fourth_order_divdamp_factor2: float = 0.004, fourth_order_divdamp_factor3: float = 0.004, fourth_order_divdamp_factor4: float = 0.004, fourth_order_divdamp_z: float = 32500.0, fourth_order_divdamp_z2: float = 40000.0, fourth_order_divdamp_z3: float = 60000.0, fourth_order_divdamp_z4: float = 80000.0)

Bases: object

Contains necessary parameter to configure a nonhydro run.

Encapsulates namelist parameters and derived parameters. TODO: (magdalena) values should be read from a configuration file. Default values are taken from the defaults in the corresponding ICON Fortran namelist files.

deepatmos_mode: bool = False
divdamp_order: DivergenceDampingOrder = np.int32(24)
divdamp_type: DivergenceDampingType = 3
extra_diffu: bool = True
fourth_order_divdamp_factor: float = 0.0025
fourth_order_divdamp_factor2: float = 0.004
fourth_order_divdamp_factor3: float = 0.004
fourth_order_divdamp_factor4: float = 0.004
fourth_order_divdamp_z: float = 32500.0
fourth_order_divdamp_z2: float = 40000.0
fourth_order_divdamp_z3: float = 60000.0
fourth_order_divdamp_z4: float = 80000.0
classmethod from_fortran_dict(atmo_dict: dict[str, Any], **overrides: Any) → NonHydrostaticConfig
iadv_rhotheta: RhoThetaAdvectionType = np.int32(2)
iau_init: bool = False
igradp_method: HorizontalPressureDiscretizationType = np.int32(3)
itime_scheme: TimeSteppingScheme = 4
l_vert_nested: bool = False
rayleigh_type: RayleighType = 2
rhotheta_offctr: float = -0.1
veladv_offctr: float = 0.25
class NonHydrostaticParams(config: NonHydrostaticConfig)

Bases: object

Calculates derived quantities depending on the NonHydrostaticConfig.

advection_explicit_weight_parameter: Final[float]

Declared as wgt_nnow_vel in ICON.

advection_implicit_weight_parameter: Final[float]

Declared as wgt_nnew_vel in ICON.

rhotheta_explicit_weight_parameter: Final[float]

Declared as wgt_nnow_rth in ICON.

rhotheta_implicit_weight_parameter: Final[float]

Declared as wgt_nnew_rth in ICON.

class SolveNonhydro(*, grid: ~icon4py.model.common.grid.icon.IconGrid, config: ~icon4py.model.atmosphere.dycore.solve_nonhydro.NonHydrostaticConfig, params: ~icon4py.model.atmosphere.dycore.solve_nonhydro.NonHydrostaticParams, metric_state_nonhydro: ~icon4py.model.atmosphere.dycore.dycore_states.MetricStateNonHydro, interpolation_state: ~icon4py.model.atmosphere.dycore.dycore_states.InterpolationState, vertical_params: ~icon4py.model.common.grid.vertical.VerticalGrid, edge_geometry: ~icon4py.model.common.grid.states.EdgeParams, cell_geometry: ~icon4py.model.common.grid.states.CellParams, owner_mask: ~gt4py.next.common.Field[~gt4py.next.common.Dims[Dimension(value='Cell', kind=<DimensionKind.HORIZONTAL: 'horizontal'>)], bool], backend: ~gt4py.next.backend.Backend | ~gt4py._core.definitions.DeviceType | BackendDescriptor | None, exchange: ~icon4py.model.common.decomposition.definitions.ExchangeRuntime, max_nudging_coefficient: float)

Bases: object

run_corrector_step(*, diagnostic_state_nh: DiagnosticStateNonHydro, prognostic_states: TimeStepPair[PrognosticState], z_fields: IntermediateFields, second_order_divdamp_factor: float, prep_adv: PrepAdvection, dtime: float, ndyn_substeps_var: int, prepare_fluxes_for_advection: bool, at_first_substep: bool, at_last_substep: bool, is_iau_active: bool, iau_wgt_dyn: float) → None
run_predictor_step(*, diagnostic_state_nh: DiagnosticStateNonHydro, prognostic_states: TimeStepPair[PrognosticState], z_fields: IntermediateFields, dtime: float, at_initial_timestep: bool, at_first_substep: bool, is_iau_active: bool, iau_wgt_dyn: float) → None

Runs the predictor step of the non-hydrostatic solver.

time_step(*, diagnostic_state_nh: DiagnosticStateNonHydro, prognostic_states: TimeStepPair[PrognosticState], prep_adv: PrepAdvection, second_order_divdamp_factor: float, dtime: float, ndyn_substeps_var: int, at_initial_timestep: bool, prepare_fluxes_for_advection: bool, at_first_substep: bool, at_last_substep: bool, is_iau_active: bool = False, iau_wgt_dyn: float = 0.0) → None

Update prognostic variables (prognostic_states.next) after the dynamical process over one substep. :param diagnostic_state_nh: diagnostic variables used for solving the governing equations. It includes local variables and the physics tendency term that comes from physics :param prognostic_states: prognostic variables :param prep_adv: variables for tracer advection :param second_order_divdamp_factor: Originally declared as divdamp_fac_o2 in ICON. Second order (nabla2) divergence damping coefficient. :param dtime: time step :param ndyn_substeps_var: number of dynamical substeps :param at_initial_timestep: initial time step of the model run :param prepare_fluxes_for_advection: Preparation for tracer advection :param at_first_substep: first substep :param at_last_substep: last substep :param is_iau_active: Incremental analysis update active during dycore step :param iau_wgt_dyn: weight scalar for the incremental analysis update

icon4py.model.atmosphere.dycore.solve_nonhydro_stencils module

Module contents