skyscapes.physical_model.exojax.presets

skyscapes.physical_model.exojax.presets#

Atmospheric-composition archetypes (VMR / MMR conversion helpers).

Each preset is a dict mapping molecule name to volume mixing ratio (VMR = mole fraction, the convention used in atmospheric science literature). vmr_dict_to_log_mmr_dict() converts to log10 mass-mixing ratios at construction time using a bulk gas (default N2) to fill the residual, returning the dict shape ExoJaxPhysicalModel.from_default_setup consumes for log_mmrs.

The four Earth-epoch presets bracket the modern oxygenated atmosphere plus three earlier states (Archean, Early Proterozoic, Late Proterozoic) drawn from a published “Earth through time” biosignature table. Source attribution lives with whoever updates these values – update the table here and note the source in your commit message or local documentation.

I got these from Natasha Latouf and she has citations… somewhere.

Example:

import jax.numpy as jnp
from skyscapes.physical_model import ExoJaxPhysicalModel
from skyscapes.physical_model.exojax.presets import (
    EARTH_MODERN_VMRS, vmr_dict_to_log_mmr_dict,
)

log_mmrs = vmr_dict_to_log_mmr_dict(EARTH_MODERN_VMRS, K=1)
model = ExoJaxPhysicalModel.from_default_setup(
    log_mmrs=log_mmrs,
    T_eq_K=jnp.full((1,), 288.0),
    T_alpha=jnp.full((1,), 0.07),
    log_surface_albedo=jnp.full((1,), jnp.log10(0.3)),
    log_gravity_cgs=jnp.full((1,), jnp.log10(981.0)),
)

Attributes#

EARTH_ARCHEAN_VMRS

EARTH_EARLY_PROTEROZOIC_VMRS

EARTH_LATE_PROTEROZOIC_VMRS

EARTH_MODERN_VMRS

Functions#

vmr_dict_to_mmr_dict(vmrs[, bulk_gas])

Convert volume-mixing-ratio dict to mass-mixing-ratio dict.

vmr_dict_to_log_mmr_dict(vmrs[, K, bulk_gas])

Convert VMR dict to log10 MMR dict with K-shaped arrays.

vmr_dict_to_earth_profile_dict(vmrs[, K, bulk_gas])

VMRs -> dict of Earth-realistic profiles per species.

Module Contents#

skyscapes.physical_model.exojax.presets.EARTH_ARCHEAN_VMRS: dict[str, float]#
skyscapes.physical_model.exojax.presets.EARTH_EARLY_PROTEROZOIC_VMRS: dict[str, float]#
skyscapes.physical_model.exojax.presets.EARTH_LATE_PROTEROZOIC_VMRS: dict[str, float]#
skyscapes.physical_model.exojax.presets.EARTH_MODERN_VMRS: dict[str, float]#
skyscapes.physical_model.exojax.presets.vmr_dict_to_mmr_dict(vmrs, bulk_gas='N2')[source]#

Convert volume-mixing-ratio dict to mass-mixing-ratio dict.

Computes mean molecular weight from the supplied VMRs plus the bulk-gas residual (mmr_bulk = 1 - sum(vmrs_tracked), in VMR space; then convert all to mass). Used internally by vmr_dict_to_log_mmr_dict(); exposed here for inspection.

Args:
vmrs: {name: VMR} for tracked molecules. Sum must be < 1

(the residual is the bulk gas).

bulk_gas: Name of the implicit residual gas ("N2" for

terrestrial atmospheres, "H2" for gas giants, …). Must appear in BULK_GAS_RECIPES.

Returns:

{name: MMR} for the tracked molecules. The bulk gas’s MMR is not returned – ExoJaxPhysicalModel computes it dynamically as the residual.

Parameters:
Return type:

dict[str, float]

skyscapes.physical_model.exojax.presets.vmr_dict_to_log_mmr_dict(vmrs, K=1, bulk_gas='N2')[source]#

Convert VMR dict to log10 MMR dict with K-shaped arrays.

Each value in the returned dict is shape (K,) so it can be passed directly as log_mmrs= to ExoJaxPhysicalModel.from_default_setup(). The same VMR is used for every planet – to give per-planet variation, modify the returned arrays before constructing the atmosphere.

Every species gets a ConstantMmr profile by default (uniform mmr at every altitude). For Earth-realistic profiles (O3 stratospheric peak, H2O cold-trap drop) use vmr_dict_to_earth_profile_dict() instead.

Parameters:
Return type:

dict[str, jaxtyping.Array]

skyscapes.physical_model.exojax.presets.vmr_dict_to_earth_profile_dict(vmrs, K=1, bulk_gas='N2')[source]#

VMRs -> dict of Earth-realistic profiles per species.

Most species get a ConstantMmr (well-mixed assumption). For H2O and O3, where Earth’s altitude profile differs sharply from uniform, the function builds:

  • H2O: TroposphericMmr with the supplied VMR below the cold trap and a 3-decade drop above (parametrized at P = 0.1 bar).

  • O3: StratosphericPeakMmr with the supplied VMR as the peak value at P = 10 mbar (Gaussian width 0.5 dex).

All values are taken from canonical Earth-atmosphere structure – customize by constructing your own profile dict if you need to fit per-planet profile parameters in an HMC retrieval.

Args:
vmrs: {name: VMR}; H2O VMR is interpreted as the

tropospheric value, O3 VMR as the stratospheric peak.

K: Number of planets (per-planet broadcasting). bulk_gas: Implicit residual gas name.

Returns:

{name: AbstractMmrProfile} ready to pass as log_mmrs= to ExoJaxPhysicalModel.from_default_setup().

Parameters:
Return type:

dict[str, skyscapes.physical_model.exojax.components.AbstractMmrProfile]