"""
Vertical cloud profile kernels.
This module contains functions that compute the vertical distribution of cloud
mass mixing ratio (q_c_lay) as a function of pressure and atmospheric conditions.
"""
from typing import Dict
import jax.numpy as jnp
from .data_constants import bar
__all__ = [
"no_cloud",
"exponential_decay_profile",
"slab_profile",
"const_profile",
]
[docs]
def no_cloud(
p_lay: jnp.ndarray,
T_lay: jnp.ndarray,
mu_lay: jnp.ndarray,
rho_lay: jnp.ndarray,
nd_lay: jnp.ndarray,
params: Dict[str, jnp.ndarray],
) -> jnp.ndarray:
"""Return zero cloud mass mixing ratio (no clouds).
Parameters
----------
p_lay : `~jax.numpy.ndarray`, shape (nlay,)
Pressure at layer centers in dyne cm⁻².
T_lay : `~jax.numpy.ndarray`, shape (nlay,)
Layer temperatures in K.
mu_lay : `~jax.numpy.ndarray`, shape (nlay,)
Mean molecular weight per layer in amu.
rho_lay : `~jax.numpy.ndarray`, shape (nlay,)
Mass density per layer in g cm⁻³.
nd_lay : `~jax.numpy.ndarray`, shape (nlay,)
Number density per layer in cm⁻³.
params : dict[str, `~jax.numpy.ndarray`]
Parameter dictionary (unused).
Returns
-------
q_c_lay : `~jax.numpy.ndarray`, shape (nlay,)
Cloud mass mixing ratio (all zeros).
"""
nlay = T_lay.shape[0]
return jnp.zeros(nlay)
[docs]
def exponential_decay_profile(
p_lay: jnp.ndarray,
T_lay: jnp.ndarray,
mu_lay: jnp.ndarray,
rho_lay: jnp.ndarray,
nd_lay: jnp.ndarray,
params: Dict[str, jnp.ndarray],
) -> jnp.ndarray:
"""Exponential decay cloud profile with hard base cutoff.
This profile follows:
q_c(P) = q_c_0 * (P / P_base)^alpha for P < P_base
q_c(P) = 0 for P >= P_base
Parameters
----------
p_lay : `~jax.numpy.ndarray`, shape (nlay,)
Pressure at layer centers in dyne cm⁻².
T_lay : `~jax.numpy.ndarray`, shape (nlay,)
Layer temperatures in K.
mu_lay : `~jax.numpy.ndarray`, shape (nlay,)
Mean molecular weight per layer in amu.
rho_lay : `~jax.numpy.ndarray`, shape (nlay,)
Mass density per layer in g cm⁻³.
nd_lay : `~jax.numpy.ndarray`, shape (nlay,)
Number density per layer in cm⁻³.
params : dict[str, `~jax.numpy.ndarray`]
Parameter dictionary containing:
- `log_10_q_c` : float
Log₁₀ cloud mass mixing ratio at the base pressure.
- `log_10_alpha_cld` : float
Log₁₀ cloud pressure power-law exponent.
- `log_10_p_base` : float
Log₁₀ base pressure in bar (converted to dyne cm⁻² internally).
Returns
-------
q_c_lay : `~jax.numpy.ndarray`, shape (nlay,)
Cloud mass mixing ratio per layer.
"""
# Retrieved parameters
q_c_0 = 10.0 ** params["log_10_q_c"]
alpha = 10.0 ** params["log_10_alpha_cld"]
p_base = 10.0 ** params["log_10_p_base"] * bar # bar → dyne cm⁻²
# Hard cutoff: clouds only exist for P < P_base
cloud_mask = p_lay < p_base
# Compute exponential profile
q_c_profile = q_c_0 * (p_lay / jnp.maximum(p_base, 1e-30)) ** alpha
# Apply hard cutoff
q_c_lay = jnp.where(cloud_mask, q_c_profile, 0.0)
return q_c_lay
[docs]
def slab_profile(
p_lay: jnp.ndarray,
T_lay: jnp.ndarray,
mu_lay: jnp.ndarray,
rho_lay: jnp.ndarray,
nd_lay: jnp.ndarray,
params: Dict[str, jnp.ndarray],
) -> jnp.ndarray:
"""Uniform slab cloud profile with hard pressure cutoffs.
The cloud is present with constant q_c between P_top and P_bot, and zero outside.
Parameters
----------
p_lay : `~jax.numpy.ndarray`, shape (nlay,)
Pressure at layer centers in dyne cm⁻².
T_lay : `~jax.numpy.ndarray`, shape (nlay,)
Layer temperatures in K.
mu_lay : `~jax.numpy.ndarray`, shape (nlay,)
Mean molecular weight per layer in amu.
rho_lay : `~jax.numpy.ndarray`, shape (nlay,)
Mass density per layer in g cm⁻³.
nd_lay : `~jax.numpy.ndarray`, shape (nlay,)
Number density per layer in cm⁻³.
params : dict[str, `~jax.numpy.ndarray`]
Parameter dictionary containing:
- `log_10_q_c` : float
Log₁₀ cloud mass mixing ratio inside the slab.
- `log_10_p_top_slab` : float
Log₁₀ pressure at the top of the slab in bar.
- `log_10_dp_slab` : float
Log₁₀ linear pressure width of the slab in bar (Δp = 10^log_10_dp_slab).
Returns
-------
q_c_lay : `~jax.numpy.ndarray`, shape (nlay,)
Cloud mass mixing ratio per layer (q_c inside slab, 0 outside).
"""
# Retrieved parameters
q_c_slab = 10.0 ** params["log_10_q_c"]
# Slab boundaries in pressure (bars → dyne cm⁻²)
P_top = 10.0 ** params["log_10_p_top_slab"] * bar # bar → dyne cm⁻²
Delta_p = 10.0 ** params["log_10_dp_slab"] * bar # bar → dyne cm⁻²
P_bot = P_top + Delta_p # P_c,top + Δp
# Hard slab cutoff: 1 inside [P_top, P_bot], 0 outside
slab_mask = jnp.logical_and(p_lay >= P_top, p_lay <= P_bot)
q_c_lay = q_c_slab * slab_mask
return q_c_lay
[docs]
def const_profile(
p_lay: jnp.ndarray,
T_lay: jnp.ndarray,
mu_lay: jnp.ndarray,
rho_lay: jnp.ndarray,
nd_lay: jnp.ndarray,
params: Dict[str, jnp.ndarray],
) -> jnp.ndarray:
"""Constant cloud mass mixing ratio throughout the entire atmosphere.
Parameters
----------
p_lay : `~jax.numpy.ndarray`, shape (nlay,)
Pressure at layer centers in dyne cm⁻².
T_lay : `~jax.numpy.ndarray`, shape (nlay,)
Layer temperatures in K.
mu_lay : `~jax.numpy.ndarray`, shape (nlay,)
Mean molecular weight per layer in amu.
rho_lay : `~jax.numpy.ndarray`, shape (nlay,)
Mass density per layer in g cm⁻³.
nd_lay : `~jax.numpy.ndarray`, shape (nlay,)
Number density per layer in cm⁻³.
params : dict[str, `~jax.numpy.ndarray`]
Parameter dictionary containing:
- `log_10_q_c` : float
Log₁₀ cloud mass mixing ratio (constant value throughout atmosphere).
Returns
-------
q_c_lay : `~jax.numpy.ndarray`, shape (nlay,)
Cloud mass mixing ratio per layer (constant value everywhere).
"""
# Retrieved parameter
q_c = 10.0 ** params["log_10_q_c"]
# Return constant profile
nlay = T_lay.shape[0]
return jnp.full(nlay, q_c)
