Source code for virtual_ecosystem.core.model_config

"""The `core.model_config` module contains pydantic models defining the configuration
settings required for the Core model in a Virtual Ecosystem simulation.
"""  # noqa: D205

from __future__ import annotations

from datetime import date
from functools import cached_property
from pathlib import Path
from typing import ClassVar

import numpy as np
from pint import DimensionalityError, Quantity, UndefinedUnitError
from pydantic import (
    ConfigDict,
    Field,
    NegativeFloat,
    PositiveFloat,
    PositiveInt,
    field_validator,
    model_validator,
)
from pyrealm.constants import CoreConst, PModelConst
from scipy import constants

from virtual_ecosystem.core.configuration import (
    DIRPATH_PLACEHOLDER,
    FILEPATH_PLACEHOLDER,
    Configuration,
)


[docs] class CoreConstants(Configuration): """Core constants for use across the Virtual Ecosystem modules. An instance of the CoreConstants dataclass provides definitions of the core constants used across an entire simulation. The core constants can be changed, as shown below, although for many this would likely generate nonsensical results. Example: >>> consts = CoreConstants() >>> consts.max_depth_of_microbial_activity 0.25 >>> consts = CoreConstants(max_depth_of_microbial_activity=0.75) >>> consts.max_depth_of_microbial_activity 0.75 """ zero_Celsius: ClassVar[float] = constants.zero_Celsius """Conversion constant from Kelvin to Celsius (°).""" standard_pressure: float = constants.atmosphere / 1000 """Standard atmospheric pressure, [kPa]""" standard_mole: float = 44.642 """Moles of ideal gas in 1 m^3 air at standard atmosphere.""" molar_heat_capacity_air: float = 29.19 """Molar heat capacity of air, [J mol-1 K-1].""" gravity: float = constants.gravitational_constant """Newtonian constant of gravitation, [m s-1].""" boltzmann_constant: float = constants.Boltzmann """The Boltzmann constant, [J K-1]""" stefan_boltzmann_constant: float = constants.Stefan_Boltzmann """Stefan-Boltzmann constant, [W m-2 K-4]. The Stefan-Boltzmann constant relates the energy radiated by a black body to its temperature.""" von_karmans_constant: float = 0.4 """Von Karman's constant, [unitless]. The von Karman's constant describes the logarithmic velocity profile of a turbulent fluid near a no-slip boundary.""" max_depth_of_microbial_activity: float = 0.25 """Maximum depth of microbial activity in the soil layers [m]. The soil model needs to identify which of the configured soil layers are sufficiently close to the surface to contain significant microbial activity that drives nutrient processes. The default value is taken from :cite:t:`fatichi_mechanistic_2019`. No empirical source is provided for this value. """ meters_to_mm: float = 1000.0 """Factor to convert variable unit from meters to millimeters.""" mm_to_kg: float = 1e-3 """Factor to convert variable unit from millimeters to kilograms of water per square metre.""" molecular_weight_air: float = 28.96 """Molecular weight of air, [g mol-1].""" gas_constant_water_vapour: float = 461.51 """Gas constant for water vapour, [J kg-1 K-1]""" seconds_to_day: float = 86400.0 """Factor to convert variable unit from seconds to day.""" seconds_to_hour: float = 3600.0 """Factor to convert variable unit from seconds to hours.""" hours_per_day: int = 24 """Number of hours per day.""" characteristic_dimension_leaf: float = 0.01 """Characteristic dimension of leaf, typically around 0.7 * leaf width, [m].""" specific_gas_constant_dry_air: float = 287.05 """Specific gas constant for dry air, [J kg-1 K-1].""" molecular_weight_ratio_water_to_dry_air: float = 0.622 """The molecular weight ratio of water to dry air. The ratio of the molar mass of water vapour (18.015 g/mol) to the molar mass of dry air (28.964 g/mol), which is approximately 0.622. This ratio is used in atmospheric calculations, particularly in determining the mixing ratio of water vapour to dry air.""" conductance_to_resistance_conversion_factor: float = 40.9 """Conductance to resistance conversion factor. This factor is used to convert between stomatal conductance in mmol m-2 s-1 and stomatal resistance in s m-1.""" density_water: float = 1000.0 """Density of water, [kg m-3].""" fungal_fruiting_bodies_c_n_ratio: float = 10.0 """Carbon to nitrogen ratio of fungal fruiting bodies, [unitless]. This constant is stored in the CoreConsts as it is used by both the animal model (to work out consumption flows) and the soil model (to work out production rates). The current default value is very much a guess. """ fungal_fruiting_bodies_c_p_ratio: float = 75.0 """Carbon to phosphorus ratio of fungal fruiting bodies, [unitless]. This constant is stored in the CoreConsts as it is used by both the animal model (to work out consumption flows) and the soil model (to work out production rates). The current default value is very much a guess. """ fungal_fruiting_bodies_decay_rate: float = np.log(2) / 50.0 """Rate constant for the decay of fungal fruiting bodies, [day^-1]. This is calculated based on the assumption that fungal fruiting bodies decay with a half-life of 50 days. This estimate should be improved based on empirical data. """ air_volumetric_heat_capacity: float = 1200.0 """Volumetric heat capacity of air at constant pressure, [J m-3 K-1]. This represents the amount of heat energy required to raise the temperature of one cubic meter of air by 1 Kelvin.""" initial_aerodynamic_resistance_canopy: float = 12.1 """Initial aerodynamic resistance of the canopy, [s m-1]. This parameter is an initial estimate of the resistance to the transfer of momentum, heat, and water vapour between the leaf surface and the atmosphere. The value is based on Australian evergreen forest, taken from :cite:t:`su_aerodynamic_2021`; note that this assumes a dense canopy. """
[docs] class GridConfiguration(Configuration): """Grid configuration. This configuration model sets the size and shape of grid cells within the simulation and then the number of cells in the X and Y directions and their locations in space. """ grid_type: str = "square" """The grid cell type. The value must be one of the options supported by the :data:`~virtual_ecosystem.core.grid.GRID_REGISTRY`.""" cell_area: PositiveFloat = Field(default=8100.0) """The area of each grid cell (m^2)""" cell_nx: PositiveInt = Field(default=9) """Number of grid cells in x direction""" cell_ny: PositiveInt = Field(default=9) """Number of grid cells in y direction""" xoff: float = -45.0 """The x offset of the grid origin""" yoff: float = -45.0 """The x offset of the grid origin"""
[docs] class TimingConfiguration(Configuration): """Configuration of the model timing. This configuration section sets the model start data, update length and run time. The update length and run time are provided as a text string that will be automatically parsed to give a total time in seconds. """ start_date: date = date(2013, 1, 1) """The simulation start date.""" update_interval: str = "1 month" """The interval at which all models are updated.""" run_length: str = "2 years" """The total run length of the simulation."""
[docs] @cached_property def update_interval_seconds(self) -> float: """Interval update length in seconds.""" return Quantity(self.update_interval).to("seconds").magnitude
[docs] @cached_property def run_length_seconds(self) -> float: """Run length in seconds.""" return Quantity(self.run_length).to("seconds").magnitude
@field_validator("update_interval", "run_length") def _validate_pint_time_quantities(cls, value) -> str: """Validates time strings can be parsed as quantities.""" try: _ = Quantity(value).to("seconds") except (DimensionalityError, UndefinedUnitError): raise ValueError(f"Cannot parse value as time quantity: {value}") return value @model_validator(mode="after") def _run_length_too_short(self) -> TimingConfiguration: """Model validation that there is enough time for at least one update.""" if self.run_length_seconds < self.update_interval_seconds: raise ValueError( f"Model run length ({self.run_length}) expires before " f"first update ({self.update_interval})" ) return self
[docs] class DataOutputConfiguration(Configuration): """Output settings for the Virtual Ecosystem model state. TODO - this is very confusingly named and structure - restructure and add class validation. """ save_initial_state: bool = False "Whether the initial state should be saved" save_continuous_data: bool = True "Whether continuous data should be saved" save_final_state: bool = True "Whether the final state should be saved" save_merged_config: bool = True "Whether to save a merged TOML file containing all config options" out_path: DIRPATH_PLACEHOLDER = Path("<DIRPATH_PLACEHOLDER>") "Directory path for output files" out_initial_file_name: str = "initial_state.nc" """File name for initial state output file""" out_folder_continuous: str = "." "Folder to save states of simulation with time to" out_continuous_file_name: str = "all_continuous_data.nc" """Name of file to save combined continuous data to""" out_final_file_name: str = "final_state.nc" """File name for final state output file""" out_merge_file_name: str = "ve_full_model_configuration.toml" """Name for TOML file containing merged configs"""
[docs] class LayersConfiguration(Configuration): """Settings for the simulation vertical structure.""" soil_layers: list[NegativeFloat] = Field(min_length=1, default=[-0.25, -1.0]) """A list of negative float values that provides the depth in metres of the soil horizons to be used in the simulation, hence also setting the number of soil layers and the horizon depth for each layer relative to the surface. The values must be unique and strictly decreasing. """ canopy_layers: PositiveInt = 10 """The maximum number of canopy layers to simulate. This is used to control the number of layers with the ``canopy`` role. Not all of these layers necessarily contain canopy during a simulation as the canopy structure within these layers is dynamic.""" above_canopy_height_offset: PositiveFloat = Field(default=2.0, allow_inf_nan=False) """A height offset relative to the canopy top that is used as the measurement height of reference climate data. It sets the the height above the canopy top of the first layer role ``above`` (metres).""" subcanopy_layer_height: PositiveFloat = Field(default=1.5, allow_inf_nan=False) """The height above ground level of the ground surface atmospheric layer, used to calculate subcanopy microclimate conditions (metres).""" surface_layer_height: PositiveFloat = Field(default=0.1, allow_inf_nan=False) """The height above ground level of the ground surface atmospheric layer (metres).""" @field_validator("soil_layers") def _soil_depths_unique_decreasing(cls, values) -> list[float]: """Check the soil depths are unique and decreasing. This runs post validation, so the inputs are a list of negative floats. """ if len(values) != len(set(values)): raise ValueError("Repeated values in soil layer depths.") strictly_decreasing = [-m for m in sorted([abs(n) for n in values])] if not values == strictly_decreasing: raise ValueError("Soil layer depths must be strictly decreasing") return values @model_validator(mode="after") def _surface_below_subcanopy(self) -> LayersConfiguration: """Check the surface height is below the subcanopy.""" if self.surface_layer_height >= self.subcanopy_layer_height: raise ValueError( f"Surface layer height ({self.surface_layer_height}) must be " f"below subcanopy layer height ({self.subcanopy_layer_height})" ) return self
[docs] class DataSource(Configuration): """Data source configuration.""" file_path: FILEPATH_PLACEHOLDER = Path("<FILEPATH_PLACEHOLDER>") var_name: str = "variable_name_placeholder"
[docs] class DataConfiguration(Configuration): """Data configuration.""" variable: tuple[DataSource, ...] = ( DataSource(var_name="variable_name_placeholder_one"), DataSource(var_name="variable_name_placeholder_two"), )
[docs] class PyrealmConfig(Configuration): """Configuration class for pyrealm constant dataclasses. These dataclasses are not pydantic models and so we permit arbitrary types. """ model_config = ConfigDict(arbitrary_types_allowed=True) core: CoreConst = CoreConst() """Core pyrealm constants""" pmodel: PModelConst = PModelConst() """Pyrealm constants for the PModel."""
[docs] class DebugConfiguration(Configuration): """Debugging options.""" truncate_run_at_update: int = -1 """This option can be used to exit a simulation at a particular update index. This can be useful for testing settings on a model and dataset without having to create a specific temporal subset. If the value is less than 1, then no truncation occurs."""
[docs] class CoreConfiguration(Configuration): """The core model configuration.""" constants: CoreConstants = CoreConstants() """Constants for the core module""" grid: GridConfiguration = GridConfiguration() """Configuration of the spatial grid""" data_output_options: DataOutputConfiguration = DataOutputConfiguration() """Configuration of the output of the Virtual Ecosystem model state""" layers: LayersConfiguration = LayersConfiguration() """Configuration of the layers in the vertical structure""" timing: TimingConfiguration = TimingConfiguration() """Configuration of the model run and step lengths""" data: DataConfiguration = DataConfiguration() """Configuration of the input variables and data sources.""" debug: DebugConfiguration = DebugConfiguration() """Configuration of debugging options.""" pyrealm: PyrealmConfig = Field(default=PyrealmConfig(), exclude=True) """Constant dataclasses for the pyrealm package. At present, the pyrealm configuration settings are excluded from model serialisation because of issues with serialising numpy arrays. This is a problem for replicating simulations where these settings have been altered. """