Model Structure Configuration

The model_structure.json file defines the building blocks of an ecosystem model for SINDBAD experiments. It consists of two main sections: model selection and pool configuration.

Model Selection

This section defines the ecosystem processes to be included in the experiment. Only listed processes are activated during execution.

Default Model Settings

The default_model section specifies default properties for all models, which can be overridden in individual model configurations.

"default_model": {
    "implicit_t_repeat": "Number of times a model runs within a single time step",
    "use_in_spinup": "Flag indicating if the model is used during spinup"
}

Model Configuration

The models section lists selected models with their corresponding approaches. In SINDBAD:

• A model represents an ecosystem process

• An approach represents the implementation method

"models": {
    "process_name": {
        "approach": "Implementation method for the process",
        "use_in_spinup": "Override default spinup behavior"
    }
}

Model Dependencies

• Use standard_sindbad_models to view the complete list of available models

• Ensure model combinations are feasible (some processes depend on others)

• Example: Snow processes require snowfall in the model structure

Pools and Storages

This section configures model components that contribute to mass balance calculations.

Pool Configuration

Each pool is defined under the pools section with its components and state variables.

"pools": {
    "pool_type": {
        "combine": "Variable name for combined pool values",
        "components": {
            "component_name": [
                "Layer configuration (number of layers or depth list)",
                "Initial storage value"
            ]
        },
        "state_variables": {
            "variable_name": "Initial value"
        }
    }
}

Pool Structure

• Each pool type (e.g., carbon, water) has its own configuration

• Components define individual storage elements

• State variables track additional pool-related metrics

Configuration Guidelines

• Ensure pool configurations match model requirements

• Verify initial values are within reasonable ranges

• Check layer configurations match forcing data dimensions