Control flow

Model Creation

A YAML format configuration file defines both the model structure (spatial domains containing biogeochemical variables and reactions that operate on them) and model parameter values. Reactions (subtypes of AbstractReaction) are identified by Type name (without module prefix) in the configuration file.

To create the model, the numerical solver or host should call create_model_from_config which reads the configuration file, and then:

• creates model Domains and Reactions (subtypes of AbstractReaction), applying any Parameters settings from the parameters: sections in the config file.
• calls set_model_geometry allowing Reactions to create and attach Grids (AbstractMesh subtypes) defining Domain sizes.
• calls register_methods! on each Reaction, allowing Reactions to define local VariableReactions and register ReactionMethods. ReactionMethods can be added to one of three groups, by calling add_method_setup! (called during model setup), add_method_initialize! (called at start of main loop) add_method_do! (called during main loop).
• links VariableReactions to VariableDomains, applying any renaming from the variable_links: sections in the config file.
• calls register_dynamic_methods! on each Reaction, allowing Reactions to create VariableReactions and register ReactionMethods that depend on VariableDomains created by other Reactions.
• relinks VariableReactions to VariableDomains to include these additional VariableReactions
• sorts each group of ReactionMethods into dispatch order based on dependencies between Variables, storing the sorted lists in the Model struct.

Model initialisation

To initialise the model, the numerical solver or host should then:

• call create_modeldata to create a ModelData struct that will store data Arrays for model Variables. A default Vector of CellRanges covering the whole model, generated by create_default_cellrange, is also stored in the cellranges_all field of the ModelData struct.
• call allocate_variables! to allocate memory for model Variables, storing these Array references in the ModelData struct.
• call check_ready to check that all Variables are allocated.
• call initialize_reactiondata! to:
  • store sorted lists of ReactionMethods and corresponding Variable array accessors in the ModelData struct. A ReactionMethod may include a prepare function to include additional buffers etc.
  • call create_dispatch_methodlists to compile default model-wide lists of initialize and do methods + corresponding Variable array accessors and CellRanges. This is stored as the in the dispatchlists_all of the ModelData struct.
• call check_configuration to call any optional check_configuration Reaction methods to validate the model configuration.
• call dispatch_setup with attribute_name = :setup, which calls Reaction setup methods (registered by add_method_setup!) to initialise any internal Reaction state and any non-state Variables (eg grid variables).

State Variable initialization

To initialise state Variables and perform any Reaction-specific initialisation, the numerical solver or host should:

• call dispatch_setup with attribute_name = :norm_value. These Reaction setup methods should set state Variables according to the :norm_value Variable attribute (which can be set in the variable_attributes: section of the config file). Reactions may also use these values internally.
  • optionally (if normalisation values are required by a numerical solver), copy the Variable normalisation values from the arrays in ModelData to the solver.
• optionally call dispatch_setup with attribute_name = :initial_value to set state Variables according to the :initial_value Variable attribute (which can be set in the variable_attributes: section of the config file).

Main loop

To calculate the model derivative, the numerical solver or host should:

• call do_deriv to call lists of initialize and do methods created by create_dispatch_methodlists

Additional notes

Operator splitting and Domain tiling: Reaction and Domain subsets

The default initialisation described above calculates the model derivative for all model Reactions for all Domains, using the cellranges_all, and dispatchlists_all fields of the ModelData struct. A custom Vector of CellRanges can be used to select a subset of ReactionMethods by Domain and operatorID, eg to implement operator splitting. Tiles within Domain eg to implement Domain decomposition for a multithreaded model can be defined by specifying CellRanges with a subset of cell indices within Domains.

The custom Vector of CellRanges should then be supplied to

• create_dispatch_methodlists to create lists of methods + corresponding CellRanges which can then be supplied to do_deriv for the model main loop.
• Optionally to VariableAggregators to create subsets of Variables and tiles within Domains for a numerical solver.

Multithreaded models

To use with a multithreaded solver eg for a spatially tiled model:

• Set threadsafe=true when calling create_modeldata. The subsequent call to allocate_variables! will then allocate Julia Threads.Atomic variables for PALEO Variables with attribute atomic: = true (usually scalar accumulator variables for totals etc).
• Supply a method_barrier implementing a thread barrier function to initialize_reactiondata!. Reactions are sorted into groups, where each group has no dependencies and later groups depend on earlier ones. The method_barrier will be inserted between these groups of independent Reactions.

Automatic differentiation

• Set eltype when calling create_modeldata to the appropriate dual number type for a (forward) automatic differentation implementation. All model arrays are held in the created ModelData struct and multiple ModelData instances can be created, allowing a combination of eg AD Jacobian and standard derivative calculations with different array types.