Connections Between Models

In APSIM Next Generation, models interact through a structured system of connections, allowing one model to access information or functionality provided by another.

These connections are established through Links (i.e. [Link] in the source code), which define how models communicate and exchange biological or physical information within a simulation.

Overview

In biological or physical terms, a Link represents a relationship or dependency between two parts of the simulation.

For example, a Leaf model depends on information from the Phenology model to know when new leaves should appear, or from the Plant model to understand the overall plant state.

These relationships are automatically established by APSIM NG when the simulation runs, ensuring that models can share data consistently.

Purpose of Links

Links serve several key purposes in APSIM NG:

Integration of biological processes — e.g., linking photosynthesis with canopy development.
Re-use of common models — e.g., several organs may link to the same Thermal Time module.
Hierarchical organization — e.g., a Plant contains Leaf, Stem, and Root components that interact through internal connections.
Parameter sharing — e.g., a Structure model accessing FinalLeafNumber from a function defined elsewhere.

Types of Connections

Different kinds of connections define how APSIM NG searches for and establishes relationships between models.
These can be understood as different biological or structural scopes:

Connection Type	Description	Scientific Analogy	Typical Use
First Available	Searches upward or outward in the model hierarchy until a matching model is found.	A plant organ seeking information from its parent plant system.	Most common; used for high-level interactions such as organ ↔︎ plant communication.
Child Connection	Connects directly to a component located within the same model.	An internal process within the same organ (e.g., leaf temperature model within Leaf).	Used for tightly coupled sub-processes or response functions.
Named Connection	Connects to a model explicitly identified by name.	Targeted reference to a specific process or dataset.	Used when multiple similar models exist (e.g., several temperature response functions).

These types correspond to APSIM’s internal connection logic (e.g., By first appearance, Child, By name), but are presented here in biological terms for clarity.

Example: Connections in a Plant Structure Model

Below is connections used in the <strong>Structure</strong> model.

Connected Component	Connection Type	Role in the Model	Description
Plant	First Available	Broader plant context	Provides structural and lifecycle information for the plant.
Phenology	First Available	Developmental driver	Supplies current growth stage and timing of key events.
Leaf	First Available	Organ-level interaction	Provides information on leaf area, appearance, and expansion.
ThermalTime	Child Connection	Internal driver	Calculates accumulated temperature used for leaf and stem development.
Phyllochron	Child Connection	Internal driver	Defines the thermal interval between successive leaf appearances.
FinalLeafNumber	Child Connection	Genotypic parameter	Determines the maximum number of leaves for this genotype.
HeightModel	Child Connection	Structural process	Computes plant height as a function of temperature and development.
BranchingRate	Child Connection	Growth process	Determines the rate of new branch initiation.
BranchMortality	Child Connection	Senescence process	Determines the rate at which branches die or are lost.

Conceptual View

In a simulation, these connections can be visualized as information pathways:

Plant
├── Phenology
├── Structure
│ ├── ThermalTime
│ ├── Phyllochron
│ ├── FinalLeafNumber
│ └── HeightModel
└── Leaf

Each arrow represents a Link, enabling data flow from one model to another — ensuring biological realism and modular consistency.

Why This Matters

Understanding Links helps scientists interpret model behavior correctly:

When a model outputs or responds to a variable, that information may originate from another connected model.
Adding or removing sub-models (e.g., replacing one ThermalTime with another) changes biological relationships.
Proper configuration of Links ensures that all processes are working in harmony — avoiding missing dependencies or duplicated effects.