The Entity-Process Framework for Integrated Agent-Based Modeling of Social-Ecological Systems

Abstract

The success of Integrated Assessment and Modeling of social-ecological systems (SESs) requires a framework allowing members of this process to share, organize and integrate their knowledge about the system under consideration. To meet this need and ease management of successful modeling processes, we present a conceptual framework for integrated agent-based modeling and simulation of SESs in the form of a formal “entity-process meta-model”, along with a distinction between three levels of models—conceptual, concrete and simulation—and characterization of the research question using indicators and scenarios. We then describe how to represent the structural and dynamic dimensions of SESs into conceptual and concrete models and to derive the simulation model from these two types of models. Finally, we discuss how our framework solves some of the challenges of integrated SES modeling: integration and sharing of heterogeneous knowledge, reliability of simulation results, expressiveness issues, and flexibility of the modeling process.

Appendix

Outline of the MAELIA simulation platform that is organised according to the Entity-Process Framework. The full details and documentation are available on the MAELIA web-site http://maelia-platform.inra.fr/ (in French; Accessed 22 May 2018).

1.1 Entities

Reading the complete actors-resources diagram of the MAELIA platform is simplified by grouping entities into four parts or “modules”. Each actor or resource (of which we give here only the labels) are endowed with attributes and operations (see below). The documentation also provides an overview description and the Actors-Resources diagram.

Agricultural Module

• Actors’ labels: farmers;

• Cognitive resources’ labels: field bloc; irrigation bloc; cultivated species; cropping plan; cropping system; crop management strategy; weather area;

• Material resources’ labels: crop; farm; islet; irrigation device; field; water withdrawal device; agricultural soil;

Hydrology Module

• Labels of material resources: dam; channel; homogeneous response unit; stream; groundwater; local reservoir ; hydrological soil; hydrological zone; average weathering zone;

Regulation Module

• Actors’ labels: dam manager; agricultural water manager; water policy;

• Cognitive resources’ labels: administrative sector; monitoring point; management unit; administrative zone;

Other uses Module

• Actors’ labels: district; industries;

• Material resources’ labels: domestic withdrawal device; industrial water withdrawal device; domestic reject device; industrial reject device;

1.2 Processes

The labels of the main processes dynamically linking the entities (and programmed in the platform) are listed below. The documentation also provides an overview description of each modelled process and the corresponding interaction diagram.

• Description of the scheduling of process activation ((by the simulator engine))

• Description of the model initialisation

Agricultural Module

• Cropping plan decision process;

• Soil-Crop Dynamics (2 alternatives models): “AqYield” (plant growth model, initially set for field crops); simple plant dynamics;

• Water withdrawal;

• Crop Management Strategy: plowing; sowing; hoeing; irrigation; harvest;

• External processes: field area change; farm economy; cropping plan (as data);

Hydrology Module

• Chanel hydrology;

• Hydrology (2 alternative models): hydrology simple model; SWAT model: [includes: land phase; water routing phase; local reservoir fulling];

• External processes: input flow; weather;

Regulation Module

• Withdrawal volume allocation; control and sanction; edict drought decree; dam water release

Other uses module

• Domestic water withdrawal; industrial water withdrawal; domestic water reject; industrial water reject;

• External process: population growth;

1.3 Template for the Description of an Entity

Definition :

System reference’ element to which it refers, role, scale;

Activities :

for actors only;

Attributes :

Name, public/private; description; data type, unit, variable/constant; read/write process accesses;

Links :

Name; description; target entity; read/write process accesses;

Operations :

Name, public/private; functional description; input/output parameters; formal description;

Initialisation :

Link to the Pre-processing section;

Discussion :

Explanation of the modelling choices, …

References :

References to external documents (if required);

Implementation :

Description of implementation details (only for the modellers);

1.4 Template for the Description of a Process

Definition :

Purpose, internal/external;

Scale :

Temporal, spatial scale(s) (if relevant);

Interface :

Type, identity and attributes of accessed entities;

Scheduling :

If applies to several entities;

Formal definition :

Specification for implementation;

Validation :

Results of sensitivity analysis, calibration, …

Discussion :

Explanation of the modelling choices, …;

References :

References to external documents (if required);

Implementation :

Description of implementation details (only for the project’s members);

Documentation provides further information on the following topics:

Pre-processing :

Description of the data sources and processing to generate each of the input files for model initialisation and external processes;

Indicators :

Available for everyone, information source entities and computation method;

Scenarios :

For each scenario, one overview page and explanations about the differences with the standard (baseline) model;

Glossary :

Definition of technical terms, parameters, acronyms, etc.