Repository navigation
Tutorials Workflows Examples
This page gathers practical guides for using OSMOSE. The pages listed here are task-oriented: they are intended to help users run an existing configuration, build or modify a configuration, add model components, run scenarios, calibrate a model, and analyse outputs.
The tutorials and workflows complement the conceptual and reference pages of the wiki. Conceptual pages explain how OSMOSE works, while workflow pages describe how to carry out specific modelling tasks.
Some extended tutorials and examples are also available in the OSMOSE guidelines repository.
-
Run an existing configuration
Explains how to run an existing OSMOSE configuration, such as a demo or a shared model application. -
Build a new configuration
Describes the main steps needed to create a new OSMOSE configuration. -
Add a species
Explains how to add a species or group to an existing configuration. -
Add fishing
Explains how to add or modify fishing processes, including fishing mortality, selectivity, catchability, effort, and related inputs. -
Add resource forcing
Explains how to include resource species or low-trophic-level forcing in an OSMOSE configuration. -
Calibrate a model
Introduces the workflow for preparing and running a calibration. -
Run scenarios
Describes how to organise simulations designed to compare alternative assumptions, management options, or environmental scenarios. -
Analyse outputs
Explains how to read, summarise, plot, and interpret OSMOSE outputs.
These pages provide short guidance on specific technical aspects of running OSMOSE.
-
Allocating Java memory for OSMOSE experiments
Explains how to adjust Java memory settings when running OSMOSE experiments, especially for large simulations, calibration runs, or memory-intensive configurations.
New users should start with Run an existing configuration before trying to build or modify a configuration.
Users preparing a new model should continue with Build a new configuration, then move to the pages on adding species, fishing, and resource forcing as needed.
Users working after a simulation should read Analyse outputs. Users preparing model comparisons or future projections should read Run scenarios.