Initialisation Troubleshooting

The following checklist can be used when diagnosing initialisation problems in OSMOSE 4.4.

General checks

• Confirm that the intended initialisation method is the one actually activated.
• Check that method-specific parameters are not being mixed unintentionally across workflows.
• Verify that all file paths are valid and point to the intended inputs.

Reconstructed population-level initialisation

• Check that population.initialization.relativebiomass.enabled = TRUE is set when this route is intended.
• Verify that biomass, yield, and catch-at-length files use the correct ndtPerYear.
• Check that observed.biomass.cutoff.size.spX is consistent with the observation process represented by the biomass index.
• Inspect the generated population.initialization.* block directly.
• Confirm that population.initialization.relativebiomass.spX, population.initialization.age.spX, population.initialization.tl.spX, and population.initialization.nschool.spX all have the same number of values.
• Confirm that population.initialization.size.spX has exactly one additional value relative to the number of classes.
• Remember that population.initialization.size.spX defines class boundaries, not class centres.
• Compare reconstructed biomass and yield with the supplied targets when such targets exist.
• Inspect larval mortality estimates for plausibility.
• Re-check growth, maturity, fecundity, lifespan, sex ratio, and length–weight parameters if reconstructed outputs look implausible.

Direct school-based initialisation

• Confirm that population.initialization.file points to the intended file.
• Verify that the file contains all required variables: species, x, y, abundance, age, length, weight, and trophiclevel.
• Check that age is expressed in years, length in centimetres, and weight in grams.
• Inspect the consistency of the school-level state, especially abundance and spatial coordinates.

Egg seeding initialisation

• Confirm that population.seeding.year.max is consistent with the intended duration of seeding.
• Check that seeding biomass is used for oviparous species and seeding abundance for viviparous species.
• Verify that the seeding values are consistent with their intended role: bootstrap versus low-level safeguard.
• Ensure that seeding does not unintentionally dominate the early dynamics when it is meant only as an anti-collapse mechanism.

Reproducibility

• Archive the actual initial-state representation used in the simulation:
  • the generated population.initialization.* block for reconstructed population-level initialisation;
  • the school-level file for direct school-based initialisation.
• Archive the upstream data-layer inputs as well, so that the origin of the generated initial state remains traceable.