Project Details
Description
The project looked into the importance of the
process error in SAM through 1) a comparison of SAM assessments for a range of
fish stocks and 2) a sensitivity analysis of the influence of the process error
in a selection of assessments. The results show that the magnitude of the
process error and the amount of correlation in the process error varied among
stocks, and were unrelated with each other. Large process error is observed
when the assessments produce a good fit to the catch data, indicating that a
good fit to the catches is not a guarantee of good assessment since all the
variability might be hidden in the process error. It was also found that large
process error was generally associated with large assessment uncertainty.
Finally, the importance of the correlation structure could not be explained. It
is considered that it is a sign that the model is unable to accommodate for
some trend (spurious or real) in the data, and moves it to the process error,
but this could not be demonstrated.
| Status | Finished |
|---|---|
| Effective start/end date | 1/01/15 → 31/12/15 |