Abstract
By implementing different observational likelihoods in a state-space age-based stock assessment model, we are able to compare the goodness-of-fit and effects on estimated fishing mortallity for different model choices. Model fit is improved by estimating suitable correlations between agegroups. We show by simulations that modelling catch as numbers-at-age is more suitable than proportions if the uncertainty of age-classifications is small
Original language | English |
---|---|
Publication date | 2015 |
Publication status | Published - 2015 |
Event | ICES Annual Science Conference 2015 - Copenhagen, Denmark Duration: 21 Sep 2015 → 25 Sep 2015 |
Conference
Conference | ICES Annual Science Conference 2015 |
---|---|
Country/Territory | Denmark |
City | Copenhagen |
Period | 21/09/2015 → 25/09/2015 |