Extensions to SCAA applications reported in: further applications of statistical catch-at-age assessment methodology to the 2J3K-O Greenland halibut resource

Abstract

This document reports refinements to the survey-based SCAA assessments reported at an earlier meeting in Vigo, and attempts to provide results for the set of future analyses requested there. Particular attention has been paid to attempting to reduce the residual patterning evident in earlier assessments through taking further account of serial correlation. These efforts seem to have been reasonably successful for the overall survey indices and commercial catch-at-age proportions, but less so for the survey catch-at-age proportions, which consequently remain somewhat overweighted in the fitting process. A simpler agestructured production model is also fitted to the data, and gives similar results to the New Baseline SCAA assessment that is developed, with Bayesian estimates of precision computed for both these approaches. Despite these efforts to incorporate serial correlation, some conflict remains amongst the different sets of input data, and partly in consequence the absolute scale of biomass is poorly determined by assessments. The most pessimistic (in stock status terms) of the SCAA variants considered produce biomass estimates that do not differ that greatly from those from XSA. Importantly however, even in those cases the SCAA assessments provide results for recent years more in line with survey index (and CPUE) trends, and give more positive projections for future abundance: for example all SCAA projections under a constant TAC of 22750 tons increase, whereas XSA projects a decrease in those circumstances.