Optimized biophysical model for Icelandic cod (Gadus morhua) larvae

A characteristic of biophysical models (BPM) is that they contain a number of parameters that are poorly known or only known within a range of possible values. This paper describes an efficient optimized BPM developed to simulate the Icelandic cod pelagic 0-group survey data. The method is based on presenting the results as a probability density function (PDF) that a particle released from a given spawning drifts downstream to a given grid location some time later. The model determines egg production model parameters (peak spawning time, spawning duration, number of eggs released) for each of 15 spawning grounds as the solution of a bound constrained optimization problem that minimizes model-data misfits in abundance and age distributions. The model is applied to simulating the 2002 and 2003 summer survey data. The model does a reasonable job of simulating the observed inshore / offshore abundance gradient and spatial age gradient for each year. Problem areas are explained from the point of view of model limitations. We caution that the results from an optimized model should always be assessed with the model's limitations in mind and with respect to whatever biological data are available.