The Use Of Geostatistical Simulation For The Estimation Of The Total Annual Egg Production Of Atlantic Mackerel (Scombrus Scombrus) With Associated Confidence Intervals.Ices Cm 2003/X:12

No abstracts are to be cited without prior reference to the author.

This paper presents some of the work undertaken as part of the recently completed EU funded project GBMAF, which involved combining Geostatistical and Bayesian methods to improve the scientific basis for the Management of the Atlantic mackerel Fisheries. Part of the research focussed on the use of geostatistical simulation to improve on the accuracy and precision of the estimation of the total annual egg production (TAEP) of Northeast Atlantic mackerel over the traditional methodology. While standard kriging-based geostatistical methods can provide unbiased estimates, calculation of the global estimation uncertainty is less straightforward. Conditional simulation offers an alternative method for obtaining confidence intervals by generating a number of equiprobable realisations of the variable surface, based on the available data and the spatiotemporal correlation structure. Sequential Gaussian simulation is used to calculate the TAEP and associated confidence intervals for each of the 9 surveyed years (1977 to 2001) to date. The results are compared to those obtained using the traditional method. It was acknowledged that the current methodology involved a number of assumptions and did not account for a number of additional sources of uncertainty. However, the results demonstrated the potential of conditional simulation as a valuable tool for the estimation of TAEP and its associated uncertainty.