Measuring uncertainty in trawl surveys: implementation in DATRAS

Abstract

Abstract The basic data from several ICES coordinated trawl and beam trawl surveys in the Baltic Sea, the North Sea, the area west of Scotland and France are stored in the DATRAS database in the ICES Secretariat; every year and for each survey member countries report data to the database. The DATRAS system offers facilities for the calculation of fish stock abundance indices, these indices are fed into the routine fish stock assessment work. The indices should be supplemented by an estimate of their uncertainty and the present paper suggests a bootstrap algorithm for implementation in DATRAS to calculate such estimates of variation for each stock/area index in a given year. The procedure is a two-step bootstrap routine whereby the length compositions and age-length key are bootstrapped independently, i.e. the uncertainty estimate is based on variability between catches in individual hauls in that year. The goal is to supplement the routine indices with uncertainties estimates in the future.