Target-type CMPs for SBT presented to the Seattle meeting are refined and extended to include the input of CKMR information through use of the associated TRO index developed by Hillary. Essentially CMPs, tuned to median recovery to 30% of the pristine TRO in 2035 are developed for each of the CPUE, gene tagging and CKMR data sets alone, and then weighted combinations of these are considered. A subset of the robustness tests which show the greatest differences in performance compared to the base/reference case (RC) OM are selected. Overall the CMP based on the CKMR data only seems to perform best for the RC, but when the selected robustness test results are also taken into account, a variant based on a weighted combination of CMPs using all three data types seems marginally preferable.
History
Department/Unit
Department of Mathematics and Applied Mathematics, University of Cape Town