George Bank Cod.ppt

George Bank Cod Base Similar to GARM2005 assessment Fleet_dome Survey_dome Survey_and_Fleet_dome Purpose: test models ability to estimate correctly when doming occurs GB cod basic setup Years 1978-2005 (27 years) Ages 1-25, 10+ Lengths 1-170 cm 3 market categories 3 surveys (CVs 20%, 30%, 25%) CV on landings 5% Growth stdevs 5.0 initial, 2.0 projection Von B params 148.2, 0.1198, -0.6173 50% Mature ~42 cm Selectivity flat-topped, 50% selected 64 cm M = 0.2 F high (0.5-2.4) then reduced in final years (0.48, 0.35) GB cod test cases Fleet_dome Moved ascending limb of selectivity curve left to allow for nearly full selectivity at peak of dome GB cod test cases Survey_dome Each survey mirrored for descending limb of selectivity curve at age 7 Georges Bank Cod AIM Georges Bank Cod Georges Bank Cod SCALE Georges Bank Cod VPA (SSB) VPA (FAA) Georges Bank Cod ASAP VPA vs ASAP (NAA) VPA vs ASAP (NAA) Precision (SSB and F) AIM: ~ 0.1, did not vary by case SCALE: 0.2-0.5 (increased with retros) VPA*: 0.05-0.4 (poor ages>6), did not vary ASAP: <0.05, did not vary *VPA CVs are for age-specific SSB, F GB Cod Conclusions Simulation set-up precluded model comparisons for AIM, ASPIC, and SCALE Between VPA and ASAP, bias was similar, but VPA estimates at age degraded faster than ASAP; ASAP more precise Dome not an issue for survey because of age-specific indices (rather than age comp. for total N index) Dome effect in fleet probably minimized because of large simulated F (fewer older fish, so misspecifying those selectivities added little bias) VPA PR VPA Fleet Dome Retro VPA q “East Coast Style” * SSB: slightly + F: slightly - ASAP SSB: slightly + F: slightly - VPA SSB: 5-30% + F: ~30% - SCALE One way trip problem ASPIC SSB: strong bias F: -40% to+20% AIM Fleet&Surv. Dome Survey Dome Fleet Dome BASE Issues with low (or no) sampling in some years SSB: slightly + F: slightly - ASAP SSB: slightly + F: slightly - VPA SSB: 5-30% + F: ~30% - SCALE One way trip problem ASPIC Division b