CRD 01-03 Table 13. Summary of ADAPT model formulations for combined
silver hake. |
|
Model Identifier
|
Main Feature
|
Survey Indices
|
Residual Patterns
|
Precision of Estimates
|
Selectivity Pattern for Older Ages
|
Comments on model fit relative to baseline
|
|
Run 11
|
Use all data
|
Fall:1-5
Spr:1-5
|
Very poor low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity
|
Baseline model
|
|
Run 12
|
Catchability change in 1980-81
|
Fall:1-5
Spr:1-5
|
Low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity
|
Little improvement to model fit
|
|
Run 13
|
Set F-ratio on plus-group to be 1/2
|
Fall:1-5
Spr:1-5
|
Low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity
|
Better residual patterns for ages 4 & 5
|
|
Run 14
|
Catchability changes in 1974-75 and 1980-81
|
Fall:1-5
Spr:1-5
|
Moderate low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity
|
Better fit to data
|
|
Run 15
|
Same as Run 14 with F-ratio set to be 1/2
|
Fall:1-5
Spr:1-5
|
Moderate low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity
|
Best fit to data
|
|
Run 16
|
Hyperdepletion: SQRT transform age-4,-5 indices
|
Fall:1-5
Spr:1-5
|
Low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity
|
No improvement to model fit
|
|
Run 17
|
Hyperstable: SQR transform age-1 to age-5 indices
|
Fall:1-5
Spr:1-5
|
Low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity
|
Modest improvement to model fit
|
|
Run 18
|
Exclude noisy age-1 index
|
Fall:2-5
Spr:2-5
|
Low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity
|
No improvement to model fit
|
|
Run 19
|
Use only fall survey data
|
Fall:1-5
|
Low to high patterns across ages
|
Very high CVs on population size estimates
|
Dome-shaped selectivity
|
No improvement to model fit
|
|
Run 20
|
Use only spring survey data
|
Spr:1-5
|
Low to high patterns across ages
|
Very high CVs on population size estimates
|
Dome-shaped selectivity
|
No improvement to model fit
|
|
Run 23
|
Increase M on ages 1 and 2
|
Fall:1-5
Spr:1-5
|
Low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity
|
Modest improvement to model fit
|
|
Run 24
|
Use only 1967-1999 data
|
Fall:1-5
Spr:1-5
|
Low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity
|
Modest improvement to model fit
|
|
Run 25
|
Use only 1981-1999 data
|
Fall:1-5
Spr:1-5
|
Moderate low to high patterns across ages
|
High CVs on population size estimates
|
Dome-shaped selectivity, except for 1997-1999
|
Better fit to data
|