Capelin September data
The foundation for modelling capelin is the yearly joint IMR-PINRO trawl-acoustic survey in September. The survey results are transferred to the Excel spreadsheet CapTab, which is considered the authoritative source of information on the capelin stock from this survey. In addition, files of historic bootstrap replicates of the September estimates are stored on files, one file for each year. Two different methods has been used: bootstrapping using a parametric model for earlier data and bootstrapping the input data for later years. In Bifrost, the bootstrapped replicates are scaled to yield the same as the CapTab data on sex-aggregated age and used when replicates of maturation and predation parameters are estimated.Bootstrap replicates based om parametric resampling of acoustic data (earlier data)
In the time period when individual Sa-values are not available the resampling of mean integrator values in each square is based on a published parametric model while the biological data are resampled non-parametrically. This is done in the notebook CapelinData.
Short description of the program
The replicates for each year are stored in the list bootstrapAgeLengthEstimates with the format:
bootstrapAgeLengthEstimates is a list over bootstrap replicates of the September estimate of capelin based on parametric resampling of acoustics, with the format:
List over sex (1: females, 2: males, 3: Total)
List of age (1-5)
List of length - The mean length in each group is given in capelinDataLengths
number (billion)
List of mean weight by length.
The basic idea behind incorporating uncertainty of the September survey into estimating replicates of parameters in Bifrost is to generate yearly replicates of the September data by resampling. Replicates of the historic time series of September data are then constructed by resampling from the the yearly replicates. The method of choice is to use non-parametric resampling throughout. In the first part of the historic period, however, the basic acoustic data (Sa values) are lacking, only mean integrator values by squares are available.
firstYearEarlierData 1972
lastYearEarlierData 2004
Bootstrap replicates based on resampling acoustic data (later data)
For the later period where all raw data from both Norwegian and Russian vessles are available the resampling is performed with the program ReCap, where non-parameric resampling has been conducted throughout.
bifrostAutumnCapelinReplicates and bifrostWinterCapelinReplicates are lists of replicates with the format:
List over area, first is area 0 (outside of Multspec system), rest are areas 1-7
List over length, 0.5 cm length intervals, first element is the length group 0.0 - 0.5 cm
Weight
List over age
Number of females (billion)
Number of males (billion)
firstYearLaterData 2005
lastYearLaterData 2010
Summary and CVs of joined time series
| File name | Date last change | Number of replicates |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1972 | {2004, 3, 30, 16, 4, 56.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1973 | {2004, 3, 30, 16, 11, 0.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1974 | {2004, 3, 30, 16, 16, 58.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1975 | {2006, 10, 3, 9, 3, 18.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1976 | {2004, 3, 30, 16, 22, 48.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1977 | {2004, 3, 30, 16, 27, 46.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1978 | {2004, 3, 30, 16, 30, 22.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1979 | {2004, 3, 30, 16, 33, 8.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1980 | {2004, 3, 30, 16, 37, 46.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1981 | {2004, 3, 30, 16, 43, 32.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1982 | {2004, 3, 30, 16, 48, 24.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1983 | {2004, 3, 30, 16, 55, 44.} | 300 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1984 | {2004, 3, 30, 17, 6, 56.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1985 | {2004, 3, 30, 17, 20, 2.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1986 | {2004, 3, 30, 17, 25, 28.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1987 | {2004, 3, 30, 17, 28, 22.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1988 | {2004, 3, 30, 17, 35, 24.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1989 | {2004, 3, 30, 17, 41, 40.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1990 | {2004, 3, 30, 18, 5, 20.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1991 | {2004, 3, 30, 18, 21, 12.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1992 | {2004, 3, 30, 18, 34, 28.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1993 | {2004, 3, 30, 18, 41, 54.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1994 | {2004, 3, 30, 18, 47, 20.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1995 | {2004, 3, 30, 18, 53, 14.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1996 | {2004, 3, 30, 19, 0, 34.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1997 | {2004, 3, 30, 19, 9, 56.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1998 | {2005, 5, 27, 12, 24, 56.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology1999 | {2005, 5, 27, 12, 36, 42.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology2000 | {2005, 5, 27, 12, 55, 26.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology2001 | {2005, 5, 27, 20, 2, 46.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology2002 | {2005, 5, 27, 20, 15, 22.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology2003 | {2005, 5, 27, 20, 24, 52.} | 349 |
| c:\Data\Capelin\BootstrapCapelinTrueTrueTrue\bootstrapSexAgeLengthAcousticBiology2004 | {2005, 5, 27, 20, 41, 42.} | 349 |
| c:\Notebooks\bifrostAutumnCapelinReplicates2005 | {2009, 8, 31, 13, 14, 46.} | 31 |
| c:\Notebooks\bifrostAutumnCapelinReplicates2006 | {2009, 12, 3, 19, 41, 48.} | 301 |
| c:\Notebooks\bifrostAutumnCapelinReplicates2007 | {2009, 12, 2, 13, 11, 22.} | 301 |
| c:\Notebooks\bifrostAutumnCapelinReplicates2008 | {2009, 12, 3, 17, 3, 38.} | 301 |
| c:\Notebooks\bifrostAutumnCapelinReplicates2009 | {2009, 12, 1, 13, 3, 10.} | 301 |
| c:\Notebooks\bifrostAutumnCapelinReplicates2010 | {2011, 2, 22, 12, 38, 17.} | 301 |
The uncertainty in the September capelin estimate has a large influence on the suggested quota. The figure below shows the time series of CV (standard deviation divided by mean) when both the old and the new method have been used.
Time series organised for Bifrost
When Bifrost is run, one replicate of the number by age by sex by length replicates of the September data is allocated to each historic year. When the present notebook is made, such replicates are drawn. The first series of replicates is the baseline, i.e. there is no resampling and the first replicate each year is used. These data are stored on the files predrawnSeptemberReplicates for later retrieval by the routine readData when Bifrost is initialised. Before storing, the number at age each year is scaled to the number at age in the data on CapTab, i.e. it is the length- and sex distribution that is resampled. Also, the weight at age is expressed in kg as oposed to gram, which is used in the raw data. The mean weight in each length group is not taken from CapTab.
0 replicates of the historic times series will be made.
When the likelihood components from the capelin September data is constructed, the variance is needed, and for this reason the lists bootstrapTime, bootstrapTimeFemales and bootstrapTimeMales are constructed, containing the CV by age for each year.
Tables for the first replicate
Year {} have no data
Year {} have no data
Explanations of routines and variables
