search

nmfs-ost / ss3-source-code

The source code for Stock Synthesis (SS3).
https://nmfs-ost.github.io/ss3-website/
Creative Commons Zero v1.0 Universal
36 stars 16 forks source link

Tag recapture - major revamp #11

Open k-doering-NOAA opened 3 years ago

k-doering-NOAA commented 3 years ago

Imported from redmine, Issue #18312 Opened by @RickMethot on 2017-11-07 Status when imported: New

need: time-vary parameters sharing of parameters fleet-specific and release specific reporting rates

k-doering-NOAA commented 3 years ago

comment from @iantaylor-NOAA on 2017-06-02: Attaching document from 2012 discussing ideas for revamp of tagging data. On today's (June 2, 2017) AMWG call, Patrick Lynch suggested that a postdoc to work on this would be a good subject for a SAAM proposal.

k-doering-NOAA commented 3 years ago

comment from @RickMethot on 2020-02-28: Ideas from Yukio: As for the proposed revision of tag i/o. I quickly read the document. It seems that this probably solve some of the difficulty to use tag model in ss, I felt when I applied ss to skipjack model.

At that time, since ss assigns different tag sheding and iverdispoersion parameter to each tag group, in practice estimation of those parameters with in ss are very difficult. The proposal will address those concern and would be very useful.

In addition it may be usefull if additional "mirroing" feature like to pick up tag id to mirror any one of parameter is also implemented. This may be added by allowing negative values in 2nd-5th columns so that -x in init_los in the row of Tag_Group_id of y, for example, indicates that init_loss parameter of Tag_Grou_id of y mirros that of Tag_Group_id x. Of cource y < x may be required. This may allow to replicate "Tag group" in MFCL

As for the reporting rate, in the case of tuna assessment in the WCPO, since several countries and SPC as well conducted their own tagging program in parallel, reporting rates are deinged to be parmetrized by fllets and "tag program". Proposed conding for reporting rates can manage the situation in WCPO partially. However it might be difficult to deal the situation that multiple tag program were conducted at the sam eperiod. The different reporting rate depending on the different tag profgram may seem to be strange. They might be attributed by the use of different dart tag (size of tag or manufactor) or by different tagger. Those factors may need to be deal by tag loss rate. However historically they may have been different reporting on several stages of reporting. They are due to the difficulty of reporting of recapture of tag released by a country and recaptured by another country. Increased colaboration among different countries are addressing those issues. But we do not how well it worked in the past.

However I may misunderstand the meaning of the affiliated tag IDs, can you tell me the concept of affiliated tag more detail.

The other idea came up to me is that to add one more dimension to the tag data to represent the "grouping" of tag group. It may say "suepr" tag group which may be used several ways. E.g. this may be used to grouping of tag_group I mentioned above. Also if we can have that additional dimension in tag data. It may also be used to structurize reporting rate to address the issue of different reporting rate by different tag program.