commfish / seak_sablefish

NSEI sablefish stock assessment
8 stars 5 forks source link

1-hr soak time #16

Closed jysullivan closed 5 years ago

jysullivan commented 5 years ago

Work with groundfish biologists to assess the utility of the 1-hr soak time survey (1988-1996). If we decide to use this as an index obtain raw data. - We decided that these data do not provide a useable index of abundance. I have taken it out of the code.

jysullivan commented 5 years ago

Sent: Friday, January 11, 2019 1:26 PM To: Sullivan, Jane Y (DFG) jane.sullivan1@alaska.gov Subject: NSEI sablefish questions

Hey Jane,

Took some digging but I was able to find more details on the 1hr soak time and the issues ADFG ran into and why they made the change to the 3 to 11 hour soak time as well has the history of the management changes over time post 1985. I highlighted the important info in the attached documents so you can get at the main points. In general for historical fishery CPUEs prior to the EQS we had almost double the amount of permits fishing and was down to 1 day seasons where we still exceeded the quota so it was definitely a race to fish scenario so soak times would be fairly short and we had poor management control using shorter season lengths so as soon as EQS was implemented this made the fishery more manageable and extended season to 30 days and then went longer as soon as the number of permit holders was reduced.

For your questions in regards to if we should include survey data that used 1-hr soak times. I think that data should be excluded as there are huge caveats with catchability on a 1-hr soak time and skimming through some of the OP plan notes it mentioned that deeper sets may have not even hit the bottom and properly soak so I think that data is fairly questionable and we can justifiably exclude it.

Working on the tagging data next.

Let me know if you have any questions, thoughts, etc.

Thanks,

-Andrew

jysullivan commented 5 years ago

Based on this information Ben, Andrew, and I decided to remove 1-hr soak time from the model.