search

pacific-hake / hake-assessment

:zap: :fish: Build the assessment document using latex and knitr
MIT License
13 stars 6 forks source link

Add numbers of fish to table of number of hauls (Table 5) #462

Closed iantaylor-NOAA closed 2 years ago

iantaylor-NOAA commented 5 years ago

From Google Doc:

6B. The SRG recommends that the JTC include information in the next assessment on the annual numbers of fish underlying each annual age-composition observation and present an analysis of the potential influence of changes in sampling. [relevant to time-varying data-weighting impacts] perhaps add the number of total fish sampled in the existing table that shows numbers of hauls (Table 5) - Chris making first stab at new/revised table wait and see what we find with temporal variation in numbers of fish sampled for age per haul

iantaylor-NOAA commented 5 years ago

@cgrandin I can't remember if the plan was to just create a new separate table or additional rows in the existing one. Perhaps you can just make a placeholder and then let us all know that we need to provide the data.

If getting data on number of fish from early years is difficult we can just focus on the recent years for now.

aaronmberger-nwfsc commented 5 years ago

We have a response to this in our Response to SRG requests and I have looked at the total number of samples through time (currently real episodic around a central mean - alas, the impact of biennial surveys. So will look at just data from fishery too. This can be something to bring to SRG as a back-pocket thing if they decide to dive into it during the review. I was also looking at some summary metric like the number of catch per unit sample or the exploitable biomass per unit sample, samples per haul, etc... just to get an idea of how variable age comp sampling has been over the years. Just FYI... but think we can 2020 milestone this for now.

andrew-edwards commented 5 years ago

Yes, keep defer it for now....

kellijohnson-NOAA commented 5 years ago

I also initially explored a random walk on the D-M parameters such that they were allowed to vary with time and the hessian was NOT Positive Definite. I didn't spend a lot of time on it though.

aaronmberger-nwfsc commented 4 years ago

We are going to start this by first looking at sampling protocols by fleet/nation as a first step to understanding this source of variation

cgrandin commented 4 years ago

I just looked at this again for 2020. It would require more data extractions from us and the building of a new table. @kellijohnson-NOAA would this be enough to address sampling protocol changes? Number of fish aged for each year and age by sector?

cgrandin commented 4 years ago

I worked out these numbers for Dave Sampson in 2020. See #680. I will add the Canadian numbers this year

aaronmberger-nwfsc commented 3 years ago

The 2020 SRG requests for the 2021 assessment (and beyond) also mention this so we should consider the two issues referenced there related to age compositions and effective sample sizes when thinking about this.

kellijohnson-NOAA commented 3 years ago

So, I extracted Nfish for shoreside and there are aged fish all the way back to 1965 from Washington. So, there is no way to know now many of the aged fish that are in the database were actually used in the pre 2008 compositions. Thoughts?

aaronmberger-nwfsc commented 3 years ago

But shoreside sampling doesn't even start until 1990, at least in terms of what's in the assessment. My guess is that previous assessors through that out for particular reasons. Do you think post-1990 might be reasonable, especially if it at least closely matches what I gathered from previous assessments? Do you think the pre-1990 are actually from the at-sea or JV vessels? I wonder if JV was landing shoreside so got labelled as such. So hard with historical data...

kellijohnson-NOAA commented 3 years ago

at-sea sampling (NORPAC) data starts in 1991 and additionally has foreign ages back to 1975. Pacfin has data going back to 1965.

kellijohnson-NOAA commented 3 years ago

@cgrandin I think we are going to table this until next year when we redo the historical age composition data right? So, can we change the milestone?

cgrandin commented 3 years ago

That sounds good to me, setting to 2022.

kellijohnson-NOAA commented 2 years ago

Well we did not "redo" any old data for this assessment, nor do I plan on being able to between now and February. So, it seems like for the 2022 assessment it would be good to get the number of fish per year for recent Canadian sampling to match what is included for the U.S. with respect to number of fish sampled for each sector per year. @cgrandin can you get this for Canada, which would be Tables 8-9 for the 2021 document?

cgrandin commented 2 years ago

I'll try, I'm swamped right now trying to convert the new SS in combination with getting the cloud runs working for retrospectives, etc.

This may have to happen later in January

cgrandin commented 2 years ago

Done in 91b5458cbd39cfa80064ff88453d5cb39fe5ab4c