Closed EmilyMarkowitz-NOAA closed 4 months ago
I feel like I need more context on this. The "content" portion of this issue seems like a misunderstanding about the fact that we recalculate all the biomass, agecomps, sizecomps for all years after each new survey year. So it's not impossible that the 1999 biomass estimates were calculated with DESIGN_YEAR = 2022. Thus, I'm not sure why we need a new column or an update to the DESIGN_YEAR column. Em, maybe we should chat about this.
Duane
On Wed, Feb 7, 2024 at 6:35 PM Em Markowitz (NOAA) @.***> wrote:
Assigned #23 https://github.com/afsc-gap-products/gap_products/issues/23 to @Duane-Stevenson-NOAA https://github.com/Duane-Stevenson-NOAA.
— Reply to this email directly, view it on GitHub https://github.com/afsc-gap-products/gap_products/issues/23#event-11740119221, or unsubscribe https://github.com/notifications/unsubscribe-auth/ANKDWAVFACXAAAC2AGLOF5TYSQ2WJAVCNFSM6AAAAABC63NMZ6VHI2DSMVQWIX3LMV45UABCJFZXG5LFIV3GK3TUJZXXI2LGNFRWC5DJN5XDWMJRG42DAMJRHEZDEMI . You are receiving this because you were assigned.Message ID: @.*** com>
-- Duane Stevenson, Ph.D. Supervisory Fish Biologist Groundfish Assessment Program NMFS, Alaska Fisheries Science Center
I had the impression from the conversation in the virtual room that much of the confusion about DESIGN_YEAR had to do with the introduction of a lookup field that the users didn't anticipate having to leverage and was less about which design year's stratum areas you used to do your computations. That said, I would suggest that both concepts, the one Duane raised and the one I raised, be explored the next time we get the team together in Office Hours.
Sounds good. It'll be good to table this until @zoyafuso-NOAA can join in on this discussion, too!
@Duane-Stevenson-NOAA is correct in that all survey years' indices are calculated with the 2022 version of the EBS stratum areas. So the DESIGN_YEAR value for YEAR = (1982-2023) for the EBS SURVEY_DEFINITION_ID = 98 should all be DESIGN_YEAR=2022. It sounds redundant but this format is important when the GOA survey time series goes from DESIGN_YEAR=1984 for YEAR = (1990-2023) and DESIGN_YEAR=2025 for YEAR = (2025-onwards) due to the restratificaiton.
This table is definitely used internally within gapindex and is the main reason for existing. I was probably not privy to this conversation at the last office hours so I am not sure how this table is used outside of that. It seems like folks want to query the GAP_PRODUCTS.AREA with just the appropriate stratum records used in the index calculation?
@EmilyMarkowitz-NOAA you’re right about the NBS years, thanks for cleaning that up. For the slope, there shouldn't be a record for the BSS with YEAR = 2023 because there wasn't a slope survey in 2023. However, when the time is appropriate, the value for DESIGN_YEAR for the BSS years will be updated from 2002 -> 2023.
Oh thanks for clearing that up, Zack. So... I think I am the one who was confused about what this table was for and suspect I have likely mislead folks/started this confusion. I thought it was documenting what stratum calculations estimates were previously calculated with (which I now see is wrong). We'll just clear it up at the next office hours I suppose? But glad I was able to fix a few other finds! @zoyafuso-NOAA, you'll go through everything here and implement stuff as needed and close the issue? Or is there anything else I can provide/help with?
Also, we currently make the AKFIN_SURVEY_DESIGN table available to AKFIN. Do we want to remove that table from the dump if it is internal/should only be needed in gapindex?
I think for now we can still provide the SURVEY_DESIGN table to AKFIN and continue discussions with users as to whether this table will still be useful to them before removing the table from the AKFIN table transfer.
Issue Description
1. Content: Shouldn't
DESIGN_YEAR
in theGAP_PRODUCT.SURVEY_DESIGN
table be the year that year's area calculations were calculated in? Here, all EBS entries haveDESIGN_YEAR
s of 2022 even for YEARs before 2022 (which is impossible). For example and if I recall correctly, design-based estimates calculated in 2023 and 2022 were calculated with area estimates where DESIGN_YEAR = 2022, but estimates from 2021 thru 2019 were originally estimated with the area calculations produced in 2019, 2018-2010 with 2010, 2009-1993 with 1993, and so on. This will be an issue with NBS and (I believe) BSS, too.2. Description: Stock assessors asked for some more direction and documentation about how to use/impliment the SURVEY_DESIGN table. Here is what is currently listed: