Skip to content

Meeting minutes

Jump to bottom
Robert Gamble edited this page Dec 6, 2019 · 61 revisions

20191206

Present: Scott, Sarah, Andy, Joe C, Ryan, Rob

Apologies: Sean

Purpose: Quick status update

Status update for issues (3 minutes per issue)

  • reactiveAtlantis - diet changing parameters (add link)

  • Need to start moving the responsibilities for dealing with Atlantis biology/catch tuning to Rob

    • ACTION: Ryan and Rob run from latest Atlantis parameter files and compare
    • ACTION: Call/Meetup to get update on what you've been doing and current issues - Ryan/Rob (next week / Tuesday)

  • clean up shop

    • UPDATE: New stuff pushed
    • ACTION: Look into Docker? (Ryan)
    • Milestone: High priority - continue working

  • Catch - Whole Region Data

    • UPDATE: Noticed some problems with species with the new catch file.
    • ACTION: Need to check the problematic species
      • HAL - Halibut (early on)
      • FLA (other flatfish - should not be much catch here, these are mostly non commercial species)
      • ANC - Anchovies (probably should not have any removals?)
      • GOO - Goosefish (it could be that I need to crank up recruits here, this has always been a problem species)
      • INV (invasive species)
      • TUN - Other tunas
      • BFT - Bluefin tuna
      • BIL - Swordfish, marlin, sailfish (no catch)
      • FDE (shallow demersal fish like shad, river herring; could be crash related)
      • BSB - Black Seabass (could be crash related)
    • Milestone: High priority, 2 weeks (Rob)

  • Catch File Parameterization

    • UPDATE: Finished for global catch, but there are errors in the output with regards to habitat_ - required length of 11 (Rob)
    • ACTION: Fix the parameter file to fix the errors.
    • Milestone: High priority, finish before next meeting (Rob)

  • Biomass - Growth

    • UPDATE: Rescaling species with new catch file
    • Milestone: High priority, next meeting (Ryan)

  • Biomass - Diet

    • UPDATE: Rescaled and diet looks better
    • ACTION: Continue tuning to data, get the data to tune to (Rob)
    • Milestone: Medium priority, 1-2 months?

  • Biomass - Migration

    • UPDATE: None
    • ACTION: Get the data to tune to - problematic species that we don't catch in the trawls (Ryan M)
    • Milestone: Medium priority, 2 months?

  • Decision on Fleet Structure

    • UPDATE: Meeting scheduled for week of December 9th - how to categorize certain fleet types
    • ACTION: Develop fleet categorizations based on gear type and location
    • Milestone: Low immediate priority but will become higher, 2-3 months

  • Develop QA/output testing scripts

    • UPDATE: None
    • ACTION: Look at what's there and continue working down the list
    • Milestone: High priority, keep working

  • Long term stability

    • UPDATE: None - Done as far as the code
    • ACTION: Get the biology fixed first
    • ACTION: Modify code to be less stringent
    • ACTION: Split into two separate issues - code for testing vs model parameterization
    • Milestone: High priority, keep working

  • [Get Joe involved]

    • UPDATE: Probably starting Monday
    • ACTION: Get his e-mail and invite to next meeting

LAST MEETING -- LAST MEETING

20191115

Purpose: Quick status update

Status update for issues (3 minutes per issue)

  • clean up shop

    • UPDATE: Ryan removed some stuff and needs to be pushed.
    • ACTION: Look into Docker? (Ryan)
    • Milestone: High priority - continue working

  • Catch - Whole Region Data

    • UPDATE: Code is more general with start and end dates
    • ACTION: Need to merge with main repository and close out
    • Milestone: High priority, 2 weeks (Rob)

  • Catch File Parameterization

    • UPDATE: Finished for global catch, but there are errors in the output with regards to habitat_ - required length of 11.
    • ACTION: Fix the parameter file to fix the errors.
    • Milestone: High priority, finish before next meeting (Rob)

  • Biomass - Growth

    • UPDATE: Rescaled MUM and clearance parameters and recruits. Reserve nitrogen values a lot better but falling. Some species just didn't grow at all or died out and just getting the forced recruits.. Atlantic states demersal fish, salmon, etc.
    • Milestone: High priority, next meeting (Ryan)

  • Biomass - Diet

    • UPDATE: Rescaled and diet looks better
    • ACTION: Continue tuning to data, get the data to tune to (Rob)
    • Milestone: Medium priority, 1-2 months?

  • Biomass - Migration

    • UPDATE: None
    • ACTION: Get the data to tune to - problematic species that we don't catch in the trawls (Ryan M)
    • Milestone: Medium priority, 2 months?

  • Decision on Fleet Structure

    • UPDATE: Meeting scheduled.
    • Milestone: Low immediate priority but will become higher, 2-3 months

  • Develop QA/output testing scripts

    • UPDATE: Works with new repository name
    • UPDATE: Folded in calibration criteria from Pathybridge's paper.
    • ACTION: Look at what's there and continue working down the list
    • Milestone: High priority, keep working

  • Long term stability

    • UPDATE: None - Done as far as the code
    • ACTION: Get the biology fixed first
    • ACTION: Modify code to be less stringent
    • ACTION: Split into two separate issues - code for testing vs model parameterization
    • Milestone: High priority, keep working

  • [Get Joe involved]

    • UPDATE: Got Joe involved
    • UPDATE: Joe B unable to get run with latest version
    • ACTION: Get everyone on same version running with latest NEUS parameterization

20191106

Purpose: Quick status update

Status update for issues (3 minutes per issue)

  • change repo name

    • UPDATE: Did it, closed
    • ACTION: Look for unintended consequences

  • clean up shop

    • UPDATE: No major changes due to Windows 10 update and Atlantis no longer working on Ryan's machine
    • ACTION: Get Atlantis working on Ryan's machine
    • ACTION: Look into Docker?
    • ACTION: Create NEUS-Atlantis-Doc repository (when things are 'final')
    • ACTION: Keep notes for R and data files in RMD or text file
    • ACTION: Make new folder for tables in R directory
    • Milestone: High priority - continue working

  • Catch - Whole Region Data

    • UPDATE: Regional catch file as one pool across whole region is done. R script and data file uploaded.
    • ACTION: Make code more general (start and end dates)
    • ACTION: Need to merge with main repository and close out
    • Milestone: Before next week, high priority

  • Catch File Parameterization

    • UPDATE: Finished - except to update for 1964 start

  • Biomass - Growth

    • UPDATE: Growth too high but stable for a number of species, esp deposit feeders
    • ACTION: Fix growth
    • Milestone: High priority, 2-3 weeks

  • Biomass - Diet

    • UPDATE: Everyone's eating deposit feeders
    • ACTION:
    • Milestone: Medium priority, 1-2 months?

  • Biomass - Migration

    • UPDATE:
    • ACTION:
    • Milestone: Medium priority, 2 months?

  • Decision on Fleet Structure

    • UPDATE:
    • ACTION: Schedule meeting to look at the discussion points
    • Milestone: Low immediate priority but will become higher, 2-3 months

  • Develop QA/output testing scripts

    • UPDATE: Visualizations available on docs folder
    • ACTION:
    • Milestone: High priority, keep working

  • Long term stability

    • UPDATE: Visualizations available
    • ACTION: Get the biology fixed first
    • ACTION: Modify code to be less stringent
    • Milestone: High priority, keep working

_Adjourned at _ 4:10 PM

20191031

Present: Rob G, Sarah G, Andy B, Ryan M, Sean L, Scott L

Apologies:

Purpose: Weekly milestone meeting -- close out Baseline

Status update for issues (3 minutes per issue)

  • change repo name

    • UPDATE: We all agree
    • ACTION: Rob G will do that and update links in Wiki and issues
    • ACTION: Ryan M and Rob G will test to make sure it works
    • Milestone: Before next week, high priority

  • clean up shop

    • UPDATE: No major changes due to Windows 10 update and Atlantis no longer working on Ryan's machine
    • ACTION: Get Atlantis working on Ryan's machine
    • ACTION: Look into Docker?
    • ACTION: Create NEUS-Atlantis-Doc repository (when things are 'final')
    • ACTION: Keep notes for R and data files in RMD or text file
    • ACTION: Make new folder for tables in R directory
    • Milestone: High priority - continue working

  • Catch - Whole Region Data

    • UPDATE: Regional catch file as one pool across whole region is done. R script and data file uploaded.
    • ACTION: Make code more general (start and end dates)
    • ACTION: Need to merge with main repository and close out
    • Milestone: Before next week, high priority

  • Catch File Parameterization

    • UPDATE: RG needs to finish this and test.
    • ACTION: Turn on catch for all groups with catch in comland
    • Milestone: Next week, high priority

  • which species groups should match trends

    • UPDATE: Split from the ecosystem indicator issue
    • ACTION: Weakfish and redfish need to be checked (Sarah G)
    • Milestone: Within 2 weeks? Medium/High priority

  • assessment biomass trends for diagnostics

    • UPDATE: Atlantic cod, mackerel and herring have been processed
    • UPDATE: Repo created to hold this (https://github.com/NOAA-EDAB/assessmentData)
    • ACTION: Change name to 'assessment outputs'.
    • ACTION: Create lookup table to map the column name of the headings to the information table we want to use
    • Milestone: Lower priority in near future, pick them off as time permits

  • Biomass - Growth

    • UPDATE: Growth too high but stable for a number of species, esp deposit feeders
    • ACTION: Fix growth
    • Milestone: High priority, 2-3 weeks

  • Biomass - Diet

    • UPDATE: Everyone's eating deposit feeders
    • ACTION:
    • Milestone: Medium priority, 1-2 months?

  • Biomass - Migration

    • UPDATE:
    • ACTION:
    • Milestone: Medium priority, 2 months?

  • Decision on Fleet Structure

    • UPDATE:
    • ACTION: Schedule meeting to look at the discussion points
    • Milestone: Low immediate priority but will become higher, 2-3 months

  • Develop QA/output testing scripts

    • UPDATE: Visualizations available on docs folder
    • ACTION:
    • Milestone: High priority, keep working

  • Long term stability

    • UPDATE: Visualizations available
    • ACTION: Get the biology fixed first
    • ACTION: Modify code to be less stringent
    • Milestone: High priority, keep working

Preview of ZZZ (15 minutes)

  • Overview of tasks
  • ...

Other business

+ ACTION: Rob G to run model meeting next week and use that to do a quick progress meeting

_Adjourned at _ 4:10 PM

20191023

Present: Rob, Ryan, Andy, Sarah, Scott, Sean

Purpose: Weekly milestone meeting -- Updates on Cleanup and Catch issues (https://github.com/NOAA-EDAB/atneus_RM/milestone/1)

Status update for issues (3 minutes per issue)

Baseline issues:

  • clean up shop
    • UPDATE: Still cleaning up shop... Geometry file with lat/lon
    • ACTION: Create one baseline set of files but include alternative parameterizations to explore different assumptions - name them consistently - we'll think about how to do this

Catch issues:

  • catch files

    • UPDATE: R script to apportion to Atlantis groups is done.
    • ACTION: Need to verify shrimp and other inverts

  • Stability issues:

    • UPDATE: Run wasn't stable under the stability criteria based on the SOE trend plots
    • ACTION: Don't want to base this on looking at a plot. Could use the ecotrend package which should be identical to the trend function we use.
    • ACTION: Check sensitivity - perhaps call a +/- 5% trend is stable
    • ACTION: Runs of longer years than 20 - could increase run to 30 or 100 years

  • Groups and indicators issues:

    • UPDATE: Groups done
    • ACTION: Indicators need to be picked - use Beth's paper as starting point
    • ACTION: Prebal
    • ACTION: Separate Groups from Indicators and mark Groups as done

  • diagnostic scripts

    • UPDATE: Checklist of stock assessments to get information for
    • ACTION: start getting data together for comparison of age, length/weight, diet (see also above discussion)
    • ACTION: Put assessments into EDAB branch drive, includes assessment PDF, R scripts, final outputs
    • ACTION: Use tabulizer package in R
    • ACTION: Use length/weight calculator in Andy's mscatch package

20191009

Present: Rob, Ryan, Andy, Sarah

Apologies: Scott, Sean

Purpose: Weekly milestone meeting -- close out Baseline

Status update for issues (3 minutes per issue)

Baseline issues:

  • clean up shop

    • UPDATE: progress made with three commits
    • ACTION: still needs further cleanup and documentation, especially the R folder, working on it today

  • which species groups should match trends

    • UPDATE: preliminary analysis is done based on survey biomass, catch, and value. discussion on whether to get assessment biomass for comparisons, yes because this is the currently accepted trend so we should match it for review. Also, some species not sampled by our survey. Will use assessment and survey B but prioritize assessments. So still not ready to close
    • ACTION: double check calculations, not sure why weakfish is showing up from biomass, seems weird
    • ACTION: Need to define a set of ecosystem indicators to match: BiomIndx.txt outputs pelagic:demersal ration, piscivores:planktivores, and a couple other things already so we can use them, just need to decide how to calculate and from what data for comparison
    • ACTION: find stock assessment biomass trend data for each of the ~25 species identified (new issue 14to address reference dataset needed for issue 10

Other issues:

  • catch files
    • UPDATE: Rob made the catch file that links functional groups to SVSPP codes and posted its location. Need to write up methods so we can repeat (done manually). Andy created packages for each database in cfdbs; svdbs, and obdbs, can use these to recreate.
    • ACTION: Need to apply catch to functional groups. Sarah did this already to determine which species comprise 80% of catch. See DiagnosticsDoc; code snippet is
 # need comland, should be able to run similar code as above to get catch and revenue
 # NOTE: THIS FILE NOT TO BE POSTED ON GITHUB DUE TO POTENTIAL CONFIDENTIALITY CONCERNS
 # ask Sean for it
 load(here("diagnostics", "comland_meatwt_deflated_stat_areas.RData"))

 # from https://github.com/NOAA-EDAB/Atlantis-Catch-Files/blob/master/Atlantis_1_5_groups_svspp_nespp3.csv
 spcodes <- readr::read_csv("Atlantis_1_5_groups_svspp_nespp3.csv")

 # time series in case we want to plot them
 comlandts <- merge(comland, spcodes) %>%
   group_by(YEAR, Name) %>%
   summarise_at(SPPLIVMT, sum)

  • long term stability

    • UPDATE: first long run done, no fishing, slight differences in biology file. will need to document differences and "why" for reviewers if need to change parameters. smaller (half to a third) growth rates with fishing than without, still constant recruitment but did change inverts. estimating higher recruitment
    • ACTION: Sarah to drop results from this run into the diagnostics script today or tomorrow
    • ACTION: Switch to 73 day output so that we have same "seasons" in each output year, evenly divides 365
    • GOAL: have persistence and long runs with no fishing that are as similar as possible to final parameterized run with fishing. And be able to explain why the differences are necessary.

  • diagnostic scripts

    • UPDATE: in progress in output_testing branch, working versions of code to test for persistence, long term stability, and code in progress for comparisons of trends (right now, making the reference dataset)
    • ACTION: see above, double check existing code and fix species name matching, etc
    • ACTION: start getting data together for comparison of age, length/weight, diet (see also above discussion)

Other business

  • Ryan also working on diets which are too invert heavy, reducing availability inconclusive at this point. Looking at percent of diet, fish are squid heavy and mammals as well. Rob will get diet comps for comparisons. We have code for diet comp in the old ECSA for summer flounder, still needs a lot of cleanup. Sarah will turn into a future issue to go into diagnostic scripts--first step get diet data in Atlantis groups, approximately 2001-2010. Same for age and growth comparisons.

Adjourned at 11:25 am

20191002

Present: Scott, Sarah, Andy, Ryan
Apologies: Sean, Rob, Laurel, Mike (in Reno!)
Purpose: Weekly milestone meeting -- close out Baseline

Status update for Baseline (3 minutes per issue)

  • Create Atlantis groups file with SVSPP/NESPP codes
  • Determine time period for hindcast comparisons
    • UPDATE:
      • Options presented in the issue.
      • Discussion: time series for hindcast comparison should not include spinup (first decades where population dynamics are not yet stabilized). Looks like 15-20 years for fish groups in current model.
      • Define spinup as 20 years for now, unless it changes with later parameterizations
      • Want to combine a time series where we can assess trend (~30 years, Hardison et al 2019) with a current snapshot for population productivity
      • Current model parameterized with ~mid-2000s hydrography, so going too far back in time may be incompatible with ecosystem trends, however temperature sensitivities are not currently turned on.
      • Once Joe incorporates the full ROMS hydro hindcasts, then we can extend the hindcast comparisons back to be compatible with that (1960s on)
      • Important side conversation: do we need to change the polygons to implement climate? this is a fundamental change to the model. Ryan thinks not likely to see cold pool in current polygon structure (checked, is there but perhaps too far offshore due to box structure). May be possible to change only mid-Atlantic polygons? Still a major model overhaul. Issue/Task for Joe: evaluate current polygon structure, decide which climate features must show up and what is the minimal set of polygons that can reproduce them?
      • Decision made: Go with 30 year hindcast (1980-2010) for trend comparison, minimal length of time series for trend detection but not too far from current conditions, not including heaving fishing by foreign fleets, current set of dominant species. But hybrid approach using current snapshot for matching population productivity characteristics.
    • ACTION:
      • start next steps (make list in issue):
        • gathering datasets (from SOE ecodata? stock assessments) for comparisons
        • by species groups (next issue)
        • list of indicators (next issue)
      • close task
  • Determine which species groups and indicators must match trends
    • UPDATE:
      • Options presented and discussed: need biomass trends but also catch, need to have fleets set up properly
      • Match for whole system or areas? Whole system minimally, areas if to be used in analysis
      • Update from Andy on mscatch matching to Atlantis polygons/regions: not done yet, need the actual Atlantis polygons, currently thinking at EPU level which means we don't need to do length extrapolations due to missing data, would need to implement at the polygon level. MScatch pulls landings data, fills in gaps for missing species/areas. Length expansion at EPU level. Still need to repackage Sean's existing work on filling in gaps landings-->EPU. Andy very close to finishing, attributing weight to length distributions after aggregating to EPU is last step. Then ready for Hydra_est, stock assessment analysis.
      • Suggest that we go no smaller than EPU for spatial comparisons of general model to avoid overfitting individual groups and sacrificing other aspects of match
      • Need to map current EPUs to Atlantis polygons, wont be a perfect match but just make a decision and document. Ryan says box 9 is only real issue, Nantucket shoals area. Designed to match surveys. Solve for MScatch.
      • Could be a consideration in designing boxes for Atlantis 2.0
      • Decision made
    • ACTION:
      • Sarah to look at ecodata for 1980-2010 by EPU to get dominant species by biomass by Oct 9
      • Need to look at stock assessments for dominant species (catch vs biomass vs revenue)
        • need scallops and lobster as well as biomass dominant groundfish
        • include species in other models (Hydra, Rpath, etc)
      • Once species agreed on, official catch comes from stock assessment inputs for the time period
      • Divide and conquer among Andy, Sarah, Rob, maybe Laurel
      • Create a fleet structure issue
  • Clean up shop
    • UPDATE: pass 1 done e4af6485b3ee10fba1dc3971750824364e039d2c
    • ACTION: create archive folder, make more progress this week

Preview of persistent biology (15 minutes)

  • Did not discuss

Other business

  • Need to close or update current issues, prioritize tasks for next milestone prior to next meeting Oct 9
  • Baseline deadline for remaining tasks extended to Oct 9
  • Need broader conversation on next steps--3-4 month outlook as well as 1-2 week, do with Joe, starting Oct 15, implementing climate and polygon reordering necessary?
  • Other discussion: Atlantis Docker from Hem, look into this for multiple people running the same codebase

Adjourned at 11:10 am

YYYYMMDD (template)

Present:
Apologies:
Purpose: Weekly milestone meeting -- close out XXX

Status update for XXX (3 minutes per issue)

  • YYY
    • UPDATE:
    • ACTION:
  • ...

Preview of ZZZ (15 minutes)

  • Overview of tasks
  • ...

Other business

  • ... _Adjourned at _

Running Atlantis

Calibration

Model Components

Model Criteria

Documentation

Clone this wiki locally