class: right, middle, my-title, title-slide # Developing a common dataset <br /> for multispecies model keyruns <br /> using a reproducible workflow in R ## WGSAM<br /> 11 October 2021 ### Sean M. Lucey, Andrew Beet, and Sarah K. Gaichas<br /> Ecosystem Dynamics and Assessment <br /> Northeast Fisheries Science Center --- # Georges Bank Multispecies Keyrun Project .pull-left[ - New England Fisheries Management Council (NEFMC) is considering a shift to place-base, ecosystem-based fisheries management (EBFM) - Example Fisheries Ecosystem Plan (eFEP) will focus on the Georges Bank Ecological Production Unit (EPU) - NEFMC's EBFM Plan Development Team proposed a management procedure using ceilings and floors - An implementation of this procedure using the Hydra model was reviewed by the Center of Independent Experts in 2018 ] .pull-right[ <img src="20211011_WGSAM_Lucey_files/figure-html/GBfoot-1.png" style="display: block; margin: auto;" /> ] --- # Georges Bank Multispecies Keyrun Project .pull-left[ - Several concerns remained - What happens with all the species in the region under proposed management regime? - How well do multispecies models perform for assessment? - How does this compare to what we are currently doing? ] .pull-right[ <img src="20211011_WGSAM_Lucey_files/figure-html/GBfoot2-1.png" style="display: block; margin: auto;" /> ] ??? Georges Bank MSKEYRUN project designed to address the first two concerns Develop potential operating models for a management strategy evaluation (MSE) that could be used to address the third --- # Multispecies/Ecosystem modeling in the Northeast US <center>  </center> --- # Motivation for a common data set .pull-left[ - Objectives will require the development (refinement) of several multispecies/ecosystem models - **LOTS** of data to sift through/ decisions to be made - Time spent organizing data is magnified when dealing with multiple species - Need to generalize solutions to common data issues - Did **NOT** want multiple versions of starting parameters ] .pull-right[  ] --- # Creating a reproducible workflow .pull-left[ - Lots of decision points - Spatial domain - Temporal domain - Focal species - Easy to forget why or even what decisions were made - Create a repository to document decisions - Tracked progress with a kanban board software ] .pull-right[  ] --- # Data, data, everywhere ... .pull-left[ ## Fisheries Independent - Survey Biomass - Length Composition - Age Composition - Diet/Consumption ] .pull-right[ ## Fisheries Dependent - Landings - Discards - Length Composition - Age Composition ] <center>  </center> ??? Fisheries independent data resides on the survey database schema (SVDBS) while fisheries dependent data is split between the commercial fisheries and observer databases --- # Creating R Packages to pull/manipulate data - Turned older R scripts into packages - Focused on making more accessible with consistent data outputs - Survey data - [Survdat](https://noaa-edab.github.io/survdat/) - Commercial landings and discards - [Comlandr](https://noaa-edab.github.io/comlandr/) - Commercial length data - [MSCatch](https://noaa-edab.github.io/mscatch/index.html) <center>  </center> --- # Creating a data repository - Some experience with [Ecodata](https://noaa-edab.github.io/ecodata/landing_page) for our State of the Ecosystem Reports - Converted documentation repository into data repository - Data and documentation in one place ([MSKeyrun](https://noaa-edab.github.io/ms-keyrun/)) --- <br> <br> <br> <br> # MSKeyrun data repository walkthrough --- background-image: url("EDAB_images/Ocean_sunset.jpg") background-size: cover ## Questions? # Thank you!