BH Recruit Scenarios

[jcaracappa1]

Summary

Parameterized the Beverton-Holt recruitment parameters (alpha/beta) for the 22 species that were using constant recruitment prior. These include: ANC, BFT, BIL, BPF, FDE, FDF, FLA, FOU, HAL, INV, MEN, OHK, OPT, PLA, SAL, SDF, TAU, TUN, TYL, WOL, WPF, and WTF. All groups have nearly an adequate amount of recruits but will be merged to dev to account for new catch data before additional calibration is performed.

New Files

• R/BH_scenario_diagnostics.R : Creates a series of diagnostic tables to confirm that the correct parameter files where used in each scenario and that each run was the expected duration
• R/plot_BH_scenarios.R : Creates a figure of each scenario with grouping by species. Can be generalized later on for species-specific experiments.
• R/scale_BH.R : Utility script to scale BH parameters
• data/BH_convert_1_index.rds & data/BH_convert_2_index.rds : Data files showing the parameters used in each scenario set

R Script Changes

• atlantis_batcher.R : Reworked batcherfile tracking and fixed a bug with run indexing
• call_run_comparisons.R & test post-processing.R : Cleaned up old parameter files references
• R/make_baseline_BH.R : Edited to do estimation for all species (not just with recruittype = 1)

Parameter File Changes

• currentVersion/RunAtlantis.sh : Added the feature to print command line output to a separate file

at_biology.prm changes

• Changed flagrecruit from 1 to 3 for all species that aren't sharks or mammals
• Re-implemented old GOO diet
• Reduced GOO predation by sharks
• mum/C tuning for LOB, ZG, MAK, BPF, SUF, WTF, PLA, GOO, TYL, RED, SDF
• Reduced mQ for TUN and BIL
• Adjusted BH alpha/beta for many species based on calibration scenarios

[jcaracappa1]

@RGamble1 Please review and merge as soon as you can. Will run the new dev and compare when finished. Attached is the run comparison for reference. The results are a bit mixed but won't get a true picture until we have the new catch.

PullRequest_197_v_dev_Model_Comparison_Biomass.pdf

[RGamble1]

Looks good @jcaracappa1