search

commfish / AlaskaHerring

Age-structured Assessment Model for Alaska Herring Stocks
2 stars 1 forks source link

Catch/maturity assumption #6

Open jysullivan opened 5 years ago

jysullivan commented 5 years ago

The assumptions related to maturity catch is different in HER and LS: HER applies the maturity curve to catch numbers-at-age, such that spawning_Nij = maturity_ij x (total_Nij – catch_Nij), and LS assumes catch to be 100% mature spawning_Nij = (maturity_ij x total_Nij) – catch_Nij.

Update from 2018-08-31:

This is a subtle difference (one parenthesis) between HER and LS that results in a lack of convergence (Hessian does not appear to be positive definite) when I code HER to match the LS assumption. I’ve been in touch with Sherri on this issue, providing output of mature and catch numbers-at-age. None of the values are negative. When I run HER with the assumption that all catch is mature, ADMB produces a report file. I generated a file displaying all the figures from this report file, and there are no glaring differences except that the estimate of survival went from ~0.5 to 0.6. One potential path forward is to code the LS model like the HER model to see if it runs and substantially changes the results.