Seminar on highly relevant topic for ModSquad, Oct 13

[Lewis-Barnett-NOAA]

Please try to attend this thesis defense by one of Curry Cunningham's students. It is at 1:30PM PST tomorrow, Friday Oct 13. Link at bottom.

ACCURACY AND PRECISION OF MODEL-BASED ABUNDANCE INDICES ESTIMATED BY
INTERCALIBRATING DATA FROM MULTIPLE SURVEYS

Indices of abundance from fishery-independent surveys represent a crucial source of data
informing the modern stock assessments used to set catch limits for groundfish in Alaska.
However, financial and logistical constraints limit the number of survey sampling events per
year, and/or the frequency with which regions are sampled, and it remains challenging for any
individual survey to sample all habitats occupied by a particular stock. Recent work has
explored the use of model-based methods which estimate trends in stock abundance by
intercalibrating data from multiple surveys, potentially addressing these challenges. However,
there has been limited prior research assessing the relative performance of alternative
methods for model-based intercalibration of survey data. In this analysis, we compare the
accuracy and uncertainty of abundance indices estimated by three intercalibration models,
Random Walk Timeseries (RWTS) models, Generalized Additive Models (GAMs), and Vector
Autoregressive Spatiotemporal (VAST) Models, using catch-rate data from three fishery-
independent surveys around Alaska. RWTS, GAM, and VAST models produced estimates with
comparable levels of precision and estimated abundance trends similar to reference indices.
Simulation analyses found that the rate at which the size-composition of the underlying stock
fluctuated was a strong influence on both accuracy and precision of model estimates, and that
this influence appeared to be compounded by the level of contrast in the size selectivity of the
constituent surveys. These results support the findings of previous work that identified the
influence of size structure and size selectivity on the success of intercalibrating indices of
abundance. We recommend that future research explore survey intercalibration model
structures that directly represent size-specific abundance and relative catch efficiency.

in partial fulfillment of the requirements of the MS in Fisheries

Tristan Sebens
MS Fisheries Candidate
Adviser: Dr. Curry Cunningham

October 13, 2023
12:30 PM AKST

Presenting: Juneau Lena 101
Zoom: https://alaska.zoom.us/j/87331949693#success