AFSC_VAST_Evaluation

Repository for evaluation of VAST package for single and multispecies spatial delta-generalized linear mixed models (delta-GLMM). The purpose of this analysis is to evaluate differences in indices and uncertainty estimates generated by VAST and design-based estimators, for a range of species in the Bering Sea and Gulf of Alaska.

Analyses

Individual analyses are contained within the examples folder, and inclue:

Name	Description
Test_Parallel	Example of how to create wrapper function to run VAST models in parallel across species.
Test_Obs_Model	Evaluation of differences in VAST model results for example species, with different observation models.
Test_DesignBased_Estimator	Simple comparison of design-based and VAST model-based indices across years and species.
Test_Knot_Number	Script to run comparison of VAST indices with differing levels of spatial complexity, as specified by the number of "knots".
Test_Autoregressive	Compares influence of different autoregressive structures for VAST model parameters over time.
Test_DeltaModel	Script to generate stratified Delta model biomass indices, and compare with design-based estimates.

VAST Package Citation

Please cite if using the software

Thorson, J.T., and Barnett, L.A.K. In press. Comparing estimates of abundance trends and distribution shifts using single- and multispecies models of fishes and biogenic habitat. ICES J. Mar. Sci. URL

VAST Package Website

Data from NOAA: Resource Assessment and Conservation Engineering Division

Please visit the following website for more information: RACE

curryc2 / AFSC_VAST_Evaluation