3. What kind of models to fit

[ericward-noaa]

1. Question of how spatially structured is recruitment:
   a. test model with spatial + spatiotemporal variation to model with spatiotemporal only

2. Model each cohort individually, with AR1 spatiotemporal fields
   a. Autocorrelation across cohorts, etc

Tinkering on how to deal with the year effects.
a. Main effects
b. Spatially varying coefficients