Data comprise a total monthly fish catch in tonnes from January 1977 to December 2006.
We fitted an observation-driven model to the logarithm of the total monthly fish catch. The model contains seasonal effects
and time-lagged terms for the preceding two months.