search

johnrsibert / tagest

Suite of software for spatially resolved fish tagging analysis
BSD 2-Clause "Simplified" License
3 stars 2 forks source link

Incorrect halflife estimates #9

Open johnrsibert opened 8 years ago

johnrsibert commented 8 years ago

Thomas Peatman reports "I've run in to issues with computing half-lives for the models fitted to PTTP data (both with and without fishing). I've been approaching this by manipulating the fleet q's in the run1.par file. Setting q's to zero has no impact on halflife estimates. I've attempted to trace back the cause. Debugging in halfproc.cpp suggests that fishing mortality is zero in the halflife calculations, regardless of the value of fleet q's."

johnrsibert commented 8 years ago

The error was incorrect computation of th average fishing effort used in the halflife computation. New fishing effort average computations are implemented in the indexed_regional_fishery_record class member function get_average_effort_array(const int month, d3_array& t) in file fish_rc3.cpp. Halflife estimates are generally higher with q_f = 0.

I recommend running halflife with graphics disabled (ie flag 78 set to 0). It is faster and avoids apparent bugs in the display.