Large changes in model behavior after version 6708
Hem Nalini Morzaria Luna - NOAA Affiliate
Cecilie Hansen
Fulton, Beth (Environment, Hobart)
G’day
What do you think has changed in demography that is causing the change in numbers/biomass?
Look at the diff between the atdemography.c code prior to 6708 and the current trunk the only changes are to things there were going into fprintf (I commented them out so not doing the assignments as not needed, help squeeze a tiny bit more speed). There should be no change to any variable values.
What am I missing?
Beth
From:
Hem Nalini Morzaria Luna - NOAA Affiliate <hemnalini.m...@noaa.gov>
Date: Friday 31 May 2024 at 12:26 PM
To: Beth Fulton <Beth....@csiro.au>, Atlantis ecosystem model <atlantis-eco...@googlegroups.com>
Cc: Raisha Lovindeer <rlovi...@eoas.ubc.ca>, Alaia Morell <ala...@uw.edu>, Isaac Kaplan - NOAA Federal <Isaac....@noaa.gov>
Subject: Large changes in model behavior after version 6708
Beth:
Hi, I am catching up to the latest model version (6716) which has the new Q10 curve. We found large changes in model behavior when running that version, and I traced the changes back to version 6708. As you can see in the figure below, group biomass changes dramatically under v6708 & 6716 (our model crashes after 17 years because of some of the exploding vertebrates).
The SVN messages suggest changes between v6706 and v6708 were for contaminant handling, but I see no changes to atContaminants.c and several changes to atdemography.c in migration and recruit distribution.
Was this the intended behavior, from a necessary change?, that we should now recalibrate to?
Thanks
Hem
--