HYCOM/CICE versus ACNFS
Jim Hunt
I'm not currently a user of HYCOM per se, merely an observer of the output. Further to Alan's heads up
over on the Arctic Sea Ice Forum, I was wondering if anyone here can explain
why the current ACNFS version of Arctic sea ice thickness looks so
different from the global HYCOM/CICE/NAVGEM version of nominally the same
thing? What are the "improvements" in the global model, and which one is closest to the "reality" in the Arctic?!
alan.wa...@hycom.org
The usual plots for this case are at:
http://www7320.nrlssc.navy.mil/GLBhycomcice1-12/skill.htm
The usual ocean data, plus sea ice fields, will be served from hycom.org.
The ocean (HYCOM) component of what will be GOFS 3.1 once it is
operational has been improved over ACNFS, but the CICE version is the
same. The sea ice differences are 1) the GOFS hindcast started from
more realistic initial conditions (primarily a better ice thickness)
than ACNFS, 2) GOFS assimilates ice concentration data across the full
Arctic domain (not just along the ice edge as in ACNFS). Neither one
assimilates sea ice thickness.
The Global system's Arctic total sea ice volume is closer than ACNFS's
to that from PIOMAS.
Compared to IceBridge's 2013 Arctic campaign, Global ice thickness is closer to the observed flight data in the regions of the Beaufort Sea and Canadian Archipelago whereas ACNFS has lower error for the region north of Greenland.
Alan.
souls...@gmail.com
That all makes sense. A supplementary question or two if I may. Are ACNFS/GOFS 3.1 still using CICE 4? If so are there any plans to move to CICE 5 at some point in the foreseeable future?
alan.wa...@hycom.org
GOFS 3.5, on a 1/25 degree grid (2 km in the Arctic), will use CICE 5. It is currently scheduled for transition to NAVOCEANO early in 2016, and might be available here before then.
An "ESPC" system that is coupled to the atmosphere (HYCOM+CICE+NAVGEM+WW3) for long term forecasting will also use CICE 5. Its IOC (initial operational capability), with an ensemble on the 1/12 degree global HYCOM/CICE grid, is scheduled for 2018. See http://www7320.nrlssc.navy.mil/pubs/2014/metzger1-2014.pdf.
Alan.
