I have a question about a notable discrepancy observed in two 10-year aquaplanet simulations.

The initialization files (ICMGG<expid>INUA, ICMGG<expid>INIT, ICMSH<expid>INIT) and fort.4 used were the same. The only difference between the simulations is that one was run on a cluster using the Cy43r3v1 source code, while the other used Cy43r3v2. However, according to the release notes of Cy43r3v2, both versions are supposed to be identical. They also had reliable acceptance tests. So, what could potentially account for the approximately 0.5 K and 1 m/s differences in zonal and time mean values?

Best regards,
Thomas


2 Comments

  1. Marcus Koehler

    Hi Thomas,

    Just to clarify, there are differences between v1 and v2 of OpenIFS 43r3 as per the release notes.  The question is only if the aquaplanet setup makes any use of these parts of the code that have changed (e.g. mass fixers, solar spectrum).  Did both simulations run on the same hardware and were they built with the same compiler and compiler version?

    I have not used the aquaplanet setup myself, so I am hopeful that other users who have more experience with this could perhaps help you further.  I also have also no feeling if a difference of 1 m/s and 0.5 K after a 10 year simulation would be very significant (it may well be?).

    Cheers,

    Marcus

  2. Thomas Batelaan

    Hi Marcus Koehler,

    Thanks for your reply. Then I was a bit confused with the sentence 'OpenIFS 43r3v1 is scientifically identical to IFS 43r3'. I will check if the mass fixers were different.

    And no, the simulations were not run on the same hardware/had other compilers.

    I asked this question on the forum because, Victoria Sinclair and I found these differences after 10 years quite large.

    Great greetings,

    ~Thomas