Hey!

I was wondering is you could verify that OpenIFS Cy40r1 callpar.F90 can produce cloud fraction (3D or layered) that are greater than 1.0?

I've been seeing values as high as 1.43 and I find this very strange? If someone is running OpenIFS could check this,I would appreciate it.

I simply check the maxval() of the STATE_TMP%A(KDIM%KIDIA:KDIM%KFDIA,:) variable.

 

5 Comments

  1. Unknown User (nagc)

    Hi Marston,

    STATE_TMP is a temporary array in callpar and yes STATE_TMP%A holds the cloud fraction variable temporary state. It is possible  within  the timestep in 40r1 to have cloud fractions greater than 1, but this is not what is output at the end of timestep from the model. Cloud fractions > 1 have no physical meaning and they are clipped to 1 when applying the tendencies for the next timestep (cloud fraction is different to the water species as it is not a conserved quantity). The cloud source/sink terms could be formulated slightly differently to avoid the need for clipping, but numerically it is done this way so that overcast cloud fractions can easily be exactly equal to 1, rather than just a bit less than 1 all the time.

    Glenn

    1. Marston Ward

      Hi Glenn,

      I thought this was the case but the relatively high values made me a bit unsure and therefore I asked.
      Thanks for the clarification. I'm going to create a wiki page for OpenIFS and COSP for future users and developers. This info will be useful.

      /M

      1. Unknown User (nagc)

        Great, let us know when the wiki page is created. If it's public we can link to it from the OpenIFS pages.

         

  2. Unknown User (nagc)

    p.s. I assume you are passing the cloud fraction into COSP. It's correct to use STATE_TMP%A but values need clipping to ensure maxval() is 1.0

     

    1. Marston Ward

      Yes, I'm passing CC to both rttov and COSP. I clipped it using the MIN() function.