Compsets, Codebase and customization for NorESM2.5 twin simulations

[mvertens]

Code Base:
The code base for the NorESM2.5 twin simulations will be noresm2_5_alpha03_v2 and will be placed in NorESMhub when its ready. I need to test this on betzy to make sure that the simulations listed in #7 can be built and run out of the box.
A few things to note:

• all AMIP and fully coupled compsets without interative ice sheets should now use DGCL%NOEVOLVE to replace CISM%NOEVOLVE. This should be done so that when interactive ice sheets are turned out things will work correctly.
• zm tuning settings from CAM should be used for when running with OSLO-AERO. This is because these simulations will be using the ZM deep convection scheme from CAM. The NorESM physics version of the modified ZM deep convection scheme from the previous version of noresm_develop cannot be used because the ZM scheme in CAM has been completely refactored, including its interface (i.e. - you cannot call the old scheme anymore). This involves setting the following namelist variables in user_nl_cam:

zmconv_c0_lnd		=  0.0075D0
zmconv_c0_ocn		=  0.0300D0
zmconv_ke		    =  5.0E-6
zmconv_ke_lnd		=  1.0E-5

• DMS and BROMO will now be sent from BLOM to CAM when BLOM%ECO is the BLOM configuration. I'm not sure that CAM will actually do anything with the BROMO fields at this point - but they will be received in the NUOPC cap.
• In comparing the COARE versus default atm/ocn flux schemes - its important to note that a new gust parameterization has been added to the default scheme only. This is turned on by the following mediator namelist which have the following out of the box values:

     add_gusts = .true.
     gust_fac = 0.0D0

In shr_flux_mod the following has been added:

    ! Convective gustiness appropriate for input precipitation.
    ! Following Regelsperger et al. (2000, J. Clim)
    ! Ug = log(1.0+6.69R-0.476R^2)
    ! Coefficients X by 8640 for mm/s (from cam) -> cm/day (for above forumla)
    ugust(gprec) = log(1._R8+57801.6_r8*gprec-3.55332096e7_r8*(gprec**2))

....

      !--- compute some needed quantities ---
      if (add_gusts) then
          vmag   = max(seq_flux_atmocn_minwind, sqrt( (ubot(n)-us(n))**2 + (vbot(n)-vs(n))**2 + (1.0_R8*ugust(min(rainc(n),6.94444e-4_r8))**2)) )
          ugust_out(n) = ugust(min(rainc(n),6.94444e-4_r8))
      else
          vmag   = max(seq_flux_atmocn_minwind, sqrt( (ubot(n)-us(n))**2 + (vbot(n)-vs(n))**2) )
          ugust_out(n) = 0.0_r8
      end if
      wind0 = max(seq_flux_atmocn_minwind, sqrt( (ubot(n)-us(n))**2 + (vbot(n)-vs(n))**2) )

ugust_out is sent back to CAM - but it looks like in this tag it only is used in the diagnostics. It would be good to make sure we want this setting in the comparison that @oyvindseland is doing.

Compset/Grid:

COMPSET: 1850_CAM%DEV%LT%NORESM%CAMoslo_CLM51%SP_CICE_BLOM%ECO_MOSART_DGLC%NOEVOLVE_SWAV_SESP
GRID: a%ne30np4.pg3_l%0.9x1.25_oi%tnx1v4_r%r05_w%null_z%null_g%gris4_m%tnx1v4

and the aliases are
N1850 and ne30pg3_f09_tn14.