Possible bug in shr_flux_atmOcn

[abozec@coaps_fsu_edu]

Hi, I'm currently comparing an "out of the box" POP run (POP/datm/dice/ from cesm1_2) with a HYCOM run using the same grid/bathy/forcing. The point of the study is to see what are the behavior of each model under the same conditions (or at least the closest as possible conditions). One of the routine I was checking was the bulk formulation atm/ocn (shr_flux_mod.F90 in shr_flux_atmOcn). It seems that in all the incrementations of cesm (from cesm1_0 to cesm1_2), the model uses thbot (air temp) for stability and NOT thvbot  (virtual temp) as used in CCSM3 and as in the original documentation of this bulk Formulation (Large and Yeager paper). Is there a scientific explanation to that change or is it a bug? If bug this : !--- compute stability & evaluate all stability functions ---
hol = shr_const_karman*shr_const_g*zbot(n)* &
(tstar/thbot(n)+qstar/(1.0_R8/shr_const_zvir+qbot(n)))/ustar**2should be !--- compute stability & evaluate all stability functions ---
hol = shr_const_karman*shr_const_g*zbot(n)* &
(tstar/thvbot(n)+qstar/(1.0_R8/shr_const_zvir+qbot(n)))/ustar**2for the 2 iterations of the calculation.From my tests with HYCOM and the same bulk formulation but using the air temp instead of virtual temp, the difference between the two formulation gives a change of the order of ~10W/m2 at the equator and the opposite at high latitudes for the latent and less than 5W/m2 for the sensible. Thanks for any infos on that,  Alex

 

[abozec@coaps_fsu_edu]

Correction on my part, sorry changing thbot to thvbot only change the latent and sensible calculation by about a few W/m2 at most (less than 5). My later comment was based on simulation where I changed something else as well, my apologies. But for consistence, I still think it shoulf be thvbot in this routine ... Regards, Alex

 

[njn01]

The forumlation in CCSM3  is incorrect (bug #764).  The problem was corrected in 2008, and the corrected version was introduced in CCSM4. All CESM1 versions include the correction.