[Wrf-users] SST turns to zero over sea

[Thorsten Beisiegel]

Dear Users,

I've a riddle and I could not find any track so far:

WRFv2.1.1 is running with Skintemperature from GFS in spite of SST. Finally in WRF I managed to get TSK  (skin temperature) changing over land with isfflx=1 giving reasonable values. Before it was constant in time.

Now over the sea areas TSK is decreasing reaching values of 200K after 72 hours simulated. 
I'm using RUC LSM in the 3 domains with isfflx=1
On the other hand when changing to isfflx=0, SST does not decrease but TSK keeps constant in time.

I noticed that the variables LANDMASK, XLAND and LU_INDEX have all a constant value of 1 all over the domains in the wrfout_d0* files and in the wrfinputd01 too. In spite of this in the wrf_real_input_em.d0* files also LANDMASK and TSK behave well in space and time.

I'm not sure if these constant values represent an error executing real.exe or if it's normal. Anyway somehow the model distiguishs water from land because TSK only decreases to unreal values over sea area.

In another run, some months ago, we achieved reasonable fields of TSK and SST over all the domain with constants values of 1 for LANDMASK, XLAND and LU_INDEX. So I don't know what change gave the reason, if it was a parameter of namelist, Registry or the input data.  

Lately I tried adding sst_update and the auxinput5_* parameters to the namelist, without success, I suppose because of XLAND=1 for all the domain.

Can anyone tell me which parameters could have an effect in order to get TSK or SST behaving well over the sea?

Many thanks

Thorsten