NLDAS-2 regrid

[Sujan Pal]

Hi all,

I am having a problem with forcing WRF-Hydro with NLDAS data. I am using standard regridding scripts and standard NLDASv2 data. My domain is large and has ocean grids (attached). The model crashes after few timesteps showing "FATAL NoahMP error". However, it runs fine with the idealized forcing. The regridded data has missing values (see below), which comes from original NLDAS2 files. I am not sure if that, or unrealistic SWdown values might be the reason (since the model runs for few timesteps and then crashes).

cdo infon 1996010114.LDASIN_DOMAIN1

 -1 :       Date     Time   Level Gridsize    Miss :     Minimum        Mean     Maximum : Parameter name

     1 : 1996-01-01 14:00:00       0   113176       0 :      35.589      42.100      48.297 : lat

     2 : 1996-01-01 14:00:00       0   113176       0 :     -85.193     -75.498     -67.080 : lon

     3 : 1996-01-01 14:00:00       0   113176   26134 :      247.48      268.38      284.88 : T2D

     4 : 1996-01-01 14:00:00       0   113176   26134 :  0.00038680   0.0027716   0.0079144 : Q2D

     5 : 1996-01-01 14:00:00       0   113176   26134 :     -9.3700     -1.4335      5.7250 : U2D

     6 : 1996-01-01 14:00:00       0   113176   26134 :     -8.3700     -2.3963      2.7400 : V2D

     7 : 1996-01-01 14:00:00       0   113176   26134 :      86267.      98234.  1.0195e+05 : PSFC

     8 : 1996-01-01 14:00:00       0   113176   26134 :      0.0000  4.6120e-07  3.8611e-05 : RAINRATE

     9 : 1996-01-01 14:00:00       0   113176   26134 :      24.240      111.09      210.35 : SWDOWN

    10 : 1996-01-01 14:00:00       0   113176   26134 :      136.12      262.89      367.79 : LWDOWN

Is it a common problem? Do I need to modify the regridding script?

Thanks,

Sujan

[Molly McAllister]

Hi Sujan, 

The model will crash if there are missing forcing data values. 

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.
To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/1761ebc9-d840-4b79-a8ea-d9e94454fb9bn%40ucar.edu.

--

------------------------------------------------------------------------------------------------------

My working day may not be your working day. Please do not feel obliged to reply to this email outside of your normal working hours.
------------------------------------------------------------------------------------------------------

Molly J. McAllister M.S. 

Associate Scientist III

National Center for Atmospheric Research

moll...@ucar.edu
Cell: 303 579 3335 - if need to contact via voice 

[Md Murad Hossain Khondaker]

Hi Sujan,

Did you find any solution to this problem? I am facing exactly same issue. My domain is also large and has some ocean grids. Every time after some timesteps, NOAH MP stops and the following error is shown.

VEGETATION!

 SWDOWN*FVEG =  2.1615262E+19

 FVEG*(SAV+SAG) =  8.2179109E+19

 FVEG*(FSRV +FSRG)=  9.0210987E+18

 GROUND!

 (1-.FVEG)*SWDOWN =  7.8384738E+19

 (1.-FVEG)*SAG =  6.3778711E+19

 (1.-FVEG)*FSRG=  5.8062383E+18

 FSRV   =  7.4657466E+18

 FSRG   =  7.4073582E+18

 FSR    =  1.4873105E+19

 SAV    =  3.7606626E+18

 SAG    =  8.1366235E+19

 FSA    =  8.5126891E+19

  ERRSW =  8.7960930E+12

===================================================================================

=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES

=   RANK 48 PID 12584 RUNNING AT compute-6-5

=   KILLED BY SIGNAL: 9 (Killed)

===================================================================================

Thanks in advvance.

Best, 

Murad 

[Sujan Pal]

Yes, my problem was due to the missing values present after regridding. I added simple lines in the NLDAS regridding script to replace the missing values with real values for all the variables before saving it into the final netcdf files. 

For example, I replaced any missing values of precipitation with 0:

else if (v.eq.6) then

    v4 = where(ismissing(v4),0,v4)

    ncdf->RAINRATE = v4

Hope this helps. 

Sujan

[Md Murad Hossain Khondaker]

Hi Sujan,

Thank you very much. My issues have been resolved thanks to your information. This helped me a lot.

Best,

Murad.

[Orzu Odinaev]

Dear Sujan, 

where exactly in the script did you add this code?

Regards,

Orzu

[Sujan Pal]

Hi Orzu,

Disregard the previous email. You can add these near Lines 230-245 of the code in the NLDAS regridding script (NLDAS2WRFHydro_regrid.ncl). Hope this helps.

Best,

Sujan