RIBX never exceeds RIC

[Lily_Zh]

Hello,

I am trying to use EPA's 12US WRF configurations using ACM PBL. I used WPS3.8 and WRF3.8. North American Model (12NAM) analysis product, NCLD 2011 landuse data (nlcd2011_ll_9s), get wrfsfdda using OBSGRID with Upper-air data in NCEP BUFR format (ds351.0) and urface data in NCEP BUFR format (ds461.0).
I know this error is related with ACM PBL, I am not sure if other options in namelist.input lead to this error and no those information from EPA wrf model evaluation document.
I also tried to use WPS4.0 and WRF3.8, or without wrfsfdda, or modis_landuse_20class_30s, or NARR data, but still exist below error:

Timing for main: time 2017-06-20_00:01:00 on domain 1: 766.94489 elapsed seconds
d01 2017-06-20_00:01:00 in med_latbound_in
d01 2017-06-20_00:01:00 module_integrate: calling solve interface
d01 2017-06-20_00:01:00 call rk_step_prep
d01 2017-06-20_00:01:00 calling inc/HALO_EM_A_inline.inc
d01 2017-06-20_00:01:00 calling inc/PERIOD_BDY_EM_A_inline.inc
d01 2017-06-20_00:01:00 call rk_phys_bc_dry_1
d01 2017-06-20_00:01:00 call init_zero_tendency
d01 2017-06-20_00:01:00 calling inc/HALO_EM_PHYS_A_inline.inc
d01 2017-06-20_00:01:00 call phy_prep
d01 2017-06-20_00:01:00 DEBUG wrf_timetoa(): returning with str = [2017-06-20_00:01:00]
d01 2017-06-20_00:01:00 call radiation_driver
d01 2017-06-20_00:01:00 Top of Radiation Driver
d01 2017-06-20_00:01:00 calling inc/HALO_PWP_inline.inc
d01 2017-06-20_00:01:00 call surface_driver
d01 2017-06-20_00:01:00 in PX Surface Layer scheme
d01 2017-06-20_00:01:00 in P-X LSM
d01 2017-06-20_00:01:00 call pbl_driver
d01 2017-06-20_00:01:00 in ACM PBL
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: <stdin> LINE: 436
RIBX never exceeds RIC, RIB(i,kte) = NaN THETAV(i,1) = NaN MOL= NaN TCONV = 0.1084705 WST = NaN KMIX = 2 UST = NaN TST = NaN U,V = NaN NaN I,J= 120 2
-------------------------------------------
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
[unset]: readline failed

Namelist is attached, if you need any more info., let me know.

Thanks in advance.

Lily

 

[Ming Chen]

Hi Lily,
(1) I looked at your nameless.input and everything looks fine except that num_land_cat = 21. if you run with NLCD data, I suppose num_land_cat should be 40.
(2) This case crashed almost immediately, which often indicates that the input data is wrong.
(3) Can you try to run the same case but without OBS nudging and see whether it works? If not, then we need to
(4) run the case with MODIS landuse data as input and see whether it works.

We need to figure out what could be the possible reason that leads to wrong input data.

 

[Lily_Zh]

Hi Ming,

Thank you for your reply and suggestions. Please see my reply below.

(1) I looked at your nameless.input and everything looks fine except that num_land_cat = 21. if you run with NLCD data, I suppose num_land_cat should be 40.
L: I used WPS_GEOG data (should be old version) from our group member. I just downloaded latest NLCD on wrf website to run geogird.exe, still 21 in geo_em.d01.nc
MAP_PROJ = 1 ;
:MMINLU = "MODIFIED_IGBP_MODIS_NOAH" ;
:NUM_LAND_CAT = 21 ;
:ISWATER = 17 ;
:ISLAKE = 21 ;
:ISICE = 15 ;
:ISURBAN = 13 ;
:ISOILWATER = 14 ; ).

I checked the index in the folder of nlcd2011_ll_9s, category_max=40, but geogrid.log still shows that:
Processing domain 1 of 1
2021-08-02 18:37:36.862 --- INFORM: For HGT_M, couldn't find interpolator sequence for resolution nlcd2011_ll_9s.
2021-08-02 18:37:36.862 --- INFORM: Using default interpolator sequence for HGT_M.
2021-08-02 18:37:36.863 --- INFORM: For HGT_M, couldn't find nlcd2011_ll_9s data source.
2021-08-02 18:37:36.863 --- INFORM: Using default data source for HGT_M.
2021-08-02 18:37:36.887 --- INFORM: For LANDUSEF, couldn't find interpolator sequence for resolution nlcd2011_ll_9s.
2021-08-02 18:37:36.888 --- INFORM: Using default interpolator sequence for LANDUSEF.
......
......
......
2021-08-02 18:37:37.436 --- INFORM: Using default data source for CANFRA.
2021-08-02 18:37:38.726 --- Processing field 1 of 28 (LANDUSEF)
......


(2) This case crashed almost immediately, which often indicates that the input data is wrong.
L: I see. I will focus on testing diff input.

(3) Can you try to run the same case but without OBS nudging and see whether it works? If not, then we need to
(4) run the case with MODIS landuse data as input and see whether it works.
L:Yes, I tested without wrfsfdda before, and also tested modis_landuse_20class_30s before, same error, I will try it again using this namelist.input.

Below is my namelist.wps.

&share
wrf_core = 'ARW',
max_dom = 1,
start_date = '2017-06-20_00:00:00','2017-06-24_00:00:00','2017-06-24_00:00:00',
end_date = '2017-06-22_00:00:00','2017-08-01_00:00:00','2017-08-01_00:00:00',
interval_seconds = 10800
io_form_geogrid = 2,
/

&geogrid
parent_id = 1, 1, 2,
parent_grid_ratio = 1, 6, 6,
i_parent_start = 1, 118, 33,
j_parent_start = 1, 55, 39,
e_we = 476, 97, 91,
e_sn = 316, 109, 91,
geog_data_res = 'nlcd2011_ll_9s',
dx = 12000,
dy = 12000,
map_proj = 'lambert',
ref_lat = 40.0,
ref_lon = -97.0,
truelat1 = 33.0,
truelat2 = 45.0,
stand_lon = -97.0,
geog_data_path = '/scratch/.../WPS_GEOG'
/

&ungrib
out_format = 'WPS',
prefix = 'FILE',
/

&metgrid
fg_name = 'FILE'
io_form_metgrid = 2,
/

Thank you.

Lily

 

[Lily_Zh]

Hi Ming,

I checked geogrid/GEOGRID.TBL and knew I used wrong name of nlcd2011 in namelist.wps. I modfied it to be geog_data_res = 'nlcd2011_9s+default'
instead of nlcd_2011_ll_9s. It can generate output with MMINLU = "NLCD40". WRF did not stop with that error at the beginning now.

Thank you for your suggestions.

Libo

 

[Ming Chen]

Libo,
Thanks for letting me know. It is great that you can fix the issue and make it work.