ERROR initInterp set_mindist: Cannot find any input points matching output point:

[Yuan Sun]

Hi all,

I am developing input urban data with more than 3 numurbl and stopped by an error in CTSMdev166 with compset: 2000_DATM%GSWP3v1_CLM51%SP_SICE_SOCN_MOSART_SGLC_SWAV_SESP

[5] ERROR initInterp set_mindist: Cannot find any input points matching output point:
[5] subgrid level, index = pft 20085
[5] lat, lon = -0.40297916237146575 , 5.3887040655324920
[5] ltype: 10
[5] ctype: 71
[5] ptype: 0
[5]
[5] Consider rerunning with the following in user_nl_clm:
[5] init_interp_fill_missing_with_natveg = .true.
[5] However, note that this will fill all missing types in the output
[5] with the closest natural veg column in the input
[5] (using bare soil for patch-level variables).
[5] So, you should consider whether that is what you want.
[5] ENDRUN:
[5] ERROR: ERROR in /home/yuansun/CESM/ctsm5.1.dev166lcz/src/init_interp/initInterpMindist.F90 at line 296

In my understanding, the ltype:10 comes from my modified code with 10 numurbl so than the ltype10-16 has no data in the standard initial clm input: lnd/clm2/initdata_map/clmi.I2000Clm50BgcCrop.2011-01-01.1.9x2.5_gx1v7_gl4_simyr2000_c190312.nc. ; ctype: 71 indicates a roof column.

I tried 'init_interp_fill_missing_with_natveg = .true.' but it leads to another error : urban inside roof energy balance error 1.9369247669649464E+037.

Thus, to skipp above errors, i tried './xmlchange CLM_FORCE_COLDSTART=on', the run failed after initialization without error recorded. I am not sure if i should blame the mediator this time.

Thanks if any comments.

Best,
Yuan

 

[oleson]

I would highly recommend starting with something simpler, like an urban single point cold start simulation. It's going to be much easier to debug.
The initial file interpolation routine isn't going to be able to handle 10 urban types without modification and init_interp_fill_missing_with_natveg is going to try to map vegetation to urban columns which is not going to work as you've seen.

 

[Yuan Sun]

Hi Keith,

Thanks for your suggestion. I will try single-point as soon as possible.

Best,
Yuan

 

[Yuan Sun]

Hi,

I found single-point run successfully but global run stopped with two log lines in cesm.log:

[4] layer, h2osoi_liq: 22819 10 -1.3954353248627895E-006
[3] layer, h2osoi_liq: 29107 10 -1.3954353248627895E-006

Since no direct error reported, i am not sure if the negative h2osoi_liq stopped the run because in the SoilWaterMovementMod.F90, line 1421:
if(h2osoi_liq(c,j) < -1e-6_r8) then
write(*,*) 'layer, h2osoi_liq: ', c,j,h2osoi_liq(c,j)
! call endrun(subname // ':: negative soil moisture values found!')
endif

But the 'call endrun' routine is commented in the latest CTSM code. Did anywhere end the run ?

Another concern is the 'drv.log' is blank in the global run.

Best,
Yuan

 

[Yuan Sun]

Hi, just updated that it is ok if h2osoi_liq(c,j) is negative, as the code set. My previous run stop came from the output settings. It works now.

Best,
Yuan