Main menu


Sudden Crash Due to ErrH2OSno

7 posts / 0 new
Last post
Sudden Crash Due to ErrH2OSno

In December of Year 46 of a CESM 1.2.2 slab ocean run with horizontal fluxes set to zero: (1850_CAM5_CLM45_CICE_DOCN%SOM_RTM_SGLC_SWAV), I get CLM crashing because of issues with snow balance. In particular, I get this in the CESM log:


WARNING:  snow balance error  nstep =       804117  indexc=         5230 ltype:

             6  errh2osno=   0.213626229745594   


clm model is stopping - error is greater than .10

 nstep =       804117  indexc=         5230  errh2osno=   0.213626229745594     

 ltype:            6

 ctype(indexc):           65

 snl:           -1

 h2osno:    1.34031126790802     

 h2osno_old:    1.39951518049922     

 snow_sources:   1.884662632477202E-008

 snow_sinks:   1.515911479245435E-004

 qflx_prec_grnd:   3.392392738458963E-005

 qflx_sub_snow:   2.217117946664299E-002

 qflx_evap_grnd:   3.706665705194147E-002

 qflx_top_soil:   0.000000000000000E+000

 qflx_dew_snow:   0.000000000000000E+000

 qflx_dew_grnd:   0.000000000000000E+000

 qflx_snwcp_ice:   0.213626229745594     

 qflx_snwcp_liq:   0.000000000000000E+000

 qflx_sl_top_soil:   0.000000000000000E+000

 clm model is stopping

 Does anyone know the significance of "snow capping"? And why there would be sudden dump of ice in a single timestep?   Nicholas Heavens Research Assistant Professor of Planetary ScienceHampton University 




This looks similar to a known bug.  The ice "dumped" is actually ponded ice on top of the soil exceeding a prescribed limit and it shouldn't be included in the qflx_snwcp_ice.  See here for a description of the bug:



This was fixed in clm4_5_6_r156.  I'm not sure what version you are using.  If you have access to the cesm respository, you could look at the change log and see what changes were made to the code.  However, the simplest thing to do might be to comment out the endrun in BalanceCheck temporarily to get past the error.


Thanks. I don't have access to the repository, but I did find this note on a listhost archive:


    William Sacks 2015-11-24 - clm4_5_6_r156 - components/clm (cesm1_5_alpha03e)

    Fix QSNWCPICE and glacial inception

    Changes diagnostic fields for clm45/clm50 compsets

    Potentially significant changes for IG compsets with clm45/clm50

That's all I really need to know.


From the error log the landtype is actually urban landunit--pervious landtype,does this bug also apply to urban landunit?


The ctype is 65, which is urban pervious road which gets treated the same as soil with respect to ice here.


I met the same problem in my simulation, where ctype=65 often comes with warning or even imbalance while have not met any with ctype=1. I was wondering what are the differences between ctype=65 and ctype=1 that may cause this problem?


I can't think of any differences that might cause that other than different forcing because the pervious road is within the urban canyon.

Log in or register to post comments

Who's new

  • g.mairiam@...
  • 408251463@...
  • 1161561668@...
  • 645505358@...
  • lbqnj@...