clm5-rtm-wiso is answer changing compared to the parent clm tag with wiso off.

bandre-ucar commented 8 years ago

Summary of Issue:

I tried running the clm_short test suite and comparing against the parent clm tag to verify that our code isn't changing answers for clm5 default with water isotopes off. It changes answers.

What is the changeset ID of the code, and the machine you are using:

svn tag: geotrace3_cac_n79_cam5_4_80

Expected behavior and actual behavior:

Expected behavior: wiso branches should be bit for bit with the upstream component tag when wiso is turned off.

Actual behavior: clm changes answers in the runoff fields received from rtm: VOLR, TWS, VOLRMCH. Related coupler fields are also different, see below. This is probably a problem with RTM rather than CLM.

Steps to reproduce the problem (should include create_newcase or create_test command along with any user_nl or xml changes):

Clean checkout of the above svn tag.
run the tests:

execca ./create_test -xml_category clm_short_45 -mach yellowstone -compiler intel -xml_mach yellowstone -xml_compiler intel -compare clm4_5_8_r182 

execca ./create_test -xml_category clm_short_45 -mach yellowstone -compiler pgi -xml_mach yellowstone -xml_compiler pgi -compare clm4_5_8_r182 

execca ./create_test -xml_category clm_short_45 -mach yellowstone -compiler gnu -xml_mach yellowstone -xml_compiler gnu -compare clm4_5_8_r182 

execca ./create_test -xml_category clm_short_50 -mach yellowstone -compiler intel -xml_mach yellowstone -xml_compiler intel -compare clm4_5_8_r182 

execca ./create_test -xml_category clm_short_50 -mach yellowstone -compiler pgi -xml_mach yellowstone -xml_compiler pgi -compare clm4_5_8_r182 

execca ./create_test -xml_category clm_short_50 -mach yellowstone -compiler gnu -xml_mach yellowstone -xml_compiler gnu -compare clm4_5_8_r182

have you modified the code? If so, it must be committed and available for testing:

no

Screen output or output files showing the error message and context:

---summarizing more details of test failures if any: --- 

/glade/scratch/andre/SMS_D_Ld3.f10_f10.ICRUCLM50BGC.yellowstone_pgi.clm-default.
C.09201512-50p/run/SMS_D_Ld3.f10_f10.ICRUCLM50BGC.yellowstone_pgi.clm-default.C.
09201512-50p.clm2.h0.0001-01-04-00000.nc.base.cprnc.out had the following fields
 that are NOT b4b  

   RMS TWS                              5.4548E-14            NORMALIZED  2.9907E-18
    RMS VOLR                             2.4724E-05            NORMALIZED  3.5212E-15
    RMS VOLRMCH                          2.4724E-05            NORMALIZED  3.5212E-15
 /glade/scratch/andre/SMS_D_Ld3.f10_f10.ICRUCLM50BGC.yellowstone_pgi.clm-default.C.09201512-50p/run/SMS_D_Ld3.f10_f10.ICRUCLM50BGC.yellowstone_pgi.clm-default.C.09201512-50p.cpl.hi.0001-01-04-00000.nc.base.cprnc.out had the following fields that are NOT b4b  

   RMS x2l_Flrr_volr                    1.7791E-17            NORMALIZED  4.4667E-15
    RMS x2l_Flrr_volrmch                 1.7791E-17            NORMALIZED  4.4667E-15

bandre-ucar commented 8 years ago

As a work around for clm5-wiso testing, I am generating new baselines from the geotrace tags that include these differences. Current baseline is: geotrace3-n79

bandre-ucar commented 8 years ago

Note, this is the same RTM problem we've discussed multiple times. I'm just documenting it.

CESM-Development / water-isotopes