chillwx@gmail_com
New Member
I am running ensembles of the 0.25° CESM using the F compset. Recently, one ensemble has failed (it initialized fine and ran for 5 days out of a 10 day simulation) with the following error message in the ccsm.log file:
237: clh GAMMA: NaNQ 0.000000000000000000E+00 0.000000000000000000E+00 NaN Q 1.00000000000000000 0.000000000000000000E+00
I have traced this to the CLM VOCEmissionMod.F90 subroutine (line 618 in the CESM1_3 release). A few lines above that is a mention in the comments stating:
! fvitt -- forc_solad240, forc_solai240 can be zero when CLM finidat is specified
! which will cause par240 to be zero and produce NaNs via log(par240)
! dml -- fsun240 can be equal to or greater than one before 10 day averages are
! set on startup or if a new pft comes online during land cover change.
! Avoid this problem by only doing calculations with fsun240 when fsun240 is
! between 0 and 1
Has anyone seen this kind of problem before and are there any solutions? I am surprised that the model initializes and runs before crashing.
237: clh GAMMA: NaNQ 0.000000000000000000E+00 0.000000000000000000E+00 NaN Q 1.00000000000000000 0.000000000000000000E+00
I have traced this to the CLM VOCEmissionMod.F90 subroutine (line 618 in the CESM1_3 release). A few lines above that is a mention in the comments stating:
! fvitt -- forc_solad240, forc_solai240 can be zero when CLM finidat is specified
! which will cause par240 to be zero and produce NaNs via log(par240)
! dml -- fsun240 can be equal to or greater than one before 10 day averages are
! set on startup or if a new pft comes online during land cover change.
! Avoid this problem by only doing calculations with fsun240 when fsun240 is
! between 0 and 1
Has anyone seen this kind of problem before and are there any solutions? I am surprised that the model initializes and runs before crashing.