CH4 streams file has to use a fixed date or results in roundoff exact restart problems

[ekluzek]

Brief summary of bug

Rather than using the current date to advance the stream data for the CH4 finundation stream a fixed date is being used because when the date varies exact restarts can cause roundoff differences. This is even though the streams file has a single time sample and uses

General bug information

CTSM version you are using: clm4_5_16_r242

Does this bug cause significantly incorrect results in the model's science? No

Configurations affected: BGC simulations

Details of bug

Here's the FORTRAN code where use of the current date is commented out and a set date is used.

         ! Explicitly set current date to a hardcoded constant value. Otherwise
         ! using the real date can cause roundoff differences that are
         ! detrected as issues with exact restart.  EBK M05/20/2017
         ! call get_curr_date(year, mon, day, sec)
         year = 1996
         mon  = 12
         day  = 31
         sec  = 0
         mcdate = year*10000 + mon*100 + day