[rrfs-mpas-jedi]Hotfix for continuous surface cycling

[chunhuazhou]

DESCRIPTION OF CHANGES:

Problem: Previous failed tasks cause cycling to pause in RRFSv2X.
Solution: This hotfix facilitates continuous surface cycling by:

• Adding a 2-hour time dependency for prep_ic.
• Outputting more mpasout files to allow the system to look back further for required cycling data.

TESTS CONDUCTED:

These changes are applied in RRFSv2X.

[guoqing-noaa]

To confirm, this PR would like to use the latest 1,2,3hrs forecasts for surface cycling if possible; if we can not get any of them in 3 hours, the prep_ic task will start anyway but then use a longer forecast (15 or 27h). Do I get it right?

+        <timedep><cyclestr offset="03:00:00">@Y@m@d@H@M00</cyclestr></timedep>
         <datadep age="00:00:05"><cyclestr offset="-1:00:00">&COMROOT;/&NET;/&rrfs_ver;/&RUN;.@Y@m@d/@H/fcst/&WGF;/fcst_f001.done</cyclestr></datadep>
         <datadep age="00:00:05"><cyclestr offset="-2:00:00">&COMROOT;/&NET;/&rrfs_ver;/&RUN;.@Y@m@d/@H/fcst/&WGF;/fcst_f002.done</cyclestr></datadep>
         <datadep age="00:00:05"><cyclestr offset="-3:00:00">&COMROOT;/&NET;/&rrfs_ver;/&RUN;.@Y@m@d/@H/fcst/&WGF;/fcst_f003.done</cyclestr></datadep>

[chunhuazhou]

To confirm, this PR would like to use the latest 1,2,3hrs forecasts for surface cycling if possible; if we can not get any of them in 3 hours, the prep_ic task will start anyway but then use a longer forecast (15 or 27h). Do I get it right?

+        <timedep><cyclestr offset="03:00:00">@Y@m@d@H@M00</cyclestr></timedep>
         <datadep age="00:00:05"><cyclestr offset="-1:00:00">&COMROOT;/&NET;/&rrfs_ver;/&RUN;.@Y@m@d/@H/fcst/&WGF;/fcst_f001.done</cyclestr></datadep>
         <datadep age="00:00:05"><cyclestr offset="-2:00:00">&COMROOT;/&NET;/&rrfs_ver;/&RUN;.@Y@m@d/@H/fcst/&WGF;/fcst_f002.done</cyclestr></datadep>
         <datadep age="00:00:05"><cyclestr offset="-3:00:00">&COMROOT;/&NET;/&rrfs_ver;/&RUN;.@Y@m@d/@H/fcst/&WGF;/fcst_f003.done</cyclestr></datadep>

Yes, that is right. Thanks!

[guoqing-noaa]

LGTM. Thanks, @chunhuazhou