Multiple LBC files for ARW, run-time option

[davegill]

TYPE: enhancement

KEYWORDS: LBC files

SOURCE: Kevin Manning and Jim Bresch (MMM/NCAR), internal

DESCRIPTION OF CHANGES:
For real-time users, the ability to start the WRF model before the
GFS model finishes allows the WRF model to complete a regional
simulation earlier (because of not delaying the model start). However, this
only works if the lateral boundary files from subsequent times are able to be
ingested by the model. As a compile-time option, this split LBC capability has
existed for years. Now that capability from the compile-time option is run-time
selectable from the namelist.

1. WRF model

Modify the compile-time option in WRF that provides the capability for split
LBC files, and now make that capability a namelist-driven run-time option.
By default, the multi_bdy_files option is FALSE.

2. real program

The user may now set the namelist.input file as follows in real:

 &time_control
 start_year                          = 2000, 2000, 2000,
 start_month                         = 01,   01,   01,
 start_day                           = 24,   24,   24,
 start_hour                          = 12,   12,   12,
 end_year                            = 2000, 2000, 2000,
 end_month                           = 01,   01,   01,
 end_day                             = 25,   25,   25,
 end_hour                            = 12,   12,   12,
 interval_seconds                    = 21600
 bdy_inname                          = "wrfbdy_d<domain>_<date>"
 /

 &bdy_control
 multi_bdy_files                     = .true.
 /

and receive the following files with a single run of the real program.

wrfbdy_d01_2000-01-24_12:00:00
wrfbdy_d01_2000-01-24_18:00:00
wrfbdy_d01_2000-01-25_00:00:00
wrfbdy_d01_2000-01-25_06:00:00

ISSUE:
Fixes #656 "Multi-file lateral boundary conditions"

LIST OF MODIFIED FILES:
modified: Registry/Registry.EM_COMMON
modified: main/real_em.F
modified: run/README.namelist
modified: share/mediation_integrate.F
modified: share/module_check_a_mundo.F
modified: test/em_real/examples.namelist

TESTS CONDUCTED:

1. Checked 24-h Jan 2000 simulation with SPLIT LBC vs traditional LBC: bit-for-bit results.
2. The lateral boundary files and the initial condition files are bit-wise identical whether the
   the original two-time-levels of multiple runs of the real program were used or if the new
   single-run of the real program is used.
3. Tried inconsistent namelist settings. They are detected:

 &time_control
 blah 
 /
 bdy_inname = "wrfbdy_d<domain>_<date>"

 &bdy_control
 multi_bdy_files = .true.
 /

Yields:

  --- ERROR: Need bdy_inname = "wrfbdy_d<domain>_<date>"
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE:  <stdin>  LINE:    2035
NOTE:       1 namelist settings are wrong. Please check and reset these options
-------------------------------------------

And the other way:

 &time_control
 bdy_inname = "wrfbdy_d<domain>_<date>"
 /

 &bdy_control
 blah
 /
 multi_bdy_files = .true.

Yields:

  --- ERROR: Remove bdy_inname = "wrfbdy_d<domain>_<date>"
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE:  <stdin>  LINE:    2035
NOTE:       1 namelist settings are wrong. Please check and reset these options
-------------------------------------------

4. Code benignly builds with NMM.

RELEASE NOTE: A run-time capability for multiple LBC files is now available. This supersedes and replaces the compile-time option. This is run-time option is accessed through &bdy_control namelist logical variable multi_bdy_files. This option requires that the lateral boundary file names in &time_control include a date: bdy_inname = "wrfbdy_d<domain>_<date>". All other functionality of the capability is identical to the compile-time option.

[davegill]

@jimbresch @kwman44
Folks,
Would you please take a look at this PR, perhaps test it out. The substantive changes are only in the Registry and mediation_integrate. You could jam those mods into your systems.

[davegill]

@weiwangncar @jimbresch @kwman44 @dudhia
Folks,
Per some prodding from Jimy yesterday, I have updated this PR to include mods to the real program. Now with a single real program run, you get all of the split files.

[davegill]

@weiwangncar @jimbresch @kwman44 @dudhia
Folks,
Also: vote on your preferred name for this option:

multi_bdy_files
multi_bc_files

Or any other suggestion would be welcome.

[dudhia]

another option time_split_wrfbdy

…

On Thu, Sep 26, 2019 at 6:35 PM Dave Gill ***@***.***> wrote: @weiwangncar <https://github.com/weiwangncar> @jimbresch <https://github.com/jimbresch> @kwman44 <https://github.com/kwman44> @dudhia <https://github.com/dudhia> Folks, Also: vote on your preferred name for this option: multi_bdy_files multi_bc_files Or any other suggestion would be welcome. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#990?email_source=notifications&email_token=AEIZ77HFVXVKUDTJXOWQ3MDQLVILRA5CNFSM4I2TXC4KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD7XKTGA#issuecomment-535734680>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEIZ77DZVEKWDL2YA3J3GFTQLVILRANCNFSM4I2TXC4A> .

[davegill]

Also, regtest results:

Tue Oct 22 13:36:31 MDT 2019
SUCCESS_RUN_WRF_d01_em_real_32_em_real_03FD vs SUCCESS_RUN_WRF_d01_em_real_33_em_real_03FD status = 0
SUCCESS_RUN_WRF_d01_em_real_32_em_real_03FD vs SUCCESS_RUN_WRF_d01_em_real_34_em_real_03FD status = 0
SUCCESS_RUN_WRF_d01_em_real_32_em_real_10 vs SUCCESS_RUN_WRF_d01_em_real_33_em_real_10 status = 0
SUCCESS_RUN_WRF_d01_em_real_32_em_real_10 vs SUCCESS_RUN_WRF_d01_em_real_34_em_real_10 status = 0
SUCCESS_RUN_WRF_d01_em_real_32_em_real_11 vs SUCCESS_RUN_WRF_d01_em_real_33_em_real_11 status = 0
SUCCESS_RUN_WRF_d01_em_real_32_em_real_11 vs SUCCESS_RUN_WRF_d01_em_real_34_em_real_11 status = 0
SUCCESS_RUN_WRF_d01_em_real_32_em_real_20NE vs SUCCESS_RUN_WRF_d01_em_real_33_em_real_20NE status = 0
SUCCESS_RUN_WRF_d01_em_real_32_em_real_20NE vs SUCCESS_RUN_WRF_d01_em_real_34_em_real_20NE status = 0
SUCCESS_RUN_WRF_d01_nmm_real_32_nmm_nest_01 vs SUCCESS_RUN_WRF_d01_nmm_real_34_nmm_nest_01 status = 0
SUCCESS_RUN_WRF_d01_nmm_real_32_nmm_nest_03 vs SUCCESS_RUN_WRF_d01_nmm_real_34_nmm_nest_03 status = 0
SUCCESS_RUN_WRF_d01_nmm_real_32_nmm_nest_04a vs SUCCESS_RUN_WRF_d01_nmm_real_34_nmm_nest_04a status = 0
SUCCESS_RUN_WRF_d01_nmm_real_32_nmm_nest_06 vs SUCCESS_RUN_WRF_d01_nmm_real_34_nmm_nest_06 status = 0
SUCCESS_RUN_WRF_d01_em_real_32_em_chem_1 vs SUCCESS_RUN_WRF_d01_em_real_34_em_chem_1 status = 0
SUCCESS_RUN_WRF_d01_em_real_32_em_chem_2 vs SUCCESS_RUN_WRF_d01_em_real_34_em_chem_2 status = 0
SUCCESS_RUN_WRF_d01_em_real_32_em_chem_5 vs SUCCESS_RUN_WRF_d01_em_real_34_em_chem_5 status = 0
Tue Oct 22 13:36:31 MDT 2019