Public Git Hosting - WRF-SFIRE.git/log

Most of the make dependency issues happened with using more than 4 job
slots at once.  Re-organized the order of compilation of objects in
dyn_nmm, share, chem to allow as more job slots as possible and
limited the maximum jobs slots for share for 6 job slots,  dyn_nmm
and chem for 16 job slots to avoid make dependency issues if specified
higher job slots;  Using an include entry instead of the dependency list
in Makefile,  separated dependency list to depend.chem for chem,
depend.dyn_nmm for dyn_nmm, depend.dyn_em for dyn_em and combined share,
frame, phys and main to depend.common.

A       chem/depend.chem
M       chem/Makefile
A       dyn_em/depend.dyn_em
M       dyn_em/Makefile
A       dyn_nmm/depend.dyn_nmm
M       dyn_nmm/Makefile
M       frame/Makefile
A       main/depend.common
M       main/Makefile
M       phys/Makefile
M       share/Makefile
M       Makefile

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6069 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

From polar folks at BPRC, an option to use sea-ice thickness field
instead of a constant 3-m sea ice thickness.

This entails a new 2D array, ICEDEPTH, to be read from the input
files, and a new namelist variable, SEAICE_THICKNESS_OPT, to allow the
user to use the ICEDEPTH field from the input
(SEAICE_THICKNESS_OPT=1), or use the default value of 3 m
(SEAICE_THICKNESS_OPT=0).

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6067 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

From the polar folks at BPRC, remove a limitation to the depth of snow
on sea ice as used in the soil heat flux calculations.

Add some comments for some suggestions regarding seaice albedo
depending on season and locale.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6066 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Adjustments to seaice_albedo_opt=1 functionality submitted by polar
folks at Byrd Polar Research Center. The snow-cover adjustment only
applies to the albedo calculation; albedos as a function of T adjusted
to avoid discontinuities.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6065 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Changes from Mansell for his NSSL MP scheme:
1. add two single-moment schemes, 19 and 20
2. bug fixes for two-moment scheme

M    phys/module_mp_nssl_2mom.F
M    phys/module_microphysics_driver.F
M    phys/module_physics_init.F
M    run/README.namelist
M    Registry/Registry.EM_COMMON
M    dyn_em/start_em.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6062 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

change from Mansell to reduce epsilon to 1e-28 from 1e-18

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6061 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Fix OpenMP trouble, caused by array-styled initialization.
Trouble was the array was MEMORY dimensioned.  The assignment
in a physics routine is only allowed to be TILE dimensioned,
which makes it thread safe.  There were multiple processors
re-assigning the entire array.  Found and fixed by John Michalakes

M    module_bl_ysu.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6053 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

With nml option nest_interp_coord (0=eta, 1=isobaric as possible),
horizontally interpolate the fine grid fields using the CG data.
Useful when initializing WRF with
input_from_file   = .t.,.t.
fine_input_stream =  0 , 2
Only T and PH are modified.

M    external/RSL_LITE/module_dm.F
M    share/mediation_interp_domain.F
M    share/Makefile
M    share/interp_fcn.F
M    Registry/Registry.EM_COMMON
M    dyn_em/module_initialize_real.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6052 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Faster and better netcdf4 compression!
1) Reduce the optimization level (so it is faster). But
reducing the optimization from 3 (of 9) down to 2 does not
really make the files much larger.
2) Add in the chuffle feature from HDF - which is almost a
zero cost item, AND it is really good at compressing fields
that are data with similar values.

M wrf_io.F90

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6051 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Users may now change the name of the lateral boundary file
in the namelist.input file. Why would they want to? Who
knows.

M mediation_integrate.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6050 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

SKEBS forcing on the lateral boundary conditions.  Thanks to
Kate for this one.

M    share/module_bc.F
M    dyn_em/module_bc_em.F
M    share/module_check_a_mundo.F
A    Registry/registry.bdy_perturb
M    Registry/Registry.EM
M    Registry/Registry.EM_CHEM
M    test/em_real/examples.namelist
M    run/README.namelist
M    dyn_em/solve_em.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6049 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Fix so that the SKEBS random numbers are reproducible,
regardless of processor count.  Thanks to Kate and Judith.

M    module_stoch.F
M    module_first_rk_step_part2.F
M    start_em.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6048 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Maybe temporary for the the next couple of weeks, but just a
simple way for folks to use the yellowstone machine with the
Intel and the PGI compilers.

M configure_new.defaults

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6047 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Fujitsu compiler directives begin with the string OCL (or ocl).
So let's make sure that we do not strip off those characters
when manufacturing our .f90 files.

M standard.c

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6046 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Still lingering instances of the
IF ( PRESENT(blah) .and.  blah.EQ.foo ) THEN
These are turned into:
IF ( PRESENT(blah) ) THEN
   IF ( blah.EQ.foo ) THEN

Thanks to the Fujitsu folks, via Jim Bresch.

M    share/module_llxy.F
M    external/esmf_time_f90/ESMF_Base.F90

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6045 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1. Fix for Gfortran compilation options on Mac OSX

Removed unsupported option -ftree-loop-linear of Gfortrann for Mac OSX

M arch/configure_new.defaults

2. Fix for netcdf lib testing issue on Mac OSX

Fixed the regular expression for grep for netcdf lib testing on Mac
OSX

M configure

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6042 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Two sets of changes to the WRFDA portion of the trunk. The first item allows assimilation of a new instrument, while the second is a minor change/bug fix.

1. WRFDA user Peiming Dong has provided additions to the WRFDA code which allow for assimilation of data from the ATMS instrument on the NPP Suomi satellite:
M       var/da/da_setup_structures/da_setup_obs_structures.inc
M       var/da/da_setup_structures/da_setup_structures.f90
M       var/da/da_radiance/da_radiance.f90
M       var/da/da_radiance/da_radiance1.f90
M       var/da/da_radiance/da_qc_rad.inc
M       var/da/da_radiance/da_radiance_init.inc
M       var/da/da_radiance/module_radiance.f90
M       var/da/da_radiance/da_read_obs_bufrtovs.inc
A       var/da/da_radiance/da_qc_atms.inc
M       var/da/da_radiance/da_setup_radiance_structures.inc
M       var/da/da_radiance/da_crtm_sensor_descriptor.inc
A       var/run/radiance_info/jpss-0-atms.info
M       var/run/VARBC.in
A       var/run/crtm_coeffs/atms_npp.TauCoeff.bin
A       var/run/crtm_coeffs/atms_npp.SpcCoeff.bin
M       Registry/registry.var

2. The rest of the changes are minor: a bug fix for assimilation of the IASI instrument (r5961), additional output for radiance assimilation with the "print_detail_rad" option, and some spelling fixes:
M       var/da/da_radiance/da_read_obs_bufriasi.inc
M       var/da/da_setup_structures/da_scale_background_errors_cv3.inc
M       var/da/da_radiance/da_crtm_init.inc
M       var/da/da_radiance/da_crtm.f90

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6040 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1) allow for anthropogenic emissions to be read at the beginning of a  restart.
M       share/mediation_integrate.F

2) Fix an uninitialized variable
M       chem/module_ctrans_grell.F

3) Fix a missed chemistry compile dependency
M       chem/Makefile

4) Prevent the organic aerosols concentration from getting too small, or going negative.
M       chem/module_aerosols_soa_vbs.F
M       chem/module_aerosols_sorgam.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6038 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

A second option for a surface drag parameterization
   from Cliff Mass, which is added to the YSU PBL scheme. The new drag
   coefficients can be selected by setting the namelist variable
   "topo_wind=2".  The current scheme/factor is retained as "topo_wind=1",
   and the default remains 0.  I have also added the necessary arrays to NMM,
   which simplifies the YSU package - although the calculation of the
   coefficients is not yet implemented for NMM (so the effect is to multiply
   by a factor =1.0 if this flag is selected)

M    phys/module_bl_ysu.F
M    dyn_nmm/module_PHYSICS_CALLS.F
M    dyn_nmm/start_domain_nmm.F
M    run/README.namelist
M    Registry/Registry.EM_COMMON
M    Registry/Registry.NMM_HWRF
M    Registry/Registry.NMM_NEST
M    Registry/Registry.NMM_CHEM
M    Registry/Registry.NMM
M    dyn_em/start_em.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6036 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

   1. bug fix for binary i/o
   M       external/io_int/io_int.F90

   2. restrict the module-use in module_IGWAVE_ADJUST.F (NMM-only) so that
   the most recent XLF compiler doesn't complain (discovered with the XLF
   compiler at NCEP...)
   M       dyn_nmm/solve_nmm.F
   M       dyn_nmm/module_IGWAVE_ADJUST.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@6007 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

mynnsfcscheme is not defined for NMM, it needs to be if-def'd out when we make it work together with fractional seaice option

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5998 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

M phys/module_surface_driver.F
module_surface_driver.F is revised to ensure MYNN surface scheme is used correctly when fractional ice cover is turned on

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5992 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

A tiny Noah-MP bug fix. A pretty innocuous situation, where an
uninitialized variable was being multiplied by zero, but I managed to
find a combination of circumstances and strict compiler options (with
ifort) where it gave me problems.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5990 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

A fix for reading innovation from disk file for very very intensive
Obs., it may pick up the "neighbor" in wrong, and if the two 'close
neighbors' just happended to have different vertical levels, the
memory issue comes. Increasing the allowed minimus location difference
from 1.0E-4 to 1.0E-6 to fix this issue.

M var/da/da_obs_io/da_search_obs.inc

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5971 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Clarify warning/message for missing namelist record
(&logging - added by HWRF - non-std msg handling)

M module_wrf_error.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5969 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

This was fixed in NAM but missed when the NAM updates were
added to WRF last year.
M module_radiation_driver.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5968 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Diagnosed PBL height - bug re-introduced with update from EMC
in V3.4 - they concur that v3.3 has the correct indexing.

M module_bl_myjpbl.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5967 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

-Change the variable ntmax, which is currently an integer setting the maximum number of iterations of cost function minimization for each outer loop. The proposed change makes this variable a vector, allowing for a different maximum number of iterations for each outer loop. The specific files changed are:
M       Registry/registry.var
M       var/da/da_minimisation/da_kmat_mul.inc
M       var/da/da_minimisation/da_minimise_lz.inc
M       var/da/da_minimisation/da_minimise_cg.inc
M       var/da/da_minimisation/da_sensitivity.inc
M       var/da/da_minimisation/da_amat_mul.inc
M       var/da/da_minimisation/da_lanczos_io.inc
M       var/da/da_main/da_solve.inc

-The second change will add support for a new instrument: IASI, the Infrared Atmospheric Sounding Interferometer, found aboard MetOp satellites. The specific files changed are:
M       Registry/registry.var
M       var/build/depend.txt
M       var/da/da_setup_structures/da_setup_obs_structures.inc
M       var/da/da_setup_structures/da_setup_structures.f90
M       var/da/da_radiance/da_radiance.f90
A       var/da/da_radiance/da_cloud_detect_iasi.inc
M       var/da/da_radiance/da_radiance1.f90
M       var/da/da_radiance/da_qc_rad.inc
A       var/da/da_radiance/da_read_obs_bufriasi.inc
M       var/da/da_radiance/da_radiance_init.inc
A       var/da/da_radiance/da_qc_iasi.inc
M       var/da/da_radiance/da_setup_radiance_structures.inc
M       var/da/da_radiance/da_crtm_sensor_descriptor.inc
A       var/run/radiance_info/metop-2-iasi.info
M       var/run/VARBC.in
A       var/run/crtm_coeffs/ODPS/iasi616_metop-a.TauCoeff.bin
A       var/run/crtm_coeffs/iasi616_metop-a.TauCoeff.bin
A       var/run/crtm_coeffs/iasi616_metop-a.SpcCoeff.bin

-The final change is the addition of a new ECMWF cloud detection scheme. The specific files changed are:
M       Registry/registry.var
M       var/da/da_define_structures/da_define_structures.f90
M       var/da/da_radiance/da_allocate_rad_iv.inc
M       var/da/da_radiance/da_cloud_detect_airs.inc
A  +    var/da/da_radiance/da_cloud_sim.inc
D       var/da/da_radiance/da_cloud_sim_airs.inc
M       var/da/da_radiance/da_radiance1.f90
M       var/da/da_radiance/da_initialize_rad_iv.inc
M       var/da/da_radiance/da_rttov.f90
M       var/da/da_radiance/da_qc_airs.inc
M       var/da/da_radiance/da_crtm.f90
M       var/da/da_radiance/da_get_innov_vector_crtm.inc

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5961 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

MODIS data referred in RUC LSM should be MODIFIED_IGBP_MODIS_NOAH

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5959 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

bug fix from Greg for radar ref calc for Thompson scheme

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5958 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Generalize the Grib1 "diffwrf-esque" shell script to work
from any location. If not called from the WRFV3/test/em_real
directory, the third argument to the script is the location
of the wgrib.exe command.

M diffwrf

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5948 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

I am proposing a commit to change file 'input_wrf.F' This is regarding the 'override_restart_timers' option in the namelist, which when set to .true. should allow users to specify other intervals they wish to use for history files, other than those specified in the wrfrst* files, when running a restart. Prior to this change, this option did not work correctly when using the 'fdda' and 'sst_update' options, which use input files at set intervals. This has been corrected and now should only use output files, and not input files, when running a restart for these two options.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5946 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

update a few README files

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5931 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

change to correct namelist var use_tot_or_hyd_p

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5930 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

update READMEs

M README.DA
M var/README.basics

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5928 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Quick fix to allow a multiply by two when there are multiple
resolutions of floating data.  Problem stems from the ambiguity
allowed in the standard.  When using -r8, what does DOUBLE
PRECISION mean?  If it is just another real*8 variable, then
multiplying by 2.0d0 is OK.  If, however, DOUBLE PRECISION is
interpreted as real*16, then 2.0d0 promotes the the product
with COMPLEX*16 variable to a COMPLEX*32 (or tries to).

M    phys/module_mp_radar.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5925 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Remember Noah-MP variable TAUSS from one timestep to another

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5920 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

A couple of fixes for the radar reflectivity from Greg Thompson.

1. Swap the module_mp_radar.F dependencies in the Makefile to
be module_mp_radar.o.
M    phys/Makefile

2. To allow the continued use of the _ACCEL capability for PGI,
the original (un-changed!) WSM5 scheme that uses the
accelerator options is still able to be called from the
microphysics driver by adding in the _ACCEL cpp flags to surround
the new subroutine arguments.  THIS IS A SHORT-TERM SOLUTION.
M    phys/module_microphysics_driver.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5917 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Bug found by Kevin.  We need a conversion factor for
the center longitude.  This routine is never actually
called in WRF, but the same mod was fixed in WPS, and
there are those who pull source from WRF for diag
purposes.

M    module_llxy.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5916 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

In preparation for a 3.4.1 release, update the version info.

M version_decl

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5915 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

add surface diagnostic variable feedback, T2, Q2, U10, V10 and psfc

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5913 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

bug fix if considering EC-type aerosol, not currently affetcting results

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5912 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1. Replace an if-def for NMM that was inadvertently removed...  (required for non-XLF compilers)

2. Calling args for NMM/RRTMG - bug fix

M    phys/module_bl_ysu.F
M    phys/module_ra_rrtmg_sw.F
M    dyn_nmm/module_PHYSICS_CALLS.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5910 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

5 small bug fixes to the chemistry

1) Correct a dimension bug in the biomass burning array

2) Add no2 biomass burning emissions that were missed for a few chemistry options.

3) Add RACM_SOA_VBS_KPP to the optical properties initialization

4) Fix dependencies in the chemistry Makefile

5) Fix a small bug in the wet deposition

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5908 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1. implemented radiance blacklisting (using reference info from both EC and NCEP).
   set namelist variable use_blacklist_rad=.true. (default is .false.) to activate it.
M       var/build/depend.txt
A       var/da/da_radiance/da_blacklist_rad.inc
M       var/da/da_radiance/da_radiance.f90
M       var/da/da_radiance/da_radiance_init.inc
M       Registry/registry.var

2. implemented radiance weighting function calculation (CRTM only).
   set namelist variable calc_weightfunc=.true. (default is .false.) to activate it.
   Original contribution from Kathryn Newman and Zhiquan Liu.
M       var/da/da_define_structures/da_define_structures.f90
M       var/da/da_main/da_solve.inc
M       var/da/da_monitor/da_rad_diags.f90
M       var/da/da_radiance/da_allocate_rad_iv.inc
M       var/da/da_radiance/da_initialize_rad_iv.inc
M       var/da/da_radiance/da_crtm.f90
M       var/da/da_radiance/da_get_innov_vector_crtm.inc
M       var/da/da_radiance/da_radiance1.f90
M       var/da/da_radiance/da_write_iv_rad_ascii.inc
M       Registry/registry.var

3. implemented fixes for using MSU and HIRS sensors on TINOS-N and NOAA-1-12
M       var/da/da_obs_io/da_read_y_unit.inc
M       var/da/da_radiance/module_radiance.f90
M       var/da/da_radiance/da_read_obs_bufrtovs.inc
M       var/da/da_radiance/da_crtm_sensor_descriptor.inc
M       var/run/VARBC.in

4. remove snow over water
M       var/da/da_transfer_model/da_transfer_wrftoxb.inc

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5906 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Inline reflectivity calculation code for 8 schemes from Greg Thompson.
This is computed on history output times only and controlled by namelist
option do_radar_ref (1 or 0) (Lin and Goddard schemes do not give
bit-for-bit reflectivity results at the commit time). It also contains
some update for Thompson scheme (e.g. graupel intercept parameter change).

M    phys/module_mp_wdm6.F
M    phys/module_mp_wsm5.F
M    phys/module_microphysics_driver.F
M    phys/module_mp_morr_two_moment.F
M    phys/Makefile
M    phys/module_mp_wsm6.F
M    phys/module_mp_thompson.F
M    phys/module_mp_wdm5.F
M    phys/module_mp_lin.F
M    phys/module_mp_gsfcgce.F
A    phys/module_mp_radar.F
M    frame/module_domain.F
M    Registry/Registry.EM_COMMON
M    dyn_em/solve_em.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5897 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

A few updates here:

- bug fix for EDMF-QNSE (by Steven Decker)

M    phys/module_bl_mfshconvpbl.F

- compatibility change with fractional seaice use (Joe Olson)

M    phys/module_surface_driver.F

- NoahMP namelists changed to single domain varialbes;
Changed description to extreme surface variables;
Fix in YSU for nest starting at a later time - caused by
adding topo_wind change.

M    Registry/Registry.EM_COMMON
M    Registry/Registry.NMM_HWRF
M    Registry/Registry.NMM_NEST
M    Registry/Registry.NMM_CHEM
M    Registry/Registry.NMM
M    dyn_em/start_em.F

- mods for isftcflx=2: 2012 AHW version:

M    phys/module_sf_sfclay.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5896 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

changing comments

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5895 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

changing sounding from constant theta to non-constant theta

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5894 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

module_surface_driver.F was revised by Fernando De Sales (UCLA) to fix the inconsistency between landuse type and ssib/ssib_seaice calling.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5892 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Make vegetation parameters consistently zero for Urban points.

Also, sneak in a correction to the README.namelist documentation, where
the dveg options were misnumbered in the comments.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5890 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

A bug fix for the NMM-calling of Tiedtke scheme...

M dyn_nmm/module_PHYSICS_CALLS.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5888 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Some changes for WRFDA. Basically, they are changes related to running WRFDA with more than one thousand processors. The file name of many output files is appended with the processor number. All "i3.3" associated with myproc need to be replaced by i4.4 when more one 1000 processors are used.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5885 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Bug fix: remove ifdef WRF_CHEM statement from the cumulus driver call.

Add checks/resets to module_check_a_mundo for the Grell scheme, cu_diag and cu_rad_feedback settings

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5883 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

For AIX, add a new stanza specifically set up for netcdf4.

M configure_new.defaults

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5880 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Add in the LIB_LOCAL argument to the call the the
external/io_netcdf/makefile when building diffwrf.
Again, most times, this is either blank or a harmless
MPI lib. It is important when we are doing netcdf4.

M Makefile

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5879 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Add in the LIB_LOCAL macro.  This is either blank or harmless MPI
libs, most of the time.  We only really care about this variable
in the makefile when doing netcdf4.

M    external/io_netcdf/makefile

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5878 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Remove a comment from ctrans_grell and fix a coefficient for the GHG scheme. Oh, and add the shallow cumumuls fix that should have been added last week.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5874 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Remove the glacial ice treatment from Noah LSM.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5871 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

YSU consistency fix for thermal roughness length

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5869 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

YSU fix for stable surface conditions (Basu) and consistency with thermal roughness length

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5868 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

MYNN addition of tke budget switch to add diagnostic outputs and minor constant changes

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5867 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

changes from MYNN and YSU updates

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5866 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

add MYNN budget switch

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5865 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

changes from MYNN and YSU updates

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5864 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

changes from MYNN and YSU updates

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5863 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Add a new chem option: a 4-bin volcanic ash transport scheme.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5856 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1) Correct the description for volcanic ash arrays - should be mixing ratio instead of concentration.

2) A small fix to the lightning parameterization related to the grid size adjustment in the parameterization

3) Add chem_opt=108 to the possible choices for MEGAN biogenics. This is a stop-gap fix as a correction needs to be added to finish this additional capability.

4) A small fix to the Grell G3 scheme and shallow cumulus.

5) A bug fix to the AFWA dust scheme

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5855 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1. Yet another tweak to:

  M   dyn_nmm/module_STATS_FOR_MOVE.F

  to get all flavors and all compilers to compile

2. A bug fix to:

  M   phys/module_cumulus_driver.F

  the adaptive time step checks were using an optional arg
  in an if-test, which fails for NMM/PGI (works fine on bluefire)

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5853 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Noah-MP updates, addressing a number of issues we've discovered in the
code since the 3.4 release.

Changes only to phys/module_sf_noahmplsm.F

-- Energy balance
-- Distribution of energy over vegetated/non-vegetated/canopy
portions of a grid cell.
-- Consistent vegetation fraction/vegetation density
-- Computation of 2m diagnostics

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5851 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1) fix a few variable descriptions in the registry.chem file where concentration should instead be mixing ratio.

2) Uncomment the reading of OL2 for global anthropogenic emissions in prep_chem_sources.

3) Correct the aging of black carbon aerosols (black carbon to organic carbon conversion rate) in gocart_aerosols

4) Fix the SO2 to SO4 conversion if aqueous phase chemistry in convective clouds is turned on for the run.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5836 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1. Adding a piece of mod for saving/restoring outer-loop control variable cvt to/from file to da_solve.inc. Using this feature to implement the multi outer-loop for 4DVAR multi-incremental run.
M var/da/da_main/da_solve.inc

2. Adding an outer_loop_restart option for wrfvar6 namelist, which used for saving/restoring outer-loop control variable cvt to/from file
M Registry/registry.var

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5834 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1) Correct the order of chemical species in registry.chem

2) Add two RACM options to the MEGAN biogenic emissions that were missed

3) skip dry deposition when dust or volcanic ash are run.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5826 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Place ifdefs for HWRF around the routine vt4_MOVE.

M module_STATS_FOR_MOVE.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5824 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Just adding in the comment line "#Registry file, WRFVAR". It prevents from rebuilding the the entire WRFDA system when only a single file is modified.

M Registry/Registry.wrfvar

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5817 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1. NMM only - ifdef a couple of sections that should have been HWRF and NMM_NEST'd
  only (moving nests, thanks Brian!)

M        dyn_nmm/module_STATS_FOR_MOVE.F

2. "clean" clean-up (thanks Tim!)  The clean script deletes the configure.wrf file
  before running `make superclean`, which means when the superclean goes into some
  subdirectories (i.e. external/io_int) it fails as the makefile has "include
  ../../configure.wrf" which no longer exists. The patch is to move the renaming of
  configure.wrf after the superclean call.

M        clean

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5815 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Remove debug statements ...

M module_cumulus_driver.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5807 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Fix troubles related to adaptive time stepping and non-zero
cudt.

1. Pull all timing for running cumulus schemes out of the
individual routines, and put the check in the driver.  This
fixes the troubles with OpenMP + adaptive dt + non-zero cudt
causing the cumulus schemes to get into a race condition as
to whether or not to run the scheme.
M    phys/module_cu_tiedtke.F
M    phys/module_cu_g3.F
M    phys/module_cu_nsas.F
M    phys/module_cu_bmj.F
M    phys/module_cu_camzm_driver.F
M    phys/module_cu_sas.F
M    phys/module_cu_gd.F
M    phys/module_cu_kf.F
M    phys/module_cumulus_driver.F
M    phys/module_cu_osas.F
M    phys/module_cu_kfeta.F

2. Several schemes should not ever use adaptive dt + non-zero cudt,
due to the assumed way that RAINCV and PRATEC are computed.  A
print statement tells users that they have selected something
incorrectly, and internally resets cudt to 0.  Impacted CU
schemes: BMJ, all SAS schemes, Tiedtke.
M    share/module_check_a_mundo.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5805 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Just before allocating arrays, check to see if they first
need to be de-allocated.

M module_stoch.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5804 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Since v3.3, the top half-level of Qv coming out of the real
program (ARW only) has been zero.  While this is only impacting
users who select the rh2qv_option=1, this is the default, and
probably very few know an option even does exist.  The problem
was that the index loop extent inside the routine rh_to_mxrat1
has "k = kts , kte-1".  Unfortunately the actual arguments use
"kte-1" for the dummy argument "kte".  We effectively get a loop
that runs from 1 to kte-2.  This leaves the kte-1 (the top level)
of the Qv empty.  The solution is to have the loop index inside
the routine run from kts to kte.

There are two rh_to_mxrat routines, and #2 is just fine.  Also,
the pressure level RH to Qv computation is now working correctly
with the fix also.

M    module_initialize_real.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5803 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Small tweaks in comments.

M configure

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5802 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1) Fix a hard-coded vegetation category index, which checked on the
wrong category for water when using the MODIS land-use database.

2) Fix a source of some serial/parallel differences that people have
been encountering with Noah-MP.  A few output variables related to the
vegetation canopy were not being assigned values over bare-ground
points.  This fix assigns values to these variables at bare-ground
points.  The contents at these points should not matter to the
simulation, because these vegetation-related terms are not used over
bare ground.

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5800 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1. Adding a regristry pre-process mod for WRFVAR
M       compile

2. Adding a few missing definitions and declarations
M       var/da/da_gen_be/da_gen_be.f90
M       var/da/da_control/da_control.f90
M       var/da/da_main/da_wrfvar_top.f90

3. Seperated and merged registry entries for WRFVAR

M       Registry/Registry.EM_COMMON
A       Registry/registry.var_chem
A       Registry/registry.var
M       Registry/Registry.wrfvar

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5798 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

NMM-only, part of the HWRF 2012 commits - missed on 5/18:
(some extra (optional) calling args for cumulus driver)

M dyn_nmm/module_PHYSICS_CALLS.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5796 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Sample namelist for using isobaric diagnostics.
AM test/em_real/namelist.input.diags

Required new registry file specifically for the p-lev diags.
A Registry/registry.diags

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5792 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Provide isobaric diagnostics for ARW, for stream 23.

M    test/em_real/examples.namelist
M    phys/module_diagnostics.F
M    phys/Makefile
M    frame/module_driver_constants.F
M    share/module_check_a_mundo.F
M    share/dfi.F
M    share/output_wrf.F
M    run/README.namelist
M    Registry/Registry.EM_CHEM
M    Registry/Registry.EM
M    dyn_em/Makefile
M    dyn_em/solve_em.F
M    dyn_em/start_em.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5791 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

The flag write_hist_at_0hr_rst now functions correctly.

M mediation_integrate.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5790 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Upon a restart, the frequency of the history and restart files
may now be modified.

M input_wrf.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5789 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Fix the way that the advective tendency arrays, advh_t and advz_t, are
handled. Since they are 4-d arrays, they need to be explicitly declared
in a package.

M Registry.EM_COMMON

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5788 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

A newer version of gfortran (rightly) complained about the inconsistent
sizes of integers.

Make the total seconds use the same definition as in ESMF time lib.
M    dyn_em/adapt_timestep_em.F

Set integers to 4-byte size, as the internal documentation states that
this is the intent of the declaration.
M    dyn_em/module_wps_io_arw.F
M    phys/module_ra_rrtmg_lw.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5787 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Provide a couple more quick examples on how to add and
subtract fields on a couple different IO streams.

M sample.txt

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5786 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Complete ndown clean-up.  Provide users with example namelists
for the three ndown stages (CG, ndown, FG).  Also, remove the
extraneous prints in the main program.

AM   test/em_real/namelist.input.ndown_1
AM   test/em_real/namelist.input.ndown_2
AM   test/em_real/namelist.input.ndown_3
M    main/ndown_em.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5785 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Allow NetCDF4 capabilities in WRF.  These are all enclosed within ifdefs,
so CPP directives are required to implement them.  Those CPP directives
are not yet included in the build.  Supplied by Huang Wei (CISL).

M    wrf_io.F90

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5784 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

Missed one file... with -r5778

A dyn_nmm/module_STATS_FOR_MOVE.F

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5782 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

1.  On some platforms, multiple processors have difficulty to read a file simultaneously, so change to rootproc read, then broadcast.
M       var/da/da_setup_structures/da_setup_be_regional.inc

2. Typo fixes.
M       var/da/da_setup_structures/da_scale_background_errors_cv3.inc
M       var/scripts/gen_be/gen_be_plot_wrapper.ksh

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5780 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d

WRF updates for the 2012 HWRF operational model.

1. Add a new logging mechanism that can be turned off at compile-time, to
   limit the i/o load.  This touches lots of files - as "write(0" and "print"
   were updated to use the wrf_message or wrf_debug mechanisms.  FATAL
   error messages are always written.

   Log buffering must be enabled by setting WRF_LOG_BUFFERING=1 before
   compiling.  The options are configured with a new &logging namelist:

   &logging
       buffer_size = 4096   ! size of the log buffer, 0 to disable
       compute_slaves_silent = T/F ! if T, non-master compute ranks will be slient
       io_servers_silent = T/F ! if T, I/O server ranks will be silent
       stderr_logging = T/F ! if F, rsl.error.* will be nearly empty
   /

   The compute_slaves_silent, io_servers_silent, and stderr_logging
   options will still work if WRF_LOG_BUFFERING is disabled.  The
   defaults are set to mimic previous behaviors.

M       configure
M       external/atm_ocn/cmpcomm.F
M       external/atm_ocn/atm_comm.F
M       external/atm_ocn/mpi_more.F
M       external/io_int/diffwrf.F
M       external/io_int/makefile
M       external/io_int/io_int.F90
M       external/io_netcdf/diffwrf.F90
M       external/io_netcdf/makefile
M       phys/module_ra_gfdleta.F
M       phys/module_pbl_driver.F
M       phys/module_sf_gfdl.F
M       phys/module_cu_sas.F
M       phys/module_physics_init.F
M       main/real_nmm.F
A       frame/clog.c
M       frame/Makefile
M       frame/module_timing.F
M       frame/module_quilt_outbuf_ops.F
M       frame/module_wrf_error.F
M       frame/wrf_debug.F
M       frame/module_io_quilt.F
A       dyn_nmm/module_STATS_FOR_MOVE.F
M       dyn_nmm/module_NEST_UTIL.F
M       dyn_nmm/module_si_io_nmm.F
M       dyn_nmm/module_initialize_tropical_cyclone.F
M       dyn_nmm/NMM_NEST_UTILS1.F
M       dyn_nmm/solve_nmm.F
M       dyn_nmm/module_TERRAIN.F
M       share/init_modules.F
M       share/mediation_nest_move.F
M       share/Makefile

2. Update shallow convection in module_cu_sas.F
M       phys/module_pbl_driver.F
M       phys/module_cu_sas.F
M       phys/module_bl_gfs.F
M       phys/module_cumulus_driver.F
M       Registry/Registry.NMM_HWRF

3. Constants tuning in module_bl_gfs.F and module_mp_HWRF.F
M       phys/module_bl_gfs.F
M       phys/module_mp_HWRF.F
M       share/module_model_constants.F

4. Re-arrange the nest-move logic in NMM (take out of module_NEST_UTIL.F, add module_STATS_FOR_MOVE.F
   - also add options for several different vortex tracking algorithms
M       dyn_nmm/module_NEST_UTIL.F
M       dyn_nmm/Makefile
M       dyn_nmm/start_domain_nmm.F
M       dyn_nmm/solve_nmm.F
M       share/mediation_nest_move.F
M       Registry/Registry.NMM_HWRF

5. A couple of minor bug fixes
M       dyn_nmm/NMM_NEST_UTILS1.F
M       dyn_nmm/start_domain_nmm.F
M       dyn_nmm/module_IGWAVE_ADJUST.F

6. Add args/flags to allow NMM to use TIEDTKE and KF cumulus
M       dyn_nmm/solve_nmm.F

7. Update a few comments - check for analysis-output-file for HWRF
M       share/mediation_integrate.F
M       Registry/Registry.NMM_HWRF

8. Add un-buffered logging from configure (to work with automated perl build scripts nicely), tweak compile-max-mem AIX opts
M       arch/Config_new.pl
M       arch/configure_new.defaults

git-svn-id: https://svn-wrf-model.cgd.ucar.edu/trunk@5778 b0b5d27b-6f0f-0410-a2a3-cb1e977edc3d