7.4 Other output files from post processor
As previously mentioned, the model forecast will produce three kinds of binary files: sigma files (r_sigf*), surface file (r_sfcf*) and flux files (r_flxf*). The post processor creates two kinds of GRIB files and related control files: pressure-flux GRIB files (r_pgbf*) and sigma GRIB files (r_sgbf*). Their control files are r_pgbf*.ctlprs and r_sgbf*.ctlsig, respectively. The previous section described how to transform a binary file into a packed GRIB file, and how to unpack a packed GRIB file and transform it into a binary file.
Besides all the files above, the current model also has two other options to produce two more binary files from the forecast. The first one is, called Ken point output, provides data at pre-determined model points. The model can save any given time period and variables at any level related to the locations of the observation station. The another kind of out put is called budget output. Please note that currently only the hydrostatic version can have associated budget terms. Additional work is still needed for the nonhydrostatic budgets.
Station-type output:
related definition in DCLSYS
for regional experiment: %DCL TECHNIQUE OPTIONS; % ##RKN = ' '; %DCL REGIONAL KEN POINTS; %DCL KEN POINT FOR REGIONAL; % #RLPNT = '50' ; %DCL maximal point allowable; % #RLTSTP = '36' ; %DCL maximal steps for saving; % #RSLVARK = '80' ; %DCL maximal number of single variable; % #RMLVARK = '8' ; %DCL maximal number of #LEVR variables; for global experiment: % ##KEN = ' '; % #LPNT = '150' ; % #LTSTP = '72' ; % #SLVARK = '80' ; % #MLVARK = '8' ;
related namelist input: NUM and CON
for regional experiment: NUM(60) determines the pointer for
latitude and longitude in CON
NUM(61) frequency in min for short wave
flux output
NUM(62) output frequency for long wave
flux
NUM(63) output frequency in min for all
other fields
NUM(64) total number of output points
for global experiment: NUM(50) to NUM(54) as the same as NUM(60)
to NUM(64) for global
for example, NUM(60)=100, then the first point latitude and
longitude is CON(NUM(60)+1) and CON(NUM(60)+#RLPNT+1), and NUM(64)
has to be less than #RLPNT.
file unit assigned in PGFCST and PRFCST:
for regional experiment: unit 79 r_kenf for global experiment: unit 67 kenf
related common block:
for regional: filenamed RSCOMGPD for global: filenamed COMGPD
related routines:
for regional: RSRSMINI : prepare some constant (file)
RSRLOOPA, RSRBPHYS, RSGRRAD2: save fields
RSRSMSAV : output
for global: KENPRE : prepare some constant (routine)
GBPHYS and GRRAD2 : save related fields (files)
KENPUT : output routine (file)
This type of station output used to be called KEN point output. It is not like other physics routines, they are plug-able. Effort should be spent to make it more plug-able, not spread over inside a routines. It should be improved in later version.
Budget-type output:
It can be turn on in DCLSYS. The variable definition is as:
%DCL REGIONAL ENERGY BUDGET; % ##A = ' '; %DCL TURN ON WITH THREE SPACES FOR ANY BUDGET; % ##T = ' '; %DCL FOR TEMPERATURE BUDGET; % ##Q = ' '; %DCL FOR MOISTURE BUDGET; % ##U = ' '; %DCL FOR U MOMENTUM BUDGET; % ##V = ' '; %DCL FOR V MOMENTUM BUDGET; % ##P = ' '; %DCL FOR PS MASS BUDGET;
in this case all the budget equation are opened, includes temperature, moisture, u v momentum and mass equations. You can have only one or two or three or four, but ##A has to be opened if any one the budget is opened.
The related source code and file unit etc as:
the related files are : RSRSMSMF
RSRLOOPA
RSFIDI
RSRLOOPB
RSRBPHYS
the related units: the output unit is 85,
but there are two swap files,
they are unit 95 and 96.
the record is (see RWRTBGT in RSRSMSMF
record 1 : 'RSM MODEL BUDGET'
record 2: THOUR,IDATE,(IVAR(N),N=1,5)
record 3 .... depends on IVAR
program to read the budget file:
read(n,'9A16)') lab
read(n,'(F8.0,9I8)') FHOUR,(IDATE(I),I=1,4),(IVAR(N),N=1,5)
do num=1,5
IF(IVAR(num).NE.0) THEN
DO N=1,IVAR(1)
KN=#LEVR
if(num.eq.5) KN=1
DO K=1,KN
if(num.eq.1)READ(n,10) (T(I,K,N),I=1,#LNGRD)
if(num.eq.2)READ(n,10) (Q(I,K,N),I=1,#LNGRD)
if(num.eq.3)READ(n,10) (U(I,K,N),I=1,#LNGRD)
if(num.eq.4)READ(n,10) (V(I,K,N),I=1,#LNGRD)
if(num.eq.5)READ(n,10) (PS(I,K,N),I=1,#LNGRD)
ENDDO
ENDDO
ENDIF
enddo
10 FORMAT(1X,6E13.6)
The field output can be found in source97/READMEBGT for hydrostatic option only:
Regional budget computation is switched by ##A ##T #NT=15 (output 16 terms) ##Q #NQ=10 (output 11 terms) ##U #NU=11 (output 12 terms) ##V #NV=11 (output 12 terms) ##P #NP=6 (output 7 terms) The output file contain one more term called base field tendency ---- T equation ---- 1. Horizontal advection 2. Vertical advection 3. Pressure changes 4. Vertical diffusion and PBL physics 5. Horizontal diffusion 6. Cumulus convection 7. Shallow convection 8. Large scale precipitation 9. Short wave radiation 10. Long wave radiation 11. Virtual temperature effect 12. Semi-implicit adjustment 13. Lateral boundary effect 14. Spectral truncation 15. Model tendency 16. Base field tendency ---- Q equation ---- 1. Horizontal advection 2. Vertical advection 3. Vertical diffusion and PBL physics 4. Horizontal diffusion 5. Cumulus convection 6. Shallow convection 7. Large scale precipitation 8. Lateral boundary effect 9. Spectral truncation 10. Model tendency 11. Base field tendency --- U & V equation --- 1. Horizontal advection 2. Vertical advection 3. Coriolis force 4. Pressure gradient force 5. Vertical diffusion and PBL physics 6. Horizontal diffusion 7. Gravity wave drag 8. Semi-implicit adjustment 9. Lateral boundary effect 10. Spectral truncation 11. Model tendency 12. Base field tendency --- PS equation ---- 1. Horizontal advection 2. Divergence effect 3. Semi-implicit adjustment 4. Lateral boundary effect 5. Spectral truncation 6. Model tendency 7. Base field tendency
The advantage of the built-in budget analysis is that the budget is completely balanced. The drawback is that budgets require lots of memory and large IO at each step. This can really slow down computations and thus this option should not be normally turned on.
webmaster: Hann-Ming Henry Juang
henry.juang@noaa.gov