ONLINE DOCUMENTATION FOR BGCMFLUX, ED=2 _______________________________________ Mike Blackburn Department of Meteorology, University of Reading. 04.12.89 INTRODUCTION. ------------- This is a brief description on running ED=2 of the diagnostics program BGCMFLUX on the output of an integration of the Reading University dry or moist baroclinic spectral model. More detailed documentation will be available shortly in the form of a UGAMP Internal Report. This program will also form the basis of a new version of the diagnostic program for use with history data from the UGAMP version of the ECMWF model. The program is stored as an update library on Cray disc, at both ULCC and RAL, in PDN=BGCMFLUX,ID=UPLIB,ED=2. At ULCC use OWN=GBAR361 and at RAL use OWN=KD. Auxiliary subroutines required are obtained precompiled from the binary libraries PDN=SUBLIB,ID=LIB and PDN=ZPLOT,ID=LIB, with OWN=GBAR361 at ULCC but OWN=PV at RAL. The source code for SUBLIB and ZPLOT is available in update libraries having ID=UPLIB if required. This documentation is stored on the front end at both sites, or is available from the author. At ULCC it is a member of a partitioned dataset on MVS, GBAR379.BDOC.TEXT(BGFLUX2). At RAL it is stored on CMS as BGFLUX2 DOC on minidisk 194 of user KD. Both have public read access. DESCRIPTION OF DIAGNOSTICS AND PRINTED / PLOTTED OUTPUT. -------------------------------------------------------- BGCMFLUX produces 9 types of diagnostic output, each having a separate switch (usually LPCn) and channel for printed output (NPCn): Type 1 Spectral coefficients of model level fields. Type 2 Latitude-longitude fields on model or isobaric levels. Type 3 Zonally averaged dynamical fields. Type 4 Kinetic energy spectra. Type 5 Global energetics. Type 6 Latitude-longitude fields on isentropic levels. Type 7 Latitude-longitude surface/physics fields. Type 8 Zonally averaged parameterisation tendencies. Type 9 Globally averaged surface/physics diagnostics. Both printed and plotted output can be produced, with the latter on microfilm and/or written to a "UGAMP data transfer file" (UTF) for plotting elsewhere. Microfilm is switched on for the entire program by KOUNTF, while switch KOUNTA controls production of the UTF. The switches and data controlling plotting of individual fields apply to both film and UTF output. See namelist INPB below for details. Printed output: By default gridded fields are written to Fortran channel 3 and only global diagnostics are copied to the output channel (Fortran channel 6, COS dataset $OUT). This allows the majority of printed output to be written to microfiche, the necessary commands to do so at ULCC and RAL are included in the example jobdecks below. For smaller volumes of printed output, all fields can be written to the output channel instead. Microfiche is posted to the user from both ULCC and RAL. Plotted output: The DIMFILM library of plotting routines used by BGCMFLUX creates a plot file which can be processed by the DICOMED film recorder at ULCC to produce 16mm or 35mm microfilm. When the program is run at ULCC, this plot file is handled automatically and film is posted to the user site. However at RAL the user must handle the plot file, copying it to 1600bpi tape(s) and arranging for these to be posted to ULCC for plotting on the DICOMED. Given typical plot file sizes (several megabytes) this is impractical, so an alternative form of output has been provided. Film output can be switched off and instead (or additionally) a "UGAMP data Transfer File" (UTF) produced, which can be transferred over JANET to either the user site or ULCC for plotting. Include DC=PU in the DISPOSE statement to avoid a shifted copy of the file being transferred. Separate programs are available at ULCC (QPLOTU) and Reading (QPLOT3), or from the author, to create plots from this data file. A procedure will shortly be made available on the Cray at RAL to send the UTF to ULCC and automatically plot it: watch this space! Units of fields: Data on the model history file are assumed to be non-dimensional and the program dimensionalises all fields before output using a local array DFAC in each output routine. Fields can then be printed and plotted either dimensional or normalised relative to the maximum value or range of the field. These options are controlled by a printing factor and contour interval for each field as follows ( aa... differs for each output type): aaFAC is a (dimensional) factor giving the units in which the field is printed. The aim is for the normalised field to have a maximum absolute value less than 1000, so that values fit into an I4 format. The factor is overridden if the contour interval is negative. aaINC is the (dimensional) contour interval for plotting, if positive. If negative, the fields are both printed and plotted normalised as follows: aaINC < -1 : for printing normalise the maximum absolute value of the field to 1000. For plotting normalise the range of the field to 1000, the contour interval being -(aaINC) in normalised units. Eg. aaINC=-200 gives a contour interval 1/5th of the field range. -1 <= aaINC < 0 : for printing normalise the maximum absolute value of the field to lie between 0.1 and 1.0 by removing powers of 10 only. For plotting normalise the range of the field to lie between 0.1 and 1.0 by removing powers of 10 only, the contour interval being -(aaINC) in normalised units. Eg. aaINC=-0.2 gives a contour interval 1/5th of the next power of 10 greater than the field range. All factors and contour intervals are dimensional and are defined in a table for each output type in the description of namelist INPB below. The dimensional units of all printed fields and dimensional contour intervals are included in the output. EXAMPLE JOBDECK AT ULCC. ------------------------ The following is a typical set of JCL, update directives and data needed to run the program at ULCC. The fiche commands and directives, dataset accesses and fetches differ at RAL (see below). JOB,US=........................... ACCESS,DN=$PL,PDN=BGCMFLUX,ID=UPLIB.OWN=GBAR361,ED=2. UPDATE,F,IN,ED. CFT,L=0,I=$CPL,ON=AZ,AIDS=LOOPALL. RELEASE,DN=$PL:$CPL. ACCESS,DN=SLIB,PDN=SUBLIB,ID=LIB,OWN=GBAR361. ACCESS,DN=ZLIB,PDN=ZPLOT,ID=LIB,OWN=GBAR361. ACCESS,DN=FILM,PDN=DIMFILM,OWN=PUBLIC. FETCH,DN=FT10,DF=TR,TEXT='DSN=xxx,DISP=OLD,UNIT=TAPE,'^ 'VOL=SER=xxx,LABEL=(xx,SL,,IN),DCB=DCB.CRAYTR'. FICHE,DN=FT03,START,COMMANDS,MSG='PLEASE POST MET. READING'. SEGLDR,L=0,CMD='LIB=SLIB,ZLIB,FILM;PRESET=INDEF',GO. FICHE,DN=FT03,END. EXIT. DUMPJOB. DEBUG,BLOCKS. /EOF *IDENT................. *D PARAM1.2,3 PARAMETER(MM=..,NN=..,NL=..,NHEM=..,MOCT=..,NWJ2=.. : ,MG=..,JG=..,,NCRAY=..,JGL=..,NTHSF=..) +further UPDATE changes to BGCMFLUX as necessary /EOF TITLE ' ', REGIONS=4:5 ,SIZE=3,LINE=1 TITLE ' METEOROLOGY ', REGIONS=4:5 ,SIZE=3,LINE=3 TITLE ' ', > REGIONS=6:15,SIZE=2,LINE=1 TITLE ' ', > REGIONS=6:15,SIZE=2,LINE=2 /EOF $INPB.................$ $INPHYS...............$ /EOF Notes: (1) If it is necessary to use the UPDATE version of SUBLIB (ID=UPLIB) then the following points should be noted. The code from BGCMFLUX must be written to the compile file ($CPL) before that from SUBLIB. The NR parameter is necessary in the UPDATE control statements relating to both files to ensure that $CPL is not rewound - an explicit REWIND is then necessary before the CFT statement and CAL must be invoked after CFT. (2) The jobdeck above assumes that the data to be processed was written to a 6250bpi standard labelled tape in transparent mode in the model integration. The DSN parameter, tape VSN and file number in the FETCH statement must be the same as those specified on writing the tape, and the JOB statement must include the ST parameter. Alternatively data stored on Cray disc should simply be ACCESSed. (3) The fiche commands and directives ensure that all data on Fortran channel 3 is written to microfiche. To include the printed fields in the output listing instead, omit the FICHE,...,END statement and use channel 6. The namelist variables NPC1 to NPC9 determine to which channel the various data types are written. If the FICHE,...,START command is also omitted, the directives and subsequent /EOF must also be removed. (4) The selection of ON=Z in CFT together with the DUMPJOB and DEBUG statements ensures that a post-mortem dump is produced in the event of job failure. EXAMPLE JOBDECK AT RAL. ----------------------- The following is a typical set of JCL, update directives and data needed to run the program at RAL. The fiche commands, dataset accesses and fetches differ at ULCC (see above). JOB,US=........................... ACCESS,DN=$PL,PDN=BGCMFLUX,ID=UPLIB.OWN=KD,ED=2. UPDATE,F,IN,ED. CFT,L=0,I=$CPL,ON=AZ,AIDS=LOOPALL. RELEASE,DN=$PL:$CPL. ACCESS,DN=SLIB,PDN=SUBLIB,ID=LIB,OWN=PV. ACCESS,DN=ZLIB,PDN=ZPLOT,ID=LIB,OWN=PV. ACCESS,DN=FILM,PDN=DIMFILM,OWN=GRAPHICS. ACCESS,DN=TEMP,PDN=xxx,LB=SL,DT=*TAPE,VOL=xxxxxx,FSEQ=xx. COPYD,I=TEMP,O=FT10. REWIND,DN=FT10. RELEASE,DN=TEMP. SEGLDR,L=0,CMD='LIB=SLIB,ZLIB,FILM;PRESET=INDEF',GO. NOTE,TEXT='1: ... user-id ... job no. ... title ... '. DISPOSE,DN=FT03,MF=VM,TEXT='TID=NCR5330A,FORM=T14'. EXIT. DUMPJOB. DEBUG,BLOCKS. /EOF *IDENT................. *D PARAM1.2,3 PARAMETER(MM=..,NN=..,NL=..,NHEM=..,MOCT=..,NWJ2=.. : ,MG=..,JG=..,,NCRAY=..,JGL=..,NTHSF=..) +further UPDATE changes to BGCMFLUX as necessary /EOF $INPB.................$ $INPHYS...............$ /EOF Notes: (1) If it is necessary to use the UPDATE version of SUBLIB (ID=UPLIB) then the following points should be noted. The code from BGCMFLUX must be written to the compile file ($CPL) before that from SUBLIB. The NR parameter is necessary in the UPDATE control statements relating to both files to ensure that $CPL is not rewound - an explicit REWIND is then necessary before the CFT statement and CAL must be invoked after CFT. (2) The jobdeck above assumes that the data to be processed was written to a 6250bpi standard labelled tape in transparent mode in the model integration, using an ACCESS,NEW statement similar to that above. The JOB statement must include *TAPE=1. Alternatively data stored on Cray disc should simply be ACCESSed directly as DN=FT10 and the COPYD, REWIND and RELEASE statements removed. (3) The NOTE and DISPOSE statements ensure that all data on Fortran channel 3 is written to microfiche. To include the printed fields in the output listing instead omit these statements and use channel 6. The namelist variables NPC1 to NPC9 determine to which channel the various data types are written. (4) The selection of ON=Z in CFT together with the DUMPJOB and DEBUG statements ensures that a post-mortem dump is produced in the event of job failure. PARAMETERS. ----------- Parameters defining model resolution and array sizes in the program must be supplied in the update directive section of the jobdeck, replacing program lines PARAM1.2,3 as follows: PARAMETER DEFAULT INTERPRETATION MM 21 Highest zonal wavenumber included in the truncation. NN 21 Highest total wavenumber included in the truncation. NL 15 Number of levels. NHEM 1 Selects hemispheric (1) or global (2) domain. MOCT 1 Only zonal wavenumbers zero and multiples of MOCT are included in the truncation. NWJ2 121 Number of odd or even coefficients included in the truncation. MG 64 Number of longitude points in grid - must be an even number with prime factors of 2 , 3 and 5 only. JG 16 No. of Gaussian latitudes between equator and pole. NCRAY 64 No. of Fourier transforms performed simultaneously. 64 should be optimal. JGL JG Controls number of latitudes for which Legendre functions and grid point orography (if used) are held in core simultaneously. The only valid choices are 1 or JG. The former reduces central memory required significantly at the expense of more I/O operations. NTHSF 5 Number of isentropic levels for potential vorticity and related diagnostics. Further parameters need to be changed only if extra fields are added to the diagnostics or very long time series are analysed. The following variables are in a separate PARAMETER statement in lines PARAM1.3,4 : PARAMETER DEFAULT INTERPRETATION NSG 12 Number of latitude-longitude fields on model or isobaric levels output by routine OPGRID. NGP NSG+2 Total number of model or isobaric level latitude- longitude fields on scratch channels NCGP(1 to NGP). NXD 28 Number of zonally averaged dynamical fields at the beginning of common blocks ZONAV and TAV which are output by routine FLXPRC. NXDF NXD+8 Total number of zonally averaged dynamical fields at the beginning of common ZONAV which are time aver- aged into TAV. NTH 4 Number of latitude-longitude fields on isentropic levels output by routine OPERTEL. NPH 6 Number of latitude-longitude surface/physics fields output by routine PHYSOP. NXP 19 Number of zonally averaged parameterisation tendency fields on the model history file, in common ZONAVP and output by routine PHYSEC. NPMAX4 1001 Maximum number of points in time series of kinetic energy output by routine NSPECT. NPMAX5 1001 Maximum number of points in time series of global energetics output by routine GDIAG. NPMAX9 1001 Maximum number of points in time series of global surface diagnostics output by routine PHYSAV. NTERM5 27 Number of terms in the global energetics produced by routine GDIAG and time averaged in common COMGEN. DATA FOR NAMELIST INPB. ----------------------- Data describing the model options and types of diagnostic output required must be supplied through namelist group INPB as follows (see also INPHYS below for the parameterised processes). Only variables for which the default is inappropriate need to be included in the jobdeck. They can be entered in free format in the form "variable=value", in any order, on several lines before the closing "$". A "-" indicates that no default is provided. DATA DESCRIBING MODEL INTEGRATION. VARIABLE DEFAULT INTERPRETATION RNTAPE - Run identifier on history records. TSPD - Number of time steps per day for the history records supplied. KITS - Number of short initial time steps used in model integration. TMEAN - NL values of basic state temperatures in Kelvin as used in model run. GA 9.81 Gravitational acceleration in m/s2. RD 287.0 Gas constant for dry air in J/kg/K. Used in non- dimensionalising temperatures. RV 461.51 Gas constant for water vapour in J/kg/K. AKAP 0.286 RD/CPD ( = R/CP for dry air). CLATNT 2.5E6 Specific heat of vapourisation of water in J/kg. RADEA 6371000. Radius of the earth in metres. Also used to non- dimensionalise lengths. WW 7.292E-5 Angular rotation of the earth in s-1. Also used to non-dimensionalise time. P0 1.E5 Reference surface pressure in N/m2. Also used to non-dimensionalise pressures. TDISS 0.125 Diffusion timescale in sidereal days on the shortest resolved scale, as used in model integration. No dissipation is included if TDISS=0. NDEL 8 Order of diffusion: ie. 8 for del 8. RESTIM 0. Restoration timescale in sidereal days used in model integration. No restoration included if RESTIM=0. DRAG NL*0. Timescale of linear drag at each level in siderial days. BASIC SWITCHES AND COUNTERS. BEGDAY - First day for which data is to be processed. ENDDAY - Last day for which data is to be processed. KOUNTD - Basic counter giving frequency of diagnostics in timesteps. History data are read, fields analysed and global diagnostics produced every (KOUNTD/TSPD) days. KOUNTP - Frequency in timesteps for output of printed fields. Spectral coefficients, fields on model/isobaric and isentropic levels, surface fields and zonal averages are all printed every (KOUNTP/TSPD) days. Must be a multiple of KOUNTD. KOUNTP=0 suppresses output. KOUNTF - Frequency in timesteps for grid-point fields to be plotted on microfilm: ie. fields are plotted every (KOUNTF/TSPD) days. Must be a multiple of KOUNTD. KOUNTF=0 suppresses ALL film output. KOUNTE - Frequency in timesteps for calculation and output of the KE spectra. Also processed at day 0.5 if the corresponding history record is included. Must be a multiple of KOUNTD. KOUNTE=0 suppresses calculation of spectra. KOUNTA - Frequency in timesteps for fields to be written to the data transfer file (UTF), for plotting at a different site. Fields are chosen by the plotting switches for each data type (see sections below), irrespective of whether film is switched on (see KOUNTF). Must be a multiple of KOUNTD. KOUNTA=0 suppresses output of file. NPCA 65 Fortran channel used for data transfer file (UTF). This can be saved or disposed to a different site. LSSD .F. All scratch files are assigned to reside on SSD if true (valid at RAL only), in which case include *SSD parameter in the JOB statement. Default is for all scratch files to reside on disc. SWITCHES AND DATA CONTROLLING OUTPUT OF SEVERAL DATA TYPES. NLAT JG Number of gaussian latitudes for which grid point values are to be printed. Applies to output types 2,3,6,7,8 (model/isobaric level and isentropic fields, surface/physics fields and zonal sections). NLAT=0 suppresses printing of all these fields. NLONG MG Number of longitudes for which grid point values are to be printed. IRSKIP 2 Skip in (x,y) grid for plotting of wind vectors on lat-long plots. ARSTD 30.0 Scaling vector magnitude equivalent to one grid interval, in same units as the vector field. ARMAX 100.0 Maximum vector magnitude plotted. ARMIN 4.0 Minimum vector magnitude plotted. ISPLIT 100 Number of points in the time series of KE spectra and global energetics and surface diagnostics after which they will be split. For example, for the default value and the choice KOUNTD=TSPD, the time series will be split into sections each 100 days long, while if KOUNTD=0.5*TSPD each section will be 50 days long (KOUNTE for the spectra). The maximum total lengths of the time series are determined by the parameters NPMAX4, NPMAX5 and NPMAX9. IZGTYP 4 Projection type for latitude-longitude plots. The only tested projections are: 1 : regular latitude-longitude. 4 : polar sterographic. Other projections allowed (see GBPLOT doc.): 2 : Lambert conformal. 3 : Polar-visual image. 5 : Equatorial visual-image. 6 : Equatorial stereographic. 7 : Oblique visual. ICOAST 0 Switch to include coastlines on plots. Default is to omit coastlines. ICOAST=1 allows cheap low- resolution option for polar stereographic projection only. ICOAST=2 gives high resolution coastlines. SWITCHES AND DATA CONTROLLING FILM OUTPUT. L35MM .F. If true selects 35mm camera, otherwise 16mm. LROT .T. Redundant with present version of Dimfilm. 16mm plots are in cine mode and 35mm landscape mode. ZINT1 0.7 Intensity setting for 'very light' lines. ZINT2 0.8 Intensity setting for 'light' lines. ZINT3 0.9 Intensity setting for 'standard' lines. ZINT4 1.0 Intensity setting for 'heavy' lines. NAME 'NO-ONE' User's name to appear on title frame. Up to 2 strings of 8 characters are allowed, each enclosed in single quotes and separated by a comma. SWITCHES AND DATA CONTROLLING VERTICAL INTERPOLATION. LINTP2 .F. Switch for interpolation of lat-long gridded fields (type 2 output) from model to isobaric levels. LINTP3 .T. Switch for interpolation of zonal mean fields (type 3 output) from model to isobaric levels. LXTRP2 .T. Switch for extrapolation of lat-long gridded fields (type 2 output) beyond extreme model levels, when fields are on isobaric levels, LINTP2=.T.. If true, appropriate extrapolation methods are used for each variable. If false all fields are held constant beyond the extreme model levels. LXTRP3 .T. Switch for extrapolation of zonal mean fields (type 3 output) beyond extreme model levels, when fields are on isobaric levels, LINTP3=.T.. If true, appropriate extrapolation methods are used for each variable. If false all fields are held constant beyond the extreme model levels. NLPR 0 Number of isobaric levels used for interpolated data with maximum NL. If zero or negative, NL isobaric levels are set up having p(k)=P0*SIGMA(k). PLOUT NL*0. Pressures in mb of the NLPR isobaric levels in order of increasing pressure. Only needed for NLPR > 0. SWITCHES AND DATA CONTROLLING USE OF OROGRAPHY. LROG .F. Switch to include orography. LROGGR .F. Switch to print lat-long distribution of orography. LROGPL .F. Switch to plot lat-long distribution of orography. ROGFAC 10. Factor, in metres, giving units in which orography is printed. ROGINC 200. Contour interval for orography in metres. LMSK2 .F. Switch to plot orographic outline or "mask" on type 2 output lat-long fields. Only relevant for isobaric level output (LINTP2=.T.). LMSK3 .F. Switch to plot lines of mean and maximum orographic height on zonal mean sections. Only relevant for isobaric level output (LINTP3=.T.). SWITCHES AND DATA FOR OUTPUT OF SPECTRAL COEFFICIENTS. NCOEFF NN Highest total wavenumber for which spectral coeff- icients are printed. Also acts as switch since zero value suppresses output. NPC1 3 Channel on which spectral coefficients are printed. LSPO NL*.F. Selection of levels for which spectral coefficients are printed. If all elements are false only (log) surface pressure is printed. SWITCHES AND DATA FOR OUTPUT OF FIELDS ON MODEL/ISOBARIC LEVELS. LPC2 .F. Switch to output fields on model/isobaric levels (choice of level type depends on LINTP2). NPC2 3 Channel on which model/isobaric level fields are printed. LSGEDY .F. Switch to remove zonal mean from these printed and plotted fields. LSGGRZ .T. Switch to print zonal averages of these fields. Values are printed at all levels, irrespective of LGPO below. LSGPLZ .F. Switch to plot zonal averages of these fields. LGPO NL*.F. Selection of levels for which model/isobaric level fields are printed. If all elements are false only single level fields (surface and msl pressures) are printed. LGFO NL*.F. Selection of levels for which model/isobaric level fields are plotted. If all elements are false only single level fields (surface and msl pressures) are plotted. LSGGR NSG*(..) Selection of variables to be printed. LSGPL NSG*(..) Selection of variables to be plotted. SGFAC NSG*(..) Dimensional factor for each variable giving units in which field is printed (see table below). Overridden if contour interval is negative. SGINC NSG*(..) If positive, dimensional contour interval for each variable (see table). If negative, field is both printed and plotted normalised. Field Field Units LSGGR LSGPL SGFAC SGINC No. Name 1 Stream function m2/s .T. .T. 1.E6 1.E7 2 Relative vorticity s-1 .T. .T. 1.E-6 1.E-5 3 Divergence s-1 .T. .F. 1.E-7 1.E-6 3 Omega mb/hr .T. .T. 0.1 1.0 4 Zonal wind m/s .T. .F. 0.1 0.0 5 Meridional wind m/s .T. .F. 0.1 0.0 7 Temperature deg C .T. .T. 0.1 4.0 8 Geopotential height metre .T. .F. 10.0 200.0 9 Specific humidity g/kg .T. .T. 0.1 1.0 10 Diabatic heating K/day .T. .T. 0.1 1.0 11 Surface pressure mb .F. .F. 1.0 20.0 12 MSL pressure - P0 mb .T. .T. 0.1 4.0 SWITCHES AND DATA FOR OUTPUT OF ZONALLY AVERAGED DYNAMICAL FIELDS. LPC3 .T. Switch to output zonally averaged dynamical fields on model/isobaric levels (choice of level type depends on LINTP3). Both time series and time- averages are output. NPC3 3 Channel on which zonal averages are printed. LXDGR NXD*(..) Selection of variables to be printed. LXDPL NXD*(..) Selection of variables to be plotted. XDFAC NXD*(..) Dimensional factor for each variable giving units in which field is printed (see table below). Overridden if contour interval is negative. XDINC NXD*(..) If positive, dimensional contour interval for each variable (see table). If negative, field is both printed and plotted normalised. Field Field Units LXDGR LXDPL XDFAC XDINC No. Name 1 Zonal wind m/s .T. .T. 0.1 5.0 2 Potential temperature Kelvin .T. .T. 0.1 5.0 3 Temperature deg C .T. .T. 0.1 5.0 4 Mean meridional circ. kg/s .T. .T. 1.E9 -0.1 5 Horiz. momentum flux (m/s)2 .T. .T. 1.0 -0.1 6 Vert. momentum flux (m/s)(mb/hr) .T. .T. 0.1 -0.1 7 Horiz. temperature flux K(m/s) .T. .T. 1.0 -0.1 8 Vert. temperature flux K(mb/hr) .T. .T. 0.1 -0.1 9 Zonal kinetic energy J/kg .T. .T. 10.0 -0.1 10 Eddy kinetic energy J/kg .T. .T. 10.0 -0.1 11 Horiz. E-P flux m3(rad) .T. .F. 1.E14 0.0 12 Vert. E-P flux m3(Pa) .T. .F. 1.E19 0.0 13 Horiz. E-P flux div. m3 .T. .F. 1.E14 0.0 14 Vert. E-P flux div. m3 .T. .F. 1.E14 0.0 15 Total E-P flux div. m3 .T. .T. 1.E14 4.E15 16 Static stability (N**2) s-2 .T. .T. 1.E-5 5.E-5 17 PV gradient (s-1)(m-1) .T. .T. 1.E-12 2.E-11 18 Total permitted waveno. .T. .T. 0.1 2.0 19 Temperature anomaly Deg C .T. .T. 0.1 1.0 20 Mean vertical motion mb/hr .T. .T. 0.1 -0.1 21 Vertical mean heat flux K(mb/hr) .T. .F. 1.0 -0.1 22 Mean meridional wind m/s .T. .T. 0.01 -0.1 23 Horiz mean heat flux K(m/s) .T. .F. 1.0 -0.1 24 Specific humidity g/kg .T. .T. 0.1 1.0 25 Relative humidity Per cent .T. .T. 1.0 10.0 26 Horiz eddy moisture flx (g/kg)(m/s) .T. .T. 0.1 -0.1 27 Vert eddy moisture flux (g/kg)(mb/hr) .T. .T. 0.1 -0.1 28 Diabatic heating K/day .T. .T. 0.1 0.5 SWITCHES AND DATA FOR OUTPUT OF KINETIC ENERGY SPECTRA. (KOUNTE) - Frequency at which spectra are output. Acts as switch, with zero value suppressing output. See counters above. NPC4 3 Channel on which KE spectra are printed. LSPPLT .F. Switch to copy both m- and n- KE spectra to channel NPLOT(1) for later plotting by the MPLOTS program. This plot file should be saved if it is to be used in a separate job. SWITCHES AND DATA FOR OUTPUT OF THE GLOBAL ENERGETICS TIME SERIES. LPC5 .T. Switch to output global energetics time series (every KOUNTD steps) and time averages. NPC5 3 Channel on which global energetics are printed. Also copied to output channel by default. LENCOP .F. Switch to copy time series of global energetics to channel NPLOT(1) for later plotting by the MPLOTS program. This plot file should be saved if it is to be used in a separate job. XUEN 0.2 Scaling factor for x-axis on energetics plots. Set equal to (ENDDAY-BEGDAY)/N to fill frame, with N=500 for 35mm film and N=250 for 16mm. IYEN 20 No. of ticks on y-axis of energy value plots. Should be related to ENRNGE. IYCON 20 No. of ticks on y-axis of energy conversion plots. Should be related to (ECONMX-ECONMN). ECONMX 10. Maximum and minimum values allowed for on the y-axis ECONMN -10. of the time series plot of energy conversions. Units of W/m2. ENRNGE 200.E4 Range of values allowed for on the y-axis of the time series plot of energy quantities. Units of J/m2. ENMN 5*(..) Minimum value allowed for on the y-axis of the time series plot of each energy quantity, in units J/m2. Defaults are: Variable Variable ENMN Number Name 1 NS 1500.E4 2 NZ 150.E4 3 NE 0. 4 KZ 0. 5 KE 0. SWITCHES AND DATA FOR OUTPUT OF POTENTIAL VORTICITY AND RELATED DIAGNOSTICS ON ISENTROPIC SURFACES. LPC6 .F. Switch to output fields on isentropic levels. NPC6 3 Channel on which isentropic fields are printed. THSURD NTHSF*() Values of potential temperature on the isentropic surfaces in Kelvin. Default is for 5 levels with theta = 350., 330., 310., 295., 280.. CPHASE 0.0 Eastward phase-speed in degrees/day subtracted to obtain isentropic relative flow. LTHGR NTH*(..) Selection of fields to be printed. LTHPL NTH*(..) Selection of fields to be plotted. The first two elements refer to U and V and must be false. THFAC NTH*(..) Dimensional factor for each variable giving units in which field is printed (see table below). Overridden if contour interval is negative. THINC NTH*(..) If positive, dimensional contour interval for each variable (see table). If negative, field is both printed and plotted normalised. * Note that PV uses double the specified contour interval at the top two levels and half it below the third level (see routine OPERTEL). Field Field Units LTHGR LTHPL THFAC THINC No. Name 1 Zonal wind - cphase m/s .F. .F. 0.1 0.0 2 Meridional wind m/s .F. .F. 0.1 0.0 3 Potential Vorticity PV-units .F. .F. 0.1 * 0.5 * 2 Pressure mb .F. .F. 1.0 50.0 SWITCHES AND DATA FOR OUTPUT OF SURFACE FIELDS ASSOCIATED WITH THE PARAMETERISED PHYSICAL PROCESSES. BOTH INSTANTANEOUS AND KOUNTD- ACCUMULATED FIELDS ARE OUTPUT. LPC7 .F. Switch to output surface fields associated with parameterisations. NPC7 3 Channel on which surface fields are printed. LPHGR NPH*(..) Selection of fields to be printed. LPHPL NPH*(..) Selection of fields to be plotted. If both convective and large scale rain are included, both are plotted on the same frame. PHFAC NPH*(..) Dimensional factor for each variable giving units in which field is printed (see table below). Overridden if contour interval is negative. PHINC NPH*(..) If positive, dimensional contour interval for each variable (see table). If negative, field is both printed and plotted normalised. Field Field Units LPHGR LPHPL PHFAC PHINC No. Name 1 Surface stress N/m2 .F. .F. 0.01 0.1 2 Surf sensible heat flux W/m2 .F. .F. 1.0 25.0 3 Surf latent heat flux W/m2 .F. .F. 1.0 25.0 4 Convective rain mm/day .F. .F. 0.1 5.0 5 Large scale rain mm/day .F. .F. 0.1 5.0 6 Total rainfall mm/day .F. .F. 0.1 5.0 SWITCHES AND DATA FOR OUTPUT OF ZONAL SECTIONS OF PHYSICS TENDENCIES. ONLY INSTANTANEOUS FIELDS ARE OUTPUT AT EACH ANALYSIS TIME, WHILE THE TIME AVERAGES USE THE KOUNTD-ACCUMULATED FIELDS. LPC8 .T. Switch to output zonally averaged tendencies of physical processes on model levels. Both time series and time-averages are output. NPC8 3 Channel on which zonal averages are printed. LXPGR NXP*(..) Selection of variables to be printed. LXPPL NXP*(..) Selection of variables to be plotted. XPFAC NXP*(..) Dimensional factor for each variable giving units in which field is printed (see table below). Overridden if contour interval is negative. XPINC NXP*(..) If positive, dimensional contour interval for each variable (see table). If negative, field is both printed and plotted normalised. Field Field Units LXPGR LXPPL XPFAC XPINC No. Name 1 U-tend Vert diffusion m/s/day .T. .F. 0.1 5.0 2 V-tend Vert diffusion m/s/day .T. .F. 0.1 5.0 3 T-tend Vert diffusion K/day .T. .F. 0.1 0.5 4 Q-tend Vert diffusion g/kg/day .T. .F. 0.1 0.5 5 T-tend Convection K/day .T. .F. 0.1 0.5 6 Q-tend Convection g/kg/day .T. .F. 0.1 0.5 7 T-tend Large scale rain K/day .T. .F. 0.1 0.5 8 Q-tend Large scale rain g/kg/day .T. .F. 0.1 0.5 9 T-tend Radiation K/day .T. .F. 0.1 0.5 10 U-tend Boundary layer m/s/day .T. .F. 0.1 5.0 11 V-tend Boundary layer m/s/day .T. .F. 0.1 5.0 12 T-tend Boundary layer K/day .T. .F. 0.1 0.5 13 Q-tend Boundary layer g/kg/day .T. .F. 0.1 0.5 14 U-tend Total m/s/day .T. .T. 0.1 5.0 15 V-tend Total m/s/day .T. .T. 0.1 5.0 16 T-tend Total K/day .T. .T. 0.1 0.5 17 Q-tend Total g/kg/day .T. .T. 0.1 0.5 18 Convection counter }Sum along .T. .F. 1.0 0.0 19 Large scale rain counter}lat. circle .T. .F. 1.0 0.0 SWITCHES AND DATA FOR OUTPUT OF THE GLOBALLY AVERAGED SURFACE DIAGNOSTICS. THE TIME SERIES OF BOTH INSTANTANEOUS AND ACCUMULATED VALUES ARE PRINTED. TIME-AVERAGES USE THE ACCUMULATED FIELDS ONLY. LPC9 .T. Switch to output globally averaged physics/surface diagnostics. Both time series (every KOUNTD steps) and time averages are output. NPC9 3 Channel on which global surface diagnostics are printed. Also copied to output channel by default. LPHCOP .F. Switch to copy time series of global surface diag- nostics to channel NPLOT(1) for later plotting by the MPLOTS program. This plot file should be saved if it is to be used in a separate job. DATA FOR NAMELIST INPHYS. ------------------------- Data describing the parameterised physical processes included in the run or for which diagnostics are required must be supplied through namelist group INPHYS. The main use of these switches is to choose which surface fields and zonally averaged tendencies are output. However the relevant switches must be set to obtain KE dissipation rates in the global energetics, since copies of the model's boundary layer and vertical diffusion codes are included in routine NTEND to calculate the momentum tendencies. Any change in these model parameterisations must be reflected in NTEND. In contrast diabatic heating is read from the model history file. VARIABLE DEFAULT INTERPRETATION LDRY .F. Switch to process a history file from the "dry" version of the spectral model. The default is to process a "moist" model history file. If LDRY is true, it is assumed there is only a single spectral record at each analysis time which omits specific humidity. All physical parameterisation diagnostics are omitted and only horizontal diffusion and linear restoration/drag are optionally included. Output of moisture and heating fields on model/isobaric levels is also switched off. The remaining variables in INPHYS are then irrelevant. BEGDP 0.0 Day at which physical parameterisations were switch- ed on in the model integration. The next analysis time is assumed to be the first at which there are (JG+3) history records rather than a single spectral record at each time. LBL .F. Switch to include boundary layer diagnostics. LVD .F. Switch to include vertical diffusion diagnostics LCR .F. Switch to include convection diagnostics. LLR .F. Switch to include large scale rain diagnostics. LRD .F. Switch to include radiation diagnostics. CD 0.001 Drag coefficient in bulk aerodynamic formulation of surface fluxes. AKVV 1. Vertical diffusion coefficient in m2/s. ************************************************************************