Afterburner

NAME

after - transforms and selects data from T-model GRIB or netCDF files

SYNOPSIS

after [ -options ] infiles resultfile < namelist

DESCRIPTION

The program "after" reads, unpacks, transforms and selects data

from T-model GRIB/netCDF files, and writes them to

SERVICE/GRIB/netCDF format for further processing (i.e. plotting).

Additionally, pressure or height interpolation is optional

(refer to parameter description of "UNITSEL").

Possible combinations of TYPE, CODE and MEAN

TYPE = 0 : CODE = 130 temperature MEAN = 0

131 u-velocity

132 v-velocity

133 specific humidity

138 vorticity

148 streamfunction

149 velocity potential

152 LnPs

155 divergence

TYPE = 10: CODE = (same as TYPE 0) MEAN = 0

TYPE = 11: CODE = (same as TYPE 0) MEAN = 0

TYPE > 11: CODE = (all codes) MEAN = 0/1

Vorticity, divergence, streamfunction and velocity potential

need special treatment in the vertical transformation.

They are not available as types 30, 40 and 41.

If you select one of these combinations, type is automatically

switched to the equivalent types 70, 60 and 61. The type of all

other variables will be switched too, because the type is a

global parameter.

OPTIONS

-P <nthreads> : Set number of OpenMP threads

-c : Print available codes and names

-d : Debug mode

NAMELIST

&SELECT

TYPE = , CODE = , LEVEL = , MULTI = , MEAN = , INTERVAL = ,

TIMESEL = , UNITSEL = , FORMAT = , PRECISION = , SZIP =

&END

default -> TYPE = 0 : Hybrid level spectral coefficients

TYPE = 10 : Hybrid level fourier coefficients

TYPE = 11 : Hybrid level zonal mean sections

TYPE = 20 : Hybrid level gauss grids

TYPE = 30 : Pressure or Height level gauss grids

TYPE = 40 : Pressure or Height level fourier coefficients

TYPE = 41 : Pressure or Height level zonal mean sections

TYPE = 50 : Pressure or Height level spectral coefficients

TYPE = 60 : Pressure or Height level fourier coefficients

TYPE = 61 : Pressure or Height level zonal mean sections

TYPE = 70 : Pressure or Height level gauss grids

default -> CODE = -1 : process all detected codes

CODE = 1 - 255 ECMWF field codes (130=Temperature, etc.)

256 - 258 : internal

259 : sqrt(u*u + v*v)

260 : largescale + convective rain

261 : total top radiation

262 : total bottom radiation

263 : net heat flux

264 : net water flux

default -> LEVEL = -1 : process all detected levels

LEVEL =: Up to NN + 1 hybrid, pressure or height levels

(depending on TYPE and UNITSEL !!!)

NN depends on model type (e.g. 19, 31 or 39)

default -> MULTI = 0 : Process only one input file

MULTI = n : Process "n" input files - each containing

one month. Put only the name of the first

input file on the command line. All subsequent

files are expected to be in the same directory.

The filenames must follow the syntax

experiment_YYMM, where YY denotes the year

and MM the month.

default -> DAYIN = 30 : Day step for incrementing input file names

DAYIN = nn : Day step is nn days

Used only in conjunction with MULTI != 0.

Since ECHAM4 output files follow the

naming convention

experiment_YYMM.DD

DAYIN is the time step in days between

subsequent input files for MULTI != 0.

default -> MEAN = 0 : Write out all terms

MEAN = 1 : Compute and write monthly or daily mean fields

!!! Only for TYPE >= 20 !!!

MEAN = 2 : Compute and write monthly or daily deviations.

!!! Only for TYPE = 20 or TYPE = 30 !!!

MEAN = 3 : Combination of MEAN=1 and MEAN=2.

Each mean field is followed by a deviation

field with an identical header record.

!!! Only for TYPE = 20 or TYPE = 30 !!!

default -> INTERVAL = 0 : Process data on monthly intervals

INTERVAL = 1 : Process data on daily intervals

e.g. To compute daily averages

set MEAN = 1 and INTERVAL = 1

INTERVAL = 2 : Process data over all inputfiles

e.g. To compute an average over

all inputfiles

set MEAN = 1 and INTERVAL = 2

default ->TIMESEL = -1 : The resultfile contains the same timesteps

per day as the input file.

Multiple values may be specified, i.e.

TIMESEL = 0, 12, 18

Note that only hours existing in the input

file can be specified.

(No time interpolation or average is done)

default ->UNITSEL = 0 : Pressure interpolation (TYPE >= 30)

or hybrid level output (TYPE <= 20).

Unlike before, pressure levels less than 100 Pa

can be processed (except 0 and negative values).

UNITSEL = 1 : Height interpolation

Values specified for LEVEL are interpreted

as height-meters (TYPE >= 30).

(Maximum LEVEL: 65535 m)

UNITSEL = 2 : Height interpolation

Values specified for LEVEL are interpreted

as height-kilometers (TYPE >= 30).

(Maximum LEVEL: 65 km)

default -> FORMAT = 0 : SERVICE

FORMAT = 1 : GRIB

FORMAT = 2 : netCDF

FORMAT = 3 : EXTRA

FORMAT = 4 : netCDF version 2

FORMAT = 6 : netCDF version 4

default -> PRECISION = 0 : use input data precision for output

PRECISION = 8 : 8 bit packed output data precision

PRECISION = 16 : 16 bit packed output data precision

PRECISION = 24 : 24 bit packed output data precision

PRECISION = 32 : 32 bit floating point output data precision

PRECISION = 64 : 64 bit floating point output data precision

default -> SZIP = 0 : No SZIP compression of GRIB records

SZIP = 1 : SZIP compression of GRIB records

EXTRAPOLATE = 0 : No extrapolation of missing values during the

interpolation from model to pressure or height level

(only available with MEAN = 0 and TYPE = 30)

default -> EXTRAPOLATE = 1 : Extrapolate missing values

EXAMPLES

To create monthly mean over all codes and hybrid levels in GRIB

format, try

after Model-file GRIB-file.avg << EOF

&SELECT TYPE=20, MEAN=1, FORMAT=1 &END

EOF

To select code 129 and 130 and interpolate to pressure levels

100, 250, 500, 850, 925 and 1000 hPa, try

after Model-file GRIB-file.sel << EOF

&SELECT

TYPE=70, FORMAT=1,

CODE=129,130,

LEVEL=10000, 25000, 50000, 85000, 92500, 100000

&END

EOF

To calculate u and v velocity from vorticity and divergence

on model level, try

after Model-file GRIB-file.sel << EOF

&SELECT

TYPE=20, FORMAT=1,

CODE=131,132

&END

EOF