Output data¶
The output data of flex_extract can be divided into the final FLEXPART input files and temporary files:
FLEXPART input files |
Temporary files (saved in debug mode) |
|
|
FLEXPART input files¶
The final output files of flex_extract are the meteorological input files for FLEXPART.
The naming convention for these files depends on the kind of data extracted by flex_extract.
Standard output files¶
In general, there is one file for each time named:
<prefix>YYMMDDHH
where YY are the last two digits of the year, MM is the month, DD the day, and HH the hour (UTC). <prefix> is by default defined as EN, and can be re-defined in the CONTROL file.
Each file contains all meteorological fields at all levels as needed by FLEXPART, valid for the time indicated in the file name.
Here is an example output which lists the meteorological fields in a single file called CE00010800 (where we extracted only the lowest model level for demonstration purposes):
$ grib_ls CE00010800
edition centre date dataType gridType stepRange typeOfLevel level shortName packingType
2 ecmf 20000108 an regular_ll 0 hybrid 91 u grid_simple
2 ecmf 20000108 an regular_ll 0 hybrid 91 v grid_simple
2 ecmf 20000108 an regular_ll 0 hybrid 91 etadot grid_simple
2 ecmf 20000108 an regular_ll 0 hybrid 91 t grid_simple
2 ecmf 20000108 an regular_ll 0 surface 1 sp grid_simple
2 ecmf 20000108 an regular_ll 0 hybrid 91 q grid_simple
2 ecmf 20000108 an regular_ll 0 hybrid 91 qc grid_simple
1 ecmf 20000108 fc regular_ll 0 surface 0 sshf grid_simple
1 ecmf 20000108 fc regular_ll 0 surface 0 ewss grid_simple
1 ecmf 20000108 fc regular_ll 0 surface 0 nsss grid_simple
1 ecmf 20000108 fc regular_ll 0 surface 0 ssr grid_simple
1 ecmf 20000108 fc regular_ll 0 surface 0 lsp grid_simple
1 ecmf 20000108 fc regular_ll 0 surface 0 cp grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 sd grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 msl grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 tcc grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 10u grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 10v grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 2t grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 2d grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 z grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 lsm grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 cvl grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 cvh grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 lcc grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 mcc grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 hcc grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 skt grid_simple
1 ecmf 20000108 an regular_ll 0 depthBelowLandLayer 0 stl1 grid_simple
1 ecmf 20000108 an regular_ll 0 depthBelowLandLayer 0 swvl1 grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 sr grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 sdor grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 cvl grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 cvh grid_simple
1 ecmf 20000108 an regular_ll 0 surface 0 fsr grid_simple
35 of 35 messages in CE00010800
Output files for long forecast¶
Flex_extract is able to retrieve forecasts with a lead time of more than 23 hours. In order to avoid collisions of time steps names, a new scheme for filenames in long forecast mode is introduced:
<prefix>YYMMDD.HH.<FORECAST_STEP>
The <prefix> is, as in the standard output, by default EN and can be re-defined in the CONTROL file. YYMMDD is the date format and HH the forecast time which is the starting time for the forecasts. The FORECAST_STEP is a 3-digit number which represents the forecast step in hours.
Output files for ensemble predictions¶
Flex_extract is able to retrieve ensembles data; they are labelled by the grib message parameter number. Each ensemble member is saved in a separate file, and standard filenames are supplemented by the letter N and the ensemble member number in a 3-digit format.
<prefix>YYMMDDHH.N<ENSEMBLE_MEMBER>
Additional fields with new precipitation disaggregation¶
The new disaggregation method for precipitation fields produces two additional precipitation fields for each time step and precipitation type (large-scale and convective). They serve as sub-grid points in the original time interval. For details of the method see Disaggregation of flux data.
The two additional fields are addressed using the step parameter in the GRIB messages, which
is set to “1” or “2”, for sub-grid points 1 and 2, respectively.
The output file names are not altered.
An example of the list of precipitation fields in an output file generated with the new disaggregation method is found below:
$ grib_ls
edition centre date dataType gridType stepRange typeOfLevel level shortName packingType
1 ecmf 20000108 fc regular_ll 0 surface 0 lsp grid_simple
1 ecmf 20000108 fc regular_ll 1 surface 0 lsp grid_simple
1 ecmf 20000108 fc regular_ll 2 surface 0 lsp grid_simple
1 ecmf 20000108 fc regular_ll 0 surface 0 cp grid_simple
1 ecmf 20000108 fc regular_ll 1 surface 0 cp grid_simple
1 ecmf 20000108 fc regular_ll 2 surface 0 cp grid_simple
Temporary files¶
Flex_extract creates a number of temporary data files which are usually deleted at the end of a successful run. They are preserved only if the DEBUG mode is switched on (see The CONTROL parameters).
MARS grib files¶
Flex_extract retrieves all meteorological fields from MARS and stores them in files ending with .grb.
Since there are limits implemented by ECMWF for the time per request and data transfer from MARS,
and as ECMWF asks for efficient MARS retrievals, flex_extract splits the overall data request
into several smaller requests. Each request is stored in its own .grb file, and the file
names are composed of several pieces of information:
<field_type><grid_type><temporal_property><level_type>.<date>.<ppid>.<pid>.grb
Description:
- Field type:
AN- Analysis,FC- Forecast,4V- 4D variational analysis,CV- Validation forecast,CF- Control forecast,PF- Perturbed forecast- Grid type:
SH- Spherical Harmonics,GG- Gaussian Grid,OG- Output Grid (typically lat / lon),_OROLSM- Orography parameter- Temporal property:
__- instantaneous fields,_acc- accumulated fields- Level type:
ML- model level,SL- surface level- ppid:
- The process number of the parent process of the script submitted.
- pid:
- The process number of the script submitted.
Example .grb files for one day of CERA-20C data:
ANOG__ML.20000908.71851.71852.grb FCOG_acc_SL.20000907.71851.71852.grb ANOG__SL.20000908.71851.71852.grb OG_OROLSM__SL.20000908.71851.71852.grb ANSH__SL.20000908.71851.71852.grb
MARS request file¶
This file is a csv file called mars_requests.csv listing the actual settings of the MARS
request (one request per line) in a flex_extract job.
It is used for documenting which data were retrieved, and for testing.
Each request consists of the following parameters, whose meaning mostly can be taken from The CONTROL parameters or The executable script - run.sh: request_number, accuracy, area, dataset, date, expver, gaussian, grid, levelist, levtype, marsclass, number, param, repres, resol, step, stream, target, time, type
Example output of a one-day retrieval of CERA-20C data:
request_number, accuracy, area, dataset, date, expver, gaussian, grid, levelist, levtype, marsclass, number, param, repres, resol, step, stream, target, time, type
1, 24, 40.0/-5.0/30.0/5.0, None, 20000107/to/20000109, 1, , 1.0/1.0, 1, SFC, EP, 000, 142.128/143.128/146.128/180.128/181.128/176.128, , 159, 3/to/24/by/3, ENDA, /mnt/nas/Anne/Interpolation/flexextract/flex_extract_v7.1/run/./workspace/CERA_testgrid_local_cds/FCOG_acc_SL.20000107.23903.23904.grb, 18, FC
1, 24, 40.0/-5.0/30.0/5.0, None, 20000108/to/20000108, 1, , 1.0/1.0, 85/to/91, ML, EP, 000, 130.128/133.128/131.128/132.128/077.128/246.128/247.128, , 159, 00, ENDA, /mnt/nas/Anne/Interpolation/flexextract/flex_extract_v7.1/run/./workspace/CERA_testgrid_local_cds/ANOG__ML.20000108.23903.23904.grb, 00/03/06/09/12/15/18/21, AN
2, 24, 40.0/-5.0/30.0/5.0, None, 20000108/to/20000108, 1, , OFF, 1, ML, EP, 000, 152.128, , 159, 00, ENDA, /mnt/nas/Anne/Interpolation/flexextract/flex_extract_v7.1/run/./workspace/CERA_testgrid_local_cds/ANSH__SL.20000108.23903.23904.grb, 00/03/06/09/12/15/18/21, AN
3, 24, 40.0/-5.0/30.0/5.0, None, 20000108, 1, , 1.0/1.0, 1, SFC, EP, 000, 160.128/027.128/028.128/244.128, , 159, 000, ENDA, /mnt/nas/Anne/Interpolation/flexextract/flex_extract_v7.1/run/./workspace/CERA_testgrid_local_cds/OG_OROLSM__SL.20000108.23903.23904.grb, 00, AN
4, 24, 40.0/-5.0/30.0/5.0, None, 20000108/to/20000108, 1, , 1.0/1.0, 1, SFC, EP, 000, 141.128/151.128/164.128/165.128/166.128/167.128/168.128/129.128/172.128/027.128/028.128/186.128/187.128/188.128/235.128/139.128/039.128/173.128, , 159, 00, ENDA, /mnt/nas/Anne/Interpolation/flexextract/flex_extract_v7.1/run/./workspace/CERA_testgrid_local_cds/ANOG__SL.20000108.23903.23904.grb, 00/03/06/09/12/15/18/21, AN
Index file¶
This file is called date_time_stepRange.idx. It contains indices pointing to specific grib messages from one or more grib files. The messages are selected with a composition of grib message keywords.
Flux files¶
The flux files contain the de-accumulated and dis-aggregated flux fields of large-scale and convective precipitation, east- and northward turbulent surface stresses, the surface sensible heat flux, and the surface net solar radiation.
flux<date>[.N<xxx>][.<xxx>]
The date format is YYYYMMDDHH as explained before. The optional block [.N<xxx>] is used for the ensemble forecast date, where <xxx> is the ensemble member number. The optional block [.<xxx>] marks a pure forecast with <xxx> being the forecast step.
Note
In the case of the new dis-aggregation method for precipitation, two new sub-intervals are added in between each time interval. They are identified by the forecast step parameter which is 0 for the original time interval, and 1 or 2, respectively, for the two new intervals.
fort files¶
There are a number of input files for the calc_etadot Fortran program named
fort.xx
where xx is a number which defines the meteorological fields stored in these files.
They are generated by the Python code in flex_extract by splitting the meteorological fields for a unique time stamp from the *.grb files, storing them under the names fort.<XX> where <XX> represents some number.
The following table defines the numbers and the corresponding content:
| Number | Content |
|---|---|
| 10 | U and V wind components |
| 11 | temperature |
| 12 | logarithm of surface pressure |
| 13 | divergence (optional) |
| 16 | surface fields |
| 17 | specific humidity |
| 18 | surface specific humidity (reduced Gaussian grid) |
| 19 | omega (vertical velocity in pressure coordinates) (optional) |
| 21 | eta-coordinate vertical velocity (optional) |
| 22 | total cloud-water content (optional) |
Some of the fields are solely retrieved with specific settings, e. g., the eta-coordinate vertical velocity is not available in ERA-Interim datasets, and the total cloud-water content is an optional field which is useful for FLEXPART v10 and newer.
The calc_etadot program saves its results in file fort.15 which typically contains:
| Number | Content |
|---|---|
| 15 | U and V wind components, eta-coordinate vertical velocity, temperature, surface pressure, specific humidity |
More details about the content of calc_etadot can be found in Vertical wind.
Note
The fort.4 file is the namelist file to control the Fortran program calc_etadot. It is therefore also an input file.
Example of a namelist:
&NAMGEN
maxl = 11,
maxb = 11,
mlevel = 91,
mlevelist = "85/to/91",
mnauf = 159,
metapar = 77,
rlo0 = -5.0,
rlo1 = 5.0,
rla0 = 30.0,
rla1 = 40.0,
momega = 0,
momegadiff = 0,
mgauss = 0,
msmooth = 0,
meta = 1,
metadiff = 0,
mdpdeta = 1
/