observation pre-processing in the arpege/aladin model
DESCRIPTION
Observation pre-processing in the ARPEGE/ALADIN model. Roger Randriamampianina Hungarian Meteorological Service ALADIN 3D-Var/ODB training course, Budapest, 6-10 June, 2006. Outline. Local pre-processing:. OULAN . BATOR. 1C RAD. BIAS CORRECTION. OULAN . - PowerPoint PPT PresentationTRANSCRIPT
Observation pre-processing Observation pre-processing in in thethe ARPEGE/ALADIN modelARPEGE/ALADIN model
Roger RandriamampianinaRoger Randriamampianina Hungarian Meteorological ServiceHungarian Meteorological Service
ALADIN 3D-Var/ODB training course, Budapest, 6-10 June, 2006ALADIN 3D-Var/ODB training course, Budapest, 6-10 June, 2006
Local pre-processing:
OULAN
BATOR
1C RAD. BIAS CORRECTION
OutlineOutline
OULAN
Source: www.wmo.intSource: www.wmo.int
Real time available data:Real time available data: NetworksNetworks ObservatioObservationns arrive through s arrive through - GTS (Global Telecommunication - GTS (Global Telecommunication
System) in ASCII or BUFR formatSystem) in ASCII or BUFR format - Direct Sat. Reception – (Budapest HRPT - Direct Sat. Reception – (Budapest HRPT
Station)Station) - Special TCP/IP lines- Special TCP/IP lines - Internet FTP in some cases - Internet FTP in some cases
ProvidersProviders - All WMO countries for most of the - All WMO countries for most of the
observation typesobservation types - Space agencies for Satellite observations - Space agencies for Satellite observations
(Eumetsat, ESA, NESDIS, NASA, (Eumetsat, ESA, NESDIS, NASA, DMSP)DMSP)
- NMS for the local national data- NMS for the local national data
TypeTypess of observations of observations
Conventional ObservationsConventional Observations Surface data:Surface data: - - SynopSynop, Ship, Ship - Bathy, Tesac- Bathy, Tesac - Buoy- Buoy Upperair data:Upperair data: - Airep, - Airep, AmdarAmdar, Acar, Acar - - TempTemp, Temp-ship, Temp-mobil, , Temp-ship, Temp-mobil,
Temp-dropTemp-drop - Pilot, Pilot-ship, - Pilot, Pilot-ship, EuroprofilEuroprofil, Profiler, Profiler - Satob, Satgeo, - Satob, Satgeo, geowind (MET-8)geowind (MET-8)
Satellite ObservationsSatellite Observations MeteosatMeteosat:: - - SEVIRISEVIRI Clear sky radiancesClear sky radiances NOAA Atovs:NOAA Atovs: - - Hirs, Amsua, Amsub, MhsHirs, Amsua, Amsub, Mhs DMSP:DMSP: - Ssm/I- Ssm/I AQUAAQUA:: -Airs-Airs
GTS
Eumetcast
…
FTP
BUFR
Intermediate ASCII
Or direct reading
oulan
Data pre-processing at HMSData pre-processing at HMS
Pre-processing chainPre-processing chain::
oulan bator
ASCII
…
screening
Reading Reading of of datadataand and coding of coding of additional additional informationinformation
Blacklisting andBlacklisting andcreation of the ODBcreation of the ODB
Complex quality Complex quality check and data check and data
selectionselection
Observation usage (in surface analysis)Observation usage (in surface analysis)
Observation usage (in variational analysis)Observation usage (in variational analysis)
Message Message typetype
Observation typeObservation type Extracted Extracted parameters parameters
Used parametersUsed parameters Parameters Parameters blacklisted or not blacklisted or not usedused
1C level Radiances
Observation: multichannels Polar Satellites
•Tb •Sat identifier•Scanning line•Scan angle (number of FOV)•Zenith angle•Azimuth angle•Satellite altitude
•Tb in some channels, depending on the nature of the surface (land, sea or ice), cloud fraction and model orography
•Tb in some channels, depending on the nature of the surface (land, sea or ice), cloud fraction and model orography
Channel numberChannel number 11 22 33 44 55 66 77 88 99 1010 1111 1212 1313 1414 1515
Over LandOver Land xx xx xx xx xx xx xx xx
Over SeaOver Sea xx xx xx xx xx xx xx xx
Over Sea iceOver Sea ice xx xx xx xx xx xx
Cloudy pixelCloudy pixel xx xx xx xx xx
Over land channels 5 and 6 are used when the model orography is less than 500m and 1500m, respectivelyOver land channels 5 and 6 are used when the model orography is less than 500m and 1500m, respectively
Example: Use of RAD1C AMSU-AExample: Use of RAD1C AMSU-A
Observation usage (Satellite radiances)
Read_Synop
Ext_lam_Synop
LSYNOP
Read_Amdar
Ext_lam_Temp
LTEMP
Read_Amsub
Ext_lam_Amsub
LTOVSAMSUB
…
…
…
OULAN
NamelistNamelistcontrolcontrol
OBSOUL
You need You need to changeto changein case of in case of
local local implementationimplementation
The oulan procedureThe oulan procedure
oulan_lam_extract
Oulan: from the export version under “UTI” directoryThe added files:
ext_amsua_aapp.f – read the native sat file (AMSU-A)ext_amsub_aapp.f – ==========x ==== (AMSU-B)ext_hirs_aapp.f – ==========x ==== (HIRS)ext_synop_netcdf.f – read the netCDF SYNOP dataext_temp_netcdf.f – read the netCDF TEMP dataoulan_lam_elev.f – read the elevation data from TLSnetcdf.inc – netCDF include fileama1c.h, amb1c.h, hirs1c.h, netcdf.h – include files
Modifications compared to the original codeModifications compared to the original code
The modified files: ext_{obs_type}.f ext_lam_{obs_type}.fWhere obs_type:
synop, temp, geowind, satob, europrofil, amsua, amsub, hirs
!!! There are two versions of the oulan_carobs in the source directory No change in the OBSOUL format Change in case of new obs. (eg. AMSU-B) (cy24 vs cy28) The number of the arguments in oulan_carobs has changed
The surface and radiosonde observations: from netCDF files;
synop, tempThe 1C radiance data:
from AAPP native format; amsua{01-20}, amsub{01-20}, hirs (to be modified)
The aircraft observation: from ASCII (FM42) and BUFR files
read separately and arranged in amdar_ascii.dat The Atmospheric Motion Vector data:
from BUFR files; read separately and arranged in geowind_ascii.dat
The windprofiler data: from BUFR files (complicated, additional changehad to be done to read the proper station identifier);
read separately and arranged in profiler_ascii.dat
Data in BUFR format are extracted out of the oulan program ! except the AMV, data are stored in different files
Data extraction from the database at the HMSData extraction from the database at the HMS
&NADIRS NDATE=20020225, NRESO=00, ALANZA = 90., ALASZA = -90., ALOOZA = -180., ALOEZA = 180., NDIFFM1 = 30, NDIFFP1 = 30, NDIFFM2 = 300, NDIFFP2 = 259, LTOVSAMSUA = .TRUE., LTOVSAMSUB = .TRUE., LTOVSHIRS = .FALSE., LTEMP = .TRUE., LSYNOP = .TRUE., LAMDAR = .TRUE., LEUROPROFIL = .TRUE., LGEOWIND = .TRUE., LSATOB = .FALSE., NINIT = 0, LRH2Q = .TRUE., /END &NANBOB NBTOVSAMSUA = 80000, NBTOVSAMSUB = 80000, NBTOVSHIRS = 8000, NBTEMP = 1000, NBSYNOP = 4000, NBAMDAR = 9000, NBEUROPROFIL = 8000, NBSATOB = 20000, NBGEOWIND = 20000, /END
The The NAMELISTNAMELIST file file
in ASCII formatin ASCII format - each observation represent- each observation representss one article consisting of: one article consisting of:
- one header (describing the observation)- one header (describing the observation)- bodies (as many as the number of the observed/measured parameters)- bodies (as many as the number of the observed/measured parameters)
Header: - total length of the observation article integer - observation type integer - observation characteristicsobservation characteristics integer - latitude real - longitude real - station identification character - date of the measurement integer - hour of the measurement integer - altitude real - number of number of the the parametersparameters (ie. bodies) integer - observation quality flags integer - site dependant info integer
First article: - date (analysis) integer - hour (network time) integer
The OBSOUL fileThe OBSOUL file
Body (ies): - parameter type integer - first vertical coordinate real - second vertical coordinate (if necessary or other) integer - observed/measured parameter real - parameter quality flag integer
Example:20041215 1242 1 10014011 48.10000 19.51667 '12756 ' 20041215 120000 153.0000000 6 1111 100000 1 -103290.0000 0.1699999976E+39 0.0000000000E+00 2064 39 101310.0000 0.1699999976E+39 271.2600098 2048 58 101310.0000 0.1699999976E+39 82.00000000 2048 7 101310.0000 0.3211538133E-03 0.2632536227E-02 2048 41 101310.0000 3.000000000 190.0000000 2048 91 101310.0000 0.1699999976E+39 100.0000000 2048
22 2 10051144 43.52000 5.22000 'EU4426 ' 20041215 100700 182.8800049 2 11111 0 2 182.8800049 0.1699999976E+39 284.3599854 4111 3 182.8800049 4.110000134 124.0000000 4111
22 3 103254090 56.95500 -7.19100 ' 055' 20041215 114500 0.1699999976E+39 2 11111 0 200 69.00000000 0.1699999976E+39 0.1699999976E+39 9 3 30160.00000 28.79999924 296.0000000 2063
synop
Amdar
AMV/geowind
The OBSOUL fileThe OBSOUL file
To be continued …
Any questions ?
Exercises Exercises
1(a,b)- 1(a,b)- create a single T(a) or V(b) obs. form create a single T(a) or V(b) obs. form aircraft data around 500 hPaaircraft data around 500 hPa
2- 2- change the obs. errors for specific humidity change the obs. errors for specific humidity for radiosonde datafor radiosonde data
idea: check the ext_lam_temp around the variable ZQEEFH
3- 3- simulate the implementation of the PILOT obs. using simulate the implementation of the PILOT obs. using measurements from radiosondemeasurements from radiosonde
idea: copy ext_pilot.F from the “UTI/oulan” directory and follow changes in the ext_lam_temp.f. Remember, that in the pilot report we have only wind measurements.
2 and 3 need the (re-)compilation of the code. To do so, please copy the oulan directory 2 and 3 need the (re-)compilation of the code. To do so, please copy the oulan directory from ~wshop01/Preproc to your Preproc directory. Use “compile_oul” for compilation and from ~wshop01/Preproc to your Preproc directory. Use “compile_oul” for compilation and
“create_oulan” to create the executable file.“create_oulan” to create the executable file.