ITSC19 NWP Survey
Please send updates or corrections to the NWP WG co-chairs.
Table 1 Overview
Institute | Data Assimilation System | Radiative Transfer Model | ATOVS assimilated | Bias Correction Scheme | Status (applies to all tables) | |||
Global | Regional or Km Scale | Radiances | Retrievals | RARS | ||||
//EC// //(Canada)// | 4DVAR & EnKF | 3DVAR | RTTOV8.7 | YES | NO | NO | Off-line | //// 1 January 2012 |
//ECMWF// (Europe) | Hybrid 4DVAR/Ens | RTTOV10.1 | YES | NO | YES | VAR | 7March 2014 | |
met.no (Norway) | 3DVAR | RTTOV7 | YES | NO | YES | Off-line | 1 January 2012 | |
US-FNMOC/NRL (USA) | 4DVAR | 3DVAR | CRTM | YES | YES(RD) | YES | Off-line | 1 January 2012 |
//DWD// (Germany) | 3DVAR | nudging | RTTOV7 | YES | NO | YES | Off-line | //// 1 January 2012 |
//Met Office// (UK) | Hybrid 4DVAR/Ens | 4DVAR | RTTOV7 | YES | NO | YES | Off-line | //// 1 January 2012 |
DMI (Denmark) | 3DVAR | RTTOV8.5 | YES | NO | YES | Off-line | 1 March 2010 | |
//JMA// (Japan) | 4DVAR | 4DVAR | RTTOV9.3 | YES | NO | YES | VAR | 1 January 2012 |
//Meteo France// (France) | 4DVAR+Ens | 3DVAR | RTTOV9.2 | YES | NO | YES | VAR | //// 1 January 2012 |
//NCEP// (USA) | 3DVAR | 3DVAR | CRTM | YES | NO | YES | VAR | 1 January 2012 |
BoM (Australia) | 4DVAR | 4DVAR | RTTOV7 | YES | NO | NO | Off-line | 1 January 2012 |
CPTEC/INPE (Brazil) | 3DVAR | 3DVAR | N/A | NO | YES | N/A | N/A | 1 January 2012 |
//OMSZ// (Hungary) | 3DVAR | RTTOV8.5 | YES | NO | YES | VAR | 1 January 2012 | |
SMHI (Sweden) | 4DVAR | RTTOV7 | YES | NO | NO | Off-line | 1 January 2012 | |
CNMCA (Italy) | LETKF | RTTOV10.2 | YES | NO (monitoring) | YES | Off-line | 1 January 2012 | |
KMA (Korea) | 4DVAR | 4DVAR | RTTOV7 | YES | NO | YES | Off-line | 1 January 2012 |
NCMRWF (India) | 3DVAR | CRTM | YES | NO | NO | VAR | 1 January 2012 |
G global
R regional
GD global deterministic
RD regional deterministic
Table prepared by ITWG/NWP WG
Table 2 NWP Systems and Use of ATOVS
Institute | Global | Regional | Data assimilation | Radiative transfer model | ATOVS radiance assim (level) | ATOVS retrievals | RARS | ||
Deterministic | Ensemble | Deterministic | Ensemble | ||||||
EC (Canada) | 35 km L80 (10d) | 66 km L40 (20m) (16d) | 15 km L80 (2d) 2.5 km L58 | 35 km L28 (3d) | GD: 4D-Var T108 L80 GE: EnKF 100 km L58 (192m) RD: 3DVAR T200 L80 | RTTOV 8.7 | Level 1b preprocessed to 1c with AAPP | NO | NO |
ECMWF (Europe) | T1279 L137 (10d) | T639 up to 10d, T319 d10-15, L62, 51m | 4D-Var T1279 L137 with T255 final inner loop | RTTOV 10.1 | Level 1c (after antenna pattern correction) | NO | EARS, AP-RARS | ||
met.no (Norway) | 8km L60 (2.75d), 12 km L60 | 3D-Var | RTTOV 7 | Level 1c | NO | EARS | |||
FNMOC/NRL (USA) | T319 L42 (7.5d) | T119 L30, 24m, 10d | GD: 4D-Var-AR T119 L42 GE: 4D-Var-AR T119 L30 RD: 3D-Var 27/9/3 km | CRTM | level 1b processed to level 1c with AAPP | Yes (RD) | YES | ||
DWD (Germany) | 20 km L60 (7d) | 7 km L40 2.8 km L50 | 2.8 km, 20m, 21h | GD: 3D-Var RD: nudging | RTTOV 7 | Level 1b | NO | YES (G) | |
Met Office (UK) | 25 km L70 (6d) | 60 km L70, 24m, 3d | 12 km L70 1.5 km L70 | 18 km, 24m, 2d | GD: 4D-Var 60 km L70 RD 12 km: 4D-Var 24 km RD 1.5 km: 3D-Var + Latent Heat nudging 1.5 km | RTTTOV 7 | Level 1b/1c | NO | EARS AP-RARS SA-RARS |
DMI (Denmark) | 0.15 deg (16 km) L40 0.09 deg (10 km) L40 0.03 deg (3.3 km) L40 0.05 deg (5.6 km) L40 | 3D-VAR 0.15 deg L40 3D-Var 0.09 deg L40 3D-Var 0.09 deg L40 Inceremental | RTTOV 8.5 | Level 1c from AAPP | NO | EARS | |||
JMA (Japan) | T959 L60 | T319 L60 (51m) (9d) | 5 km L50 | T319 L60, 11m, 5d | GD: 4D-Var T319 L60 RD: 4D-Var 15 km | RTTOV 9.3 | Level 1c | NO | EARS, AP-RARS, SP-RARS |
Météo France (France) | T798C2.4 L70 (4d) | T538C2.4 L65 (35m) | 2.5 km L60 | 25 km (35m) (4d) | GD: 4D-Var+ensemble T323C1 L70 RD: 3D-Var 2.5 km | RTTOV 9.2 | level 1c | NO | EARS, AP-RARS |
NCEP (USA) | T574 L64 (7.5d) T190 L64 (16d) | T254L42, 84m, 1-8d T190L42, 84m, 8-16d | 12 km L60 4 km L50 | 16 km L35 (21m) (3.6d) | GD: 3D-Var T574 L64 RD: 3D-Var 12 km | CRTM | Level 1b | NO | Yes (G) |
BoM (Australia) | 80 km L50 (10d) | 37.5 km L50 | GD: 4D-Var 100 km L50 RD: 4D-Var 75 km L50 | RTTOV 7 | Level 1d from AAPP Met Office | NO | NO | ||
CPTEC/INPE (Brazil) | 45 km L64 (7d) | 100 km L28, 40m, 15d | 15 km L50 | 20 km, 21m, 10d | GD: 3D-Var 40 km RD: 3D-Var 15 km | CRTM | YES | NO | YES |
OMSZ (Hungary) | 8 km L49 (2.25d) 2.5 km L60 (2d) | 8 km (11m) L49 (2d) | RD: 3D-Var 8 km L49 RD: 3D-Var 2.5 km L60 | RTTOV 8.5 | Level 1c | NO | EARS | ||
SMHI (Sweden) | 10 km L60 | 4D-Var | RTTOV7 | YES | NO | EARS | |||
CNMCA (Italy) | 7 km L40 2.8 km L50 | R: LETKF 10 km | RTTOV9.2 | Level 1c | NO | EARS | |||
KMA (Korea) | 25 km L70, 10.5d | 40 km L70, 10d | 12 km L70 1.5 km L70 | G: 4D-Var 75 km L70 R 12 km: 4D-Var 36 km L70 R 1.5 km: 3D-Var + FGAT | RTTOV7 | Level 1d from AAPP from Met Office | NO | EARS, AP-RARS, SA-RARS | |
NCMRWF (India) | T574L64 | 3DVAR | CRTM | Level 1b | NO | NO |
CODES
G Global
R Regional
GD Global deterministic
RD Regional deterministic
GE Global ensemble
RE Regional ensemble
Tn Truncation wavenumber
Ln Number of levels
(nd) number of days
(nm) number of members
Table 3 Use of ATOVS Radiances in NWP
Centre | NWP Systems using ATOVS | Thinning | Bias Corr. | Other Preproc | AMSU-A | AMSU-B/MHS | HIRS | |||||||||||||||||||||||||||
Satellite | Channels | Satellite | Channels | Satellite | Channels | |||||||||||||||||||||||||||||
N15 | N16 | N17 | N18 | N19 | AQ | MO | land/sea | sea/low topo | sea | N15 | N16 | N17 | N18 | N19 | AQ | MO | land/sea | sea/low topo | sea | N15 | N16 | N17 | N18 | N19 | AQ | MO | land/sea | sea/low topo | sea | |||||
EC (Canada) | GD, RD, GE | GD, RD: 150km GE: 250km(1) | HK off-line dynamic (2) | screen for CLW and precip | X | X | X | X | X | X | 7-14 | 6 | 4-5 | X | X | X | X | 3-4 | 2,5 | |||||||||||||||
ECMWF (Europe) | GD,GE | 125km | variational | X | X | X | X | X | X | 7-14 | 5-6 | X | X | X | 3-5 | X | 12 | 4-7,11, 14,15 | ||||||||||||||||
met.no (Norway) | RD (HIRLAM) | 50 km | HK off-line static | cloud clearing | X | X | X | 4-10 | ||||||||||||||||||||||||||
FNMOC/NRL (USA) | GD, GE | 140km(4) | HK off-line dynamic(5) | screen for high CLW | X | X | X | X | X | X | 6-14 | 4-5 | X | X | X | 4-5 | ||||||||||||||||||
DWD (Germany) | GD | 200km | off-line static(6) | screen for CLW and precip. | X | X | X | X | 5-14 | |||||||||||||||||||||||||
Met Office (UK) | GD,RD, GE | G=125/154km R=80km | off-line | calibration Remap to HIRS | X | X | X | X | 6-14 | 4-5 | X | X | X | X | 3-5 | G | G | 4-7,11 12,15 | ||||||||||||||||
DMI (Denmark) | RD (HIRLAM) | 0.9 deg (100km) | HK off-line static | X | X | X | X | X | 4-10(7) | X | X | X | 3-5 | |||||||||||||||||||||
JMA (Japan) | GD,GE | G: AMSU-A 250km, AMSU-B 180km R: 45 km | off-line for scan bias, remainder with variational | cloud/rain classification | X | X | X | X | X | X | 8-13 | 6-7 | 4-5 | X | X | X | X | 3-5 | ||||||||||||||||
Météo France (France) | GD,RD | 125km; 80 km IASI RD | variational | X | X | X | X | X | X | 5-13(8,9) | 5-13 | 5-13 | X | X | X | 3-5 | 3-5 | 3-5 | X | X | X | 12 | 4-7,11, 14,15 | |||||||||||
NCEP (USA) | GD,RD | G=145km R=120km IR, 60km MW | variational | (10) | X | X | X | X | X | 1-13,15 | X | X | G | X | 1-5 | X | X | 2-15 | ||||||||||||||||
BoM (Australia) | GD,RD | G=154km R=80km | HK off-line(11) | (12) | X | X | X | X | 7-14 | 6 | 4-5 | X | X | 3-5 | X | X | X | X | 4-7,11 12,15 | |||||||||||||||
CPTEC/INPE (Brazil) | GD, RD | 145 km | X | X | X | X | X | X | X | X | X | X | ||||||||||||||||||||||
OMSZ (Hungary) | RD | 80km | variational | X | X | X | 5-12 | X | X | 3-5 | ||||||||||||||||||||||||
SMHI (Sweden) | RD (HIRLAM) | 0.9 deg | HK off-linestatic | X | X | 5-10 | ||||||||||||||||||||||||||||
CNMCA (Italy) | RD | R=120 km | off-line dynamic | Screen for CLW and precip | X | X | X | X | 7-10 | 5-6 | 5-10 | (3) | (3) | (3) | (3) | (3) | (3) | (3) | 3-4 | 5 | ||||||||||||||
KMA (Korea) | GD, RD, GE | G=154 km R=80 km | off-line | (120 | X | X | X | X | X | 7-14 | 4-14 | X | X | X | X | 3-5 | X | X | X | 2-8, 10, 15 | ||||||||||||||
NCMRWF (India) | GD | variational | X | X | X | X | X | 1-10, 12, 13 | X | X | X | X | 1-5 | X | X | X | 2-15 |
NOTE: All centres reject some channels from certain satellites for which the radiance data are known to be bad. These details are not provided here.
CODES
G Global
R Regional
GD Global deterministic
RD Regional deterministic
GE Global ensemble
RE Regional ensemble
Nxx,XX NOAA-XX
MO Metop
AQ Aqua
off-line dynamic bias corrections updated and applied to data before each analysis
variational e.g VarBC, where bias corrections are computed during variational analysis as part of minimization
HK Harris and Kelly type bias correction (thickness predictors)
sea/low topo assimilation over ocean and low topograohy land
CLW cloud liquide water
NOTES
(1) Thinning to be reduced to 150 km (currently being tested)
(2) Bias correction is off-line static for AMSU-A channels 11-14
(3) AMSU-B/MHS assimilation is being tested
(4) Start with taking every 4th point along scan, from every other scan line (skipping pattern to sample every scan point), then further reduce to approximately 140 km.
(5) VarBC code developed, to be tested this spring.
(6) Switch to an online bias correction scheme soon
(7) Channels 4-5 not assimilated over sea ice
(8) AMSU-A ch. 5-6 not assimilated over sea ice
(9) AMSU-A channels in range 5-13 not assimilated: N15 (6,11) N16 (5-10,12) AQ (5-7) MO (7)
(10) Calibration and for RARS data removal of antenna correction (G), Calibration ( R )
(11) Bias correction soon to become offline adaptive
(12) Cailibration and pre-processing via AAPP at Met Office. Cloud detection, land/sea/ice determination in OPS pre-processor as part of Met Office UM suite.
Table 4 Use of ATOVS Retrievals in NWP
Centre | ATOVS retrievals | ||||||||
Used in radiance assimilation e.g. Qc | Assimilated | NWP system | Scheme/Product | Variable(s) | Vertical resolution | Horizontal sampling | Excluded | Bias correction | |
EC (Canada) | NO | NO | |||||||
ECMWF (Europe) | NO | NO | |||||||
met.no (Norway) | NO | NO | |||||||
FNMOC/NRL (USA) | NO | YES | RD (COAMPS, NAVDAS) | NESDIS | Temperature | Full to 10 mb | 1.5 degrees (167 km) | over land | NO |
DWD (Germany) | NO | NO | |||||||
Met Office (UK) | YES (1) (in GD) | NO | |||||||
DMI (Denmark) | NO | NO | |||||||
JMA (Japan) | NO | NO | |||||||
Météo France (France) | NO | NO | |||||||
NCEP (USA) | NO | NO | |||||||
BoM (Australia) | YES (2) | NO | |||||||
CPTEC=INPE (Brazil) | NO | NO | |||||||
OMSZ (Hungary) | NO | NO | |||||||
SMHI (Sweden) | |||||||||
CNMCA (Italy) | NO | NO | RD | EUMETSAT | Temperature | Up to 10 hPa | |||
KMA (Korea) | YES | NO | |||||||
NCMRWF (India) | NO | NO |
NOTES
(1) Run 1D-var prior to assimilation which analyses skin temperature and for advanced IR sounders cloud cover and emissivity which is used in the forward calculations in 4D-var.
(2)
1D-Var
Table 5 Use of AIRS and IASI Radiances in NWP
Note: For some centres, further information on channel numbers and observation errors for AIRS and IASI can be found at the end of this document. Also, please note the definitions of the bands below the table.
Centre | AIRS | IASI | ||||||||||||||
15 microns (1) | Window + O3 (2) | H2O (3) | Short Wave (4) | 15 microns (1) | Window + O3 (2) | H2O (3) | Short Wave (4) | |||||||||
Land | Ocean | Land | Ocean | Land | Ocean | Land | Ocean | Land | Ocean | Land | Ocean | Land | Ocean | Land | Ocean | |
EC (Canada) | 4 | 19 | 0 | 10 | 26 | 33 | 0 | 24 | 14 | 43 | 0 | 19 | 0 | 0 | 0 | 0 |
ECMWF (Europe) | 48 | 81 | 0 | 32 | 0 | 7 | 0 | 19 | 0 | 153 | 0 | 28 | 0 | 9 | 0 | 1 (*) |
met.no (Norway) | ||||||||||||||||
FNMOC/NRL (USA) | 31 | 51 | 13 | 25 | 51 | |||||||||||
DWD (Germany) | ||||||||||||||||
Met Office (UK) | 46 | 75 | 0 | 19 | 0 | 47 | 0 | 0 | 87 | 87 | 14 | 21 | 30 | 30 | 0 | 0 |
DMI (Denmark) | ||||||||||||||||
JMA (Japan) | ||||||||||||||||
Meteo France (France) | 0 | 54 | 0 | 0 | 0 | 0 | 0 | 0 | 50 | 64 | 0 | 4 | 9 | 9 | 0 | 0 |
NCEP (USA) | 75 | 75 | 11 | 11 | 20 | 20 | 14 | 14 | 116 | 116 | 45 | 45 | 4 | 4 | ||
BoM (Australia) | 20 | 12 | 16 | 14 | ||||||||||||
CPTEC/INPE (Brazil) | ||||||||||||||||
SMHI (Sweden) | ||||||||||||||||
CNMCA (Italy) | ||||||||||||||||
KMA (Korea) | 0 | 20 | 0 | 11 | 0 | 16 | 0 | 0 | 95 | 124 | 0 | 29 | 24 | 30 | 0 | 0 |
NCMRWF (India) | 75 | 11 | 20 | 14 | 116 | 45 | 4 |
Notes:
(1) 15 microns: 650 – 770 cm^-1
(2) Window + 03: 770 – 1210 cm^-1
(3) H2O: 1210 – 2000 cm^-1
(4) Short wave: 2000 – 2700 cm^-1
(*) Channel 5480 – this is really a WV channel, but falls into the short-wave band in the above categorisation.
Table 6 Treatment of cloudy sounder radiances in NWP
Centre | HIRS | AIRS | IASI | AMSU-A | AMSU-B/MHS |
---|---|---|---|---|---|
EC (Canada) | n/a | IC | IC | IC | IC |
ECMWF (Europe) | IC & SV | IC & SV | IC & SV | IC | IC |
met.no (Norway) | |||||
FNMOC/NRL (USA) | n/a | IC | IC | IC | IC |
DWD (Germany) | HH | ||||
Met Office (UK) | HH | SV (with dynamic selection of weakly cloud-affected channels) | SV (with dynamic selection of weakly cloud-affected channels) | AC | IC |
DMI (Denmark) | |||||
JMA (Japan) | IC | IC | |||
Meteo France (France) | IC | IC & SV | IC & SV | IC | IC |
NCEP (USA) | IC | IC | IC | IC | IC |
BoM (Australia) | HH | HH | IC | IC | |
CPTEC/INPE (Brazil) | |||||
SMHI (Sweden) | |||||
KMA (Korea) | IC | IC | |||
NCMRWF (India) |
Notes:
HH: Hole hunting: search for FOVS for which all channels are clear, and only assimilate these clear-sky radiances. Cloud-affected radiances are not assimilated.
IC: Insensitive channels: search for channels and spots which are considered clear, and only assimilate these clear radiances; clouds might be present for lower-peaking channels, but only channels insensitive to these clouds are assimilated. Cloud-affected radiances are not assimilated.
CC: Cloud clearing: method to transform potentially cloud-affected observed radiances to clear radiances, and then assimilate these modified radiances.
SV: Sink variable: cloud parameters are retrieved or analysed as a sink variable in 1D/3D/4D-Var, allowing the assimilation of certain cloud-affected radiances. The retrieved or analysed cloud information does not directly affect the model cloud fields.
AC: Assimilate clouds: assimilation of cloud-affected radiances, using information from model clouds and with feedback on model clouds.
Additional information
Some centres have kindly provided further information on channel usage and observation errors, especially for AIRS and IASI, but occasionally also for other instruments. These are provided below:
EC (Canada) as of 1 January 2012
FNMOC/NRL (USA) AMSUA AIRS AQUA and IASI METOPA as of 1 January 2012
Meteo France (France) as of 1 January 2012
NCEP (USA) as of 1 January 2012