Thomas August (EUMETSAT)

1
Status of the operationalIASI L2 version 5
products

• Thomas August (EUMETSAT)
• Tim Hultberg, Dorothée Coppens, Arlindo Arriaga,
  Lars Fiedler,
• Xavier Calbet, Dieter Klaes, Peter Schlüssel,
  Rose Munro, Marc Crapeau

2
Outline

1. Overview of the IASI mission
2. The IASI L2 processor (PPF) version 5
3. Discussion on T, q validation
4. Summary

3
I. The IASI mission The instrument
Infrared Fourier transform interferometer Spectral
range 645 to 2760 cm-1 (15.5-3.62 µm) Spectral
sampling 0.25 cm-1 Footprint 12 km
(Nadir) Swath 1000km ( 48.3)
4
I. The IASI mission Metop and the EPS program

• European contribution to The Initial Joint
  Polar-orbiting System (IJPS)
• Satellites in complementary orbits designed to
  ensure global data coverage at intervals of no
  more than 6 hours.

• Prime objective provide continuous, long-term
  data sets in support of operational
  meteorological and environmental forecasting and
  global climate monitoring.
• Lifetime 3 satellites, more than 14 years
  operation from October 2006.
• Imager (AVHRR) and MW (AMSU,MHS) companions

5
I. The IASI mission The operational L2 products
Disseminated in NRT (sensing 3h)
Vertical profiles Temperature Vertical profiles
Humidity Land Sea Surface Temperature
TWT
EMS
Surface emissivity
Cloud detection and characterisation
CLD
OZO
Ozone total partial (0-6, 0-12, 0-16 km)
columns
TRG
Slide 5
6
II. IASI L2 PPF version 5 Why upgrade ?
Processing chain version 5 operational since 14
September 2010 !
7
II. IASI L2 PPF version 5 The main steps

1. Input data pre-processing
2. Cloud detection and characterisation
3. Statistical retrievals T, q, Ts, e, trace gases
4. Optimal Estimation Method (OEM) T, q, Ts, O3

8
II. IASI L2 PPF version 5 The static databases
Data type Origin Resolution
Land/Sea mask AAPP quadtree
Surface elevation GTOPO 30 (US Geological Survey) 30 x 30 (1 km)
Emissivity Atlas Global land surface IR emissivity database UW-Madison (Borbas et al., 2007) 0.05 x 0.05 lat/lon 10 wavelengths Monthly means
Emissivity Atlas Global land surface climatology from IASI measurements (Zhou et al., submitted to IEEE TGRS ) 0.5x0.5 Full spectrum (via PCs) Monthly means
9
II. IASI L2 PPF version 5 Measurements aux.
products

• IASI L1C radiances
• Noise-filtering via PC reconstruction
• ECMWF 6h-forecasts 00, 06, 12, 18 UTC
• Up to 3h difference with sensing-time
• Time-interpolation
• AVHRR
• L1B used if IASI L1C format lt 11.0
• Cloud-fraction embedded in IASI 1C used for cloud
  screening
• AMSU MHS 1B
• Not used yet.

10
II. IASI L2 PPF version 5 Cloud detection
NWP test
AVHRR test
Opt. depth test

• NWP cloud test
• OBS CALC(NWP, ems, ) _at_ chans 751 1023
• Cloudy if OBS-CALC gt 1K
• Relies on accurate forecasts and surface ems
• AVHRR integrated cloud-fraction
• IASI PSF weighted count of AVHRR cloudy pixels
• Cloudy if CFR gt 2
• Cloud optical depth test
• EOF retrieval, D. Zhou et al., GRL 2005
• Triggers final acceptance of cloudy retrievals

18/11/2009 White clear
11
II. IASI L2 PPF version 5 Statistical retrievals
T, q profiles

• EOF retrieval (D. Zhou et al., 2005, GRL)
• Regression with different coefficients for
  different viewing angles
• Full profiles available under clear and partly
  cloudy conditions
• Clear sky first-guess in the final iterative
  retrieval
• Clear sky q profiles enters final L2 products
• Cloudy IFOVs after users request, the cloudy T,q
  profiles retrievals were added to the final IASI
  L2 products. Distributed since 02/12/2010.

12
II. IASI L2 PPF version 5 Statistical retrievals
SST
Impact of the cloud aerosol detection on the
product quality
Global figures Cold bias 0.4 K s 0.4 K
Outside aerosol areas Cold bias 0.25 K s lt 0.3
K IASI L2P GHRSST
(AATSR IASI) SST
13
II. IASI L2 PPF version 5 Statistical retrievals
LST

• 19-24 March 2010 Night
• LANDSAF_MSG IASI

Global figures (excluding deserts) ?
0.99 Warm bias 1 K s lt 2 K
Slide 13
14
II. IASI L2 PPF version 5 The iterative
retrieval OEM

• Standard Optimal Estimation Method (OEM) after
  Rodgers
• J (x - xa)T . B-1 . (x - xa) (F(x) y)T .
  R-1 . (F(x) y)
• Marquardt-Levenberg minimisation (5 iterations
  max.)
• Active parameters Tskin and T, q O3 profiles
• Background term computed in the EOF space of the
  atmospheric parameters T -gt 28 PCs, q -gt 18 PCs,
  O3 -gt 9 PCs
• One unique global a priori and covariance matrix
• 316 channels after Collard (Collard et al., QJRMS
  2007)
• global radiance tuning and measurement error
  covariance matrix prior clear sky OBS-CALC
• Forward model RTIASI-4 optimised for faster
  computations
• First guess statistical retrievals

15
II. IASI L2 PPF version 5 T, q profiles
non-regression check

• Non-regression check and validation result from
  comparisons against ECMWF analyses.
• 3 datasets
• One Year orbits centred on analyses synoptic
  times
• Focus days (8 Sep 2007, 5 March and 9 May 2008)
• 19-24 March 2010, with latest L1c products
• Detailed results presented in EUM/MET/TEN/09/0448

16
II. IASI L2 PPF version 5 T, q profiles
non-regression check
17
II. IASI L2 PPF version 5 Temp. profiles
non-regression check

• 19-24 March 2010
• IASI - ECMWF
• Yield increased x1.7
• rms 1K 200-800hPa

rms 0.7 K !!!
rms 0.7 K !!!
- v5 - v4
18
II. IASI L2 PPF version 5 WV profiles
non-regression check

• 19-24 March 2010
• (IASI - ECMWF)
• Bias reduced in BDL

EOF FG
OEM
- v5 - v4
19
II. IASI L2 PPF version 5 OEM Future
developements

• Include covariance Tskin T WV and possibly O3
  in BG
• Synergistic use of MW measurements AMSU, MHS
• Add physical constraints (super-saturation and
  adabiatic lapse rate) to restrict solution space
• More channels needed for humidity retrievals
  external study for MTG-IRS by C. Serio (Univ.
  Basilicata), paper by Rodgers in Adv. Space Res.
  1998
• Migration to RTTOV-10
• Upgraded spectroscopy
• MW instruments gt residual cloud contamination
  effect
• PCRTM gt use more channels in retrievals
• Simulation of clouds aerosols

20
III. Discussion on T,q validation T _at_ NOAA/NPROVS
T _at_ 500 hPa
T _at_ 700 hPa
T _at_ 850 hPa
PPFv4
PPFv5
Credits NOAA / NESDIS Center for Satellite
Applications and Research.
21
III. Discussion on T,q validation WV _at_
NOAA/NPROVS
WV _at_ 500 hPa
WV _at_ 700 hPa
WV _at_ 850 hPa
PPFv5
Credits NOAA / NESDIS Center for Satellite
Applications and Research.
22
III. Discussion on T,q validation Sondes _at_
Sodankylä

• Humidity profiles High variability, model vs
  sondes differences
• Credits X. Calbet (EUM) and R. Kivi (FMI)

13/07/2007
3s noise
Calbet et al., AMTD 2010
23
III. Discussion on T,q validation ConcordIasi

• Co-operation with F. Rabier and L. Lavanant
  (Meteo-France)

• 3 main field experiments during Austral springs
  2008, 2009 and 2010
• In 2010, 19 super-pressure balloons (CNES)
  launched from McMurdo, 6 with a scientific
  payload sounding the stratosphere, and 13 of the
  driftsonde type developed at NCAR.
• From the 13 driftsondes flying at around 20km
  altitude, 640 dropsondes were dropped over
  Antarctica and the surrounding seas, coincident
  with Metop overpasses.
• Additional measurements performed at other
  stations in Antarctica Dumont dUrville,
  Concordia and Rothera.
• Drop-ondes coincident with Metop and A-Train
  overpasses

24
Summary

• PPFv5 operational since 14/09/2010 !
• Significant improvements T, LST, clouds (O3, CO)
• Products further characterised SST, LST
• Some cloudy retrievals
• Validation documentation available on-line
  www.eumetsat.int
• Improve/characterise further T and q in bottom
  layers
• Reduce use of NWP forecasts
• Seasonal variations in long-term monitoring
  origins ? How to account for it in CDRs ?
• clouds/aerosol detection/simulation impact on WV
  profiles quality ?
• What WV products references for the validation of
  satellite retrieved profiles with ground
  resolution of 12-30 km ?

25

• Thank you for your attention !
• thomas.august_at_eumetsat.int