Assimilation of AIRS/IASI data at ECMWF PeterBauer - PowerPoint PPT Presentation

Assimilation of AIRS/IASI data at ECMWF Peter
Bauer Peter
Bauer,
European
Centre
for
Medium-Range
Weather
Forecasts Tony
McNally,
Andrew
Collard,
Marco
Matricardi,
Wei
Han,
Carla
Cardinali,
Niels
Bormann •
Initial
performance
/
impact
assessment •
Upgrades:
Addition
of
water
vapour
channels,
cloud-affected
radiances,
ozone •
Comprehensive
observing
system
experiments Slide 1 •
Future
upgrades •
Summary Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

Data assimilation system (4D-Var)  The observations are used to correct errors in the short forecast from the previous analysis time.  Every 12 hours we assimilate 4 – 8,000,000 observations to ~3,000,000 from AIRS& IASI! correct the 100,000,000 variables that define the model’s virtual atmosphere.  This is done by a careful 4-dimensional interpolation in space and time of the available observations; this operation Slide 2 takes as much computer power as the 10-day forecast. Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
ECMWF Ⓒ


Data sources: Satellites Radiances ( → brightness temperature = level 1): • AMSU-A on NOAA-15/18/19, AQUA, Metop • AMSU-B/MHS on NOAA-17/18/19, Metop • SSM/I on F-15, AMSR-E on Aqua • HIRS on NOAA-17/19, Metop • AIRS on AQUA, IASI on Metop • MVIRI on Meteosat-7, SEVIRI on Meteosat-9, GOES-11/12, MTSAT-1R imagers Bending angles ( → bending angle = level 1): • COSMIC (6 satellites), GRAS on Metop Ozone ( → total column ozone = level 2): • Total column ozone from SBUV on NOAA-17/18, OMI on Aura Atmospheric Motion Vectors ( → wind speed = level 2): • Meteosat-7/9, GOES-11/12, MTSAT-1R, MODIS on Terra/Aqua Sea surface parameters ( → wind speed and wave height = level 2): • Near-surface wind speed from Seawinds on QuikSCAT, ERS-2 scatterometer, Slide 3 ASCAT on Metop • Significant wave height from RA-2/ASAR on Envisat, Jason altimeter Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

Initial performance assessment Upgrades: Addition of water vapour channels, cloud-affected radiances Comprehensive observing system experiments Future upgrades Summary Slide 4 Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

Current use of AIRS/IASI data • AIRS
CO 2 
and
H 2 O
channels
assimilated
since
October
2003
(324
channels,
1/9
FOV). • IASI
CO 2 /H 2 O
channels
assimilated
since
June
2007/March
2009
(8461
channels,
1/4
FOV). Slide 5 • Assimilated
in
clear-sky
areas
and
above
clouds;
since
September
2009
in
fully
overcast situations,
AIRS
(not
IASI)
over
land
surfaces/sea-ice. • Continuous
revision
of
channel
usage,
quality
control:
Ozone
channels,
PC
RT. Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

Noise: AIRS vs. IASI data IASI AIRS (after
April
2007
calibration
change) FG-departure standard
deviation Mean
FG-departure Δ TB
[K] after
bias
correction Mean
FG-departure before
bias
correction Slide 6 λ
[μm] λ
[μm] (A.
Colla llard) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
ECMWF Ⓒ


IASI: Model minus observations First-guess
departure
standard
deviations in
15
μm
CO 2 -band Calculated Observed Slide 7 (A.
Colla llard) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

IASI: Model minus observations First-guess
departure
standard
deviations in
H 2 O-band Calculated Observed Slide 8 (A.
Colla llard) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

Initial performance assessment Upgrades: Addition of water vapour channels, cloud-affected radiances Comprehensive observing system experiments Future upgrades Summary Slide 9 Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

IASI H 2 O channel impact 10
IASI
water
vapour
channels Grey
channels
are
the
120
H 2 O
channels
 distributed
via
the
GTS Slide 10 (A.
Colla llard) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

IASI H 2 O channel impact 10
IASI
water
vapour
channels:
Fit
to
other
moisture
sounder
radiances Normalised Best value at ~1.5K to unity here Slide 11 (A.
Colla llard) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

IASI/AIRS cloud detection AIRS
channel
226
at
13.5micron A
non-linear
pattern
recognition
algorithm
is
applied
to
departures (peak
about
600hPa) of
the
observed
radiance
spectra
from
a
computed
clear-sky background
spectra. obs-calc
(K) obs Vertically
ranked
channel
index This
identifies
the
characteristic
signal
of
cloud
in
the
data
and allows
contaminated
channels
to
be
rejected. AIRS
channel
787
at
11.0
micron unaffected
 unaffected
 (surface
sensing
window
channel) channels
 channels
 assimilated assimilated hPa) Pressure
(hPa CLOUD CLOUD contaminated
 contaminated
 channels
 channels
 rejected rejected Slide 12 Temperature
Jacobian
(K) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
ECMWF Ⓒ


Assimilation of cloud-affected channels • by
adding
cloud
top
pressure
and
effective
cloud
fraction
to
control
vector (via
sink
variable),
for
retrieved
effective
cloud
cover
=1; • no
cloudy
RT
calculations
required,
conservative
linearization
point. Single
cycle
HIRS,
AIRS,
IASI
overcast
/
clear Slide 13 (T.
McNally lly) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

Assimilation of cloud-affected channels Temperature
forecast
error
RMSE
difference (EXP-CTRL,
77
cases,
own
analyses) 200
hPa 200
 Positive: deterioration Negative: improvement 0.2+ K shading 500
hPa 500
 700
hPa Slide 14 (T.
McNally lly) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
ECMWF Ⓒ


Initial performance assessment Upgrades: Addition of water vapour channels, cloud-affected radiances Comprehensive observing system experiments Future upgrades Summary Slide 15 Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

AIRS/IASI impact CTRL plus AIRS EU NH US SH Slide 16 (T.
McNally lly) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

AIRS/IASI impact CTRL plus IASI NH EU SH US Slide 17 (T.
McNally lly) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

AIRS/IASI impact CTRL plus both NH EU SH US Slide 18 (T.
McNally lly) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF

Advanced diagnostics GOES-Rad MTSAT-Rad MET 9-Rad MET 7-Rad Relative FC error reduction per system AMSU-B MHS AMSR-E SSMI GPS-RO IASI AIRS AMSU-A HIRS TEMP-mass DRIBU-mass AIREP-mass The forecast sensitivity SYNOP-mass SCAT-wind MODIS-AMV (Cardinali, 2009, QJRMS, MET-AMV MTSAT-AMV GOES-AMV PILOT-wind 135, 239-250) denotes the TEMP-wind DRIBU-wind AIREP-wind sensitivity of a forecast error SYNOP-wind 0 2 4 6 8 10 12 14 16 18 20 metric (dry energy norm at 24 FEC % or 48-hour range) to the observations. The forecast GOES-Rad MTSAT-Rad MET 9-Rad sensitivity is determined by MET 7-Rad AMSU-B MHS AMSR-E the sensitivity of the forecast SSMI GPS-RO IASI error to the initial state, the AIRS AMSU-A HIRS TEMP-mass innovation vector, and the DRIBU-mass AIREP-mass SYNOP-mass Kalman gain. SCAT-wind MODIS-AMV MET-AMV MTSAT-AMV GOES-AMV PILOT-wind TEMP-wind DRIBU-wind AIREP-wind Slide 19 SYNOP-wind Relative FC error reduction per observation 0 5 10 15 20 25 30 FEC per OBS % (C.
Cardin inali li) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
ECMWF Ⓒ


Advanced diagnostics – MW sounder denial black cntrl 3 AMSU-A, 2 MHS vs 1 AMSU-A, 0 MHS O3 GOES-Rad MTSAT- MERIS Met-Rad Met-Rad AMSU-B MHS SSMI GPS-RO IASI AIRS AMSU-A HIRS SCAT Met-AMV GOES- PILOT TEMP DRIBU AIREP SYNOP 0 1 2 3 4 5 6 7 8 9 FEC % Slide 20 Forecast error reduction [%] (C.
Cardin inali) li) Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
ECMWF Ⓒ


Initial performance assessment Upgrades: Addition of water vapour channels, cloud-affected radiances Comprehensive observing system experiments Future upgrades Summary Slide 21 Assimilation
of
AIRS/IASI
data
at
ECMWF 
 


P.
Bauer Ⓒ
 Ⓒ
ECMWF