An update of AMSR-E total precipitable water retrieval algorithm and the application for JMA NWP

1
An update of AMSR-E total precipitable water
retrieval algorithmand the application for JMA
NWP

• Masahiro Kazumori
• Numerical Prediction Division
• Japan Meteorological Agency

2
Contents

• Utilization of AMSR-E data in JMA NWP
• Update of TPW algorithm for AMSR-E
• TPW retrieval from SSMIS brightness temperature
• Assimilation experiment of SSMIS TPW in JMA NWP
• Summary and plan

3
Operational NWP Models in JMA

• Meso-scale NWP model
• Disaster prevention information
• 8 times/day implementation
• 5km horizontal resolution
• 4D-Var Data AssimilationSystem

• Global NWP model
• One-week and short range forecast
• 4 times/day implementation
• 20km horizontal resolution
• 4D-Var DataAssimilationsystem

GSM
MSM
4
Utilization of AMSR-E data in MSM
Retrieval assimilation (Total Precipitable Water
and Rain Rate from AMSR-E, TMI and SSMI) Impacts
Better rainfall forecasts
A case study Fukui Heavy Rain in
2004 Assimilation of the Aqua/AMSR-E data to
Numerical Weather Predictions, Tauchi et al.,
IGARSS04 Poster
5
Recent updates of MSM DA system
Improvements of rainfall, wind and temperature
forecasts

• Introduction of a new 4D-Var data assimilation
  system based on Non hydrostatic model, JNoVA, on
  April 7, 2009

Comparison of 3-hr rainfall distribution for
Typhoon case in 16 Aug. 2006.
RA Obs.
New (24-hr fcst.)
Old (24-hr fcst.)
??????????????(???)?297?

• Changes
• Horizontal resolution from 10km to 5km
• Number of vertical model layer from 40 to 50.
  (Model Top 22km)
• Assimilation Time window from 6-hour to 3-hour

6
Impacts of MW data (TPW and Rain rate)on
moisture analysis in MSM
AMSR-E data Coverage(TPW O-B)
Analysis Increment of TPW
With MW data
WO MW data
Much increment from the assimilation of MW
data. Satellite precipitation and moisture
measurements are important data source in data
sparse area (over ocean).
18UTC April 19, 2009
7
TPW retrieval algorithm for AMSR-E
TPW data are retrieved in the pre-process of the
DA system
Input AMSR-E 19,23,37V and H pol. Brightness
Temperature JAXA L1B Ancillary data AMSR-E L2
SST, SSW, T850 (JMA GANAL)

• Original code was developed by Mr. Takeuchi at
  JMA (Takeuchi 2002).
• TPW algorithm was adopted as standard algorithm
  for AMSR-E in JAXA
• and also have been used in NWP at JMA

Takeuchi, Y., Algorithm theoretical basis
document of the algorithm to derive total water
vapor content from ADEOS-II/AMSR, Special Issue
on AMSR Retrieval Algorithms, EORC
Bulletin/Technical Report, JAXA, 2002.
8
Update of TPW algorithm for AMSR-E

• LUT in the algorithm was updated by using 3-yr
  RAOB and AMSR-E collocated dataset (2006-2008).
• Updated LUTs
• T850 and Mean atmospheric temperature table
• Wind speed correction table and extension to
  strong wind condition gt20m/s
• Conversion table PWI (Precipitalbe water index)
  to PWA (Precipitable water amount)
• Correction coefficients on SST dependency
• No use of internal BT conversion from ver.2 to
  ver.1

TPW Verification against RAOB (2009.1-5)
Collocation criteria Within 60min. 150km
mm
mm
NEW Num 1349 Min -18.836 Max 19.008
Ave -0.135 Std 3.355 Current Num 1344
Min -18.532 Max 15.366 Ave 0.817 Std 4.071
AMSR-E TPW
AMSR-E TPW
RAOB TPW
RAOB TPW
mm
mm
9
TPW Comparison in Tropics
JAXA AMSR-E L2 TPW
New AMSR-E TPW
2008.9.15
RSS AMSR-E GRID DATA
mm
Excessive TPW amount in deep convective area
(Tropics)
Similar to RSS TPW product.
10
TPW Comparison in High latitudes
JAXA AMSR-E L2 TPW
New AMSR-E TPW
2008.9.15
RSS AMSR-E GRID DATA
mm
Global Analysis Psea Surface wind speed
2008.9.15 00UTC
m/s
11
TPW retrieval from SSMIS radiance

• JMA obtains DMSP F-16 and 17 SSMIS data (SDR)
  from NOAA/NESDIS.
• Issues on SSMIS data
• Data quality of lower air sounding channels (LAS)
  were not good. (i.e. Solar contamination to warm
  calibration target and antenna emissivity).
• However, Microwave Imaging channels (ENV) can be
  used for NWP purpose and atmospheric moisture
  monitoring.
• SSMIS frequencies
• 19.35 (V,H), 22.235 (V), 37 (V,H), 91.655 (V,H)
  GHz, zenith angle 53.1 deg.
• AMSR-E frequencies
• 18.7 (V,H), 23.8 (V,H), 36.5(V,H), 89 (V,H) GHz,
  zenith angle 55.0 deg.
• AMSR-E TPW algorithm can be applicable for SSMIS
  TPW retrieval.
• LUTs for SSMIS were made by using collocated data
  SSMIS and RAOB.

12
TPW comparison with SSMIS and SSMI
F-16 SSMIS
F-17 SSMIS
2008.9.15
F-13 SSMI (from RSS)
mm
Retrieved TPW from F16 and F17 SSMIS seems to be
consistent with F-13 SSMI TPW from RSS.
13
Assimilation experiment of SSMIS TPW in MSM
MW-PW, MW-RR, RA
1st Guess TPW
SSMIS TPW O-B
Analysis time
Assimilation time window
12
10
11
09
mm
K
12UTC June 9, 2009
TPW increment WO SSMIS
TPW increment W SSMIS
Increase of MW data coverage bring much moisture
information in oceanic region
mm
mm
14
Assimilation experiment of SSMIS TPW in MSM
Operational regional NWP model require much
satellite data because rapid updated forecast is
essential to provide timely information for
disaster prevention
Current data coverage (MW-PW,MW-RRRA) in MSM DA
system
12UTC Jun. 09, 2009
00
03
06
09
12
15
18
21
15
Assimilation experiment of SSMIS TPW in MSM
Analysis increment of total column water vapor
are limited in the satellite data coverage
Analysis increment of total column water vapor
12UTC Jun. 09, 2009
16
Assimilation experiment of SSMIS TPW in MSM
Addition of SSMIS TPW can much increase the
observation information in MSM DA system
DMSP F16F17 SSMIS TPW data coverage in MSM DA
system
12UTC Jun. 09, 2009
17
Summary and plan

• Total Precipitable Water derived from Microwave
  Imager is crucial data source over ocean in JMA
  NWP system.
• Improvement of MW TPW retrieval algorithm
• Update of LUT by using 3-yr RAOB collocated
  dataset.
• Reduced bias in comparison with RAOB, and
  comparison with RSS TPW product
• Available under strong wind condition (over
  20m/s)
• No use of internal BT conversion from ver.2 to
  ver.1
• Bias -0.14 mm and RMS 3.36 mm. Meet
  GCOM-W/AMSR2 RA requirements
• Applicable for SSMIS and other MW Imager data.
• Assimilation of SSMIS (F16F17) TPW retrievals
  in JMA MSM DA system
• Increase the data coverage and fill the data in
  empty analysis time window
• Provided atmospheric water vapor information
  over the ocean ? Hope to see better rainfall
  forecast
• Retrieval of SSMIS TPW is ready. Assimilation
  experiments are planned to confirm the impacts
  for the weather forecast.