GRAS SAF CDOP PT7 Meeting

1
ROM SAF and Radio Occultation Products
Kent B. Lauritsen ROM SAF team Danish
Meteorological Institute, Copenhagen,
Denmark Contents - Short introduction to
ROM SAF and Radio Occultations (RO) - Status of
NRT operations - Offline processing and offline
and climate data products - Overview of some
development and climate related activities
EUM 2012, 3-7 September 2012, Sopot, Poland
2
ROM Satellite Application Facility Consortium

• Leading Entity Danish Meteorological Institute
  (Copenhagen, Denmark)
• Kent B. Lauritsen, Hans Gleisner, Johannes K.
  Nielsen, Frans Rubek, Helge Jønch- Sørensen,
  Stig Syndergaard, Hallgeir Wilhelmsen, Kristian
  Rune Larsen
• Partners ECMWF (Reading, UK)
• Sean Healy
• Institute dEstudis Espacials de
  Catalunya (IEEC, Barcelona, Spain)
• Estel Cardellach, Santi Oliveras
• Met Office (Exeter, UK)
• Dave Offiler, Chris Burrows, Ian
  Culverwell

3
Radio Occultation Meteorology (ROM) SAF objectives

• One of EUMETSATs 8 SAFs
• Operational processing and archiving center for
  radio occultation (RO) data from Metop (GRAS) and
  other RO missions
• Data products and software deliverables
• Near-real time (NRT) radio occultation products
• - operational products in NRT (refractivity,
  temperature, pressure, humidity)
• Offline radio occultation products
• - offline profiles (bending angle, refractivity,
  temperature, pressure, humidity)
• - climate products gridded data of bending
  angle, refractivity, temperature, humidity,
  geopotential height
• - reprocessed RO data sets
• Radio Occultation Processing Package (ROPP)
• - routines for assimilation and processing of RO
  data

4
Atmosphere profiling by radio occultation
t2
Bending angle (BA) vertical profiles of - N
(refractivity) - T (temperature) - P
(pressure) - q (humidity)
time t1
Metop/GRAS
(setting occ)
t3
GPS
2 GPS signals L1 19 cm L2 24 cm
(rising occ)

• While a GPS satellite sets or rises behind
  the horizon
• Additional bending of the GPS signals ray path
  due to refraction in the atmosphere
• The GPS receiver measures the excess Doppler
  shift

5
General principle of RO data processing
Bending angle obtained from the measured phases
and amplitudes and the positions and velocities
of the two satellites using Doppler shift and
geometric optics or wave optics inversion
(canonical transform) Ionosphere corrected
bending angle obtained by linear combination of
the bending angles corresponding to the two GPS
frequencies L1 and L2 Refractivity obtained
from the bending angle as a function of height
using the Abel Transform inversion (assuming
spherical symmetry and statistical
optimization) Dry temperature and pressure
obtained by using the ideal gas law and the
assumption of hydrostatic equilibrium (set
humidity ? 0) Pressure, temperature and
specific humidity (water vapor) obtained using
an ancillary temperature, humidity and, e.g., the
1D-Var algorithm
6
Atmospheric sounding with RO
L 300 km Z 0.1 1.5 km


7
Global distribution of RO profiles
1 day (Metop-A) 1 month (Metop-A)
8
ROM SAF NRT refractivity product (GRM-01)

• Produced from geometric optics (GO) level 1b
  operational NRT bending angles from GRAS/Metop-A
  data from EUMETSAT CAF
• Used by NWP centers worldwide for assimilation or
  as QC when assimilating EUMETSAT CAF level 1b
  bending angles
• Limited information contents at low altitudes
  (due to GO and so far not using raw sampling
  data)
• Future plan
• Production of improved product based on wave
  optics processing will increase the information
  content in the lower troposphere to be done
  after EUMETSAT CAF NRT upgrade end of 2012
• Improve statistical optimization (SO) by using an
  enhanced background climatology

9
Monthly refractivity (GRM-01) statistics
2011-2012
Feb 2011
Mar 2011
Apr 2011
Jun 2011
May 2011
Feb 2012
Mar 2012
Apr 2012
May 2012
Jun 2012
- Very similar bias above 30 km month for month
between 2011 2012 statistics - Seasonal
variation Due to ECMWF model, measurements/stat.
opt., or both?
10
BAROCLIM Bending angle RO climatology
ROM/GRAS SAF VS study with Ulrich Foelsche and
Barbara Scherllin-Pirscher (U. Graz) Purpose to
derive a bending angle (BA) climatology from
averaged RO data and being able to use the
climatology in the statistical optimization
initialization process
BAROCLIM characteristics COSMIC data from
08/2006 to 07/2011 Careful outlier rejection
Calculate monthly mean profiles for 10 zonal
bands Mean profiles are still noisy at high
(impact) altitudes (gt60 km) Statistical
optimization between 60 km and 80 km no MSIS lt60
km, no RO gt80 km
11
BAROCLIM compared to ECMWF
Systematic differences are very small below 40
km (ECMWF assimilates RO bending angles)
Systematic positive difference (gt0.5 ) at 40 km
to 45 km at all latitudes Positive difference
(larger than 2) above 50 km Negative
systematic difference (larger than 2 ) above 50
km at high southern latitudes Differences are
mainly attributable to ECMWF
12
Average-profile vs. single-profile processing
Statistical optimization based on averaging of BA
profiles directly. Relative difference between
observations (average-profile and single-profile)
and ECMWF.
13
Offline GRAS processing BA Refr. statistics
Processing of April 2012 GRAS data from the
EUMETSAT CAF offline prototype (ftp)
BA
Refr.
14
Refractivity differences relative to
ECMWFCOSMIC RO data for 2007, 2008, 2009
Change in ECMWF cycle 32r3 in 2008 i) COSMIC
RO assimilated to surface ii) updated
convection and entrainment physics
Ref M. E. Gorbunov et al, J. Atm. Ocean. Tech.
(2011)
15
ROM SAF offline gridded products climate data
ROM SAF climate data products are built upon the
existing offline products. The climate products
extend the range of such products offered to
users.
Climate data product 2D zonal grid 1 climate errors Time resolution Spatial 2 resolution Formats,graphical Formats,numerical
GRM-17 Bending angle yes Monthly 5 deg latitude PNG, JPG ASCII, netCDF
GRM-18 Refractivity yes Monthly 5 deg latitude PNG, JPG ASCII, netCDF
GRM-19 Temperature yes Monthly 5 deg latitude PNG, JPG ASCII, netCDF
GRM-20 Spec. humidity yes Monthly 5 deg latitude PNG, JPG ASCII, netCDF
GRM-21 Geopot. height yes Monthly 5 deg latitude PNG, JPG ASCII, netCDF
1 A latitude-height grid where the height can be
expressed in MSL altitude, geopotential height,
or in terms of pressure. 2 The height resolution
of the grid is determined by the height
resolution of the profiles 200 m
16
RO climate data intercomparison ROtrends
Fractional anomalies in the 12-20 km layer w.r.t.
annual cycle CHAMP RO data from 2001-2008
The slopes of trend lines agree between 6 RO
processing centers
17
RO data in climate science

• MSU/AMSU vs. RO temperatures Steiner et al.
  (2007, 2009), Ho et al. (2008), Ladstädter et al.
  (2010) ROM/GRAS VS study

MSU data show a stronger cooling trend than RO
(particularly in the tropics). Possible
explanation strong tropical warming in the upper
troposphere better resolved in RO (due to the
high vertical resolution of RO data).
18
RO data and climate change detection

• Feasibility of using RO data (bending angles) for
  trend detectionRinger and Healy (2008)

Analysis of zonal mean bending angles in
a transient climate experiment trends may
be distinguished from natural variability after
10 to 16 years.
19
Concluding remarks

• Selected activities in ROM SAF CDOP-2 phase
• Metop-B production of operational NRT and
  offline products and climate data
• Reprocessing of Metop, COSMIC, CHAMP and other RO
  data with consistent algorithms and validation
  and production of offline climate data records
• Continue to enhance the ROPP package with new
  routines for e.g. tropopause height and planetary
  boundary layer height calculations
• For more information
• - Poster 56 by Johannes K. Nielsen on 1D-Var
  retrieval of temperature and humidity
• - Poster 37 by Hans Gleisner on level 3 offline
  climate data
• - ROM SAF website http//www.romsaf.org