ECMWF et’Training Course- Data Assimilation and use of satellite data (3 May 2005) The Global Observing System Overview of data sources Data coverage Data used Data monitoring Use for model verification François Lalaurette and Jean-Noël Thépaut
Mar 27, 2015
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
The Global Observing System
Overview of data sources
Data coverage
Data used
Data monitoring
Use for model verification
François Lalaurette and Jean-Noël Thépaut
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Overview of data sources
SYNOP / SHIP/ METAR
Meteorological/ Aeronautical weather stations (2m, except wind: 10m)
Ships (variable height, default=25m)
Some moored buoys (5m)
BUOYS
Moored (TAO, PIRATA)
Drifters
used parameters: wind, pressure, temperature
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Moored TAO buoy
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Overview of data sources (cont’d)
TEMPSHIP / DROPSONDES
ASAPs (commercial lines) in replacement of weather ships (stationary)
Dropsondes from scientific aircrafts (NOAA, UKMO, DLR); used for FASTEX, NORPEX (winter adaptative observing network experiments), NA-TREC and Tropical Cyclones;
parameters: Temperature, Wind, Pressure, Humidity
PROFILERS
UHF/VHF Doppler "clear air" radars (US and Europe);
parameter: wind
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data reception (radiosondes)
J1994
FMAMJJASONDJ1995
FMAMJJASONDJ1996
FMAMJJASONDJ1997
FMAMJJASONDJ1998
FMAMJJASONDJ1999
FMAMJJASONDJ2000
FMAMJJASONDJ2001
FMAMJJASONDJ2002
FMAMJJASONDJ2003
FMAMJJASONDJ2004
FMAMJJASONDJ2005
FM0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Fre
qu
en
cy
*10
00
Temperature 500 hPa - GLOBALMonthly counts of Radiosondes received at ECMWF
00 UTC 12 UTC 06 UTC 18 UTC
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Profilers
Profiler site near Haskell, OK (http://www-dd.fsl.noaa.gov/)
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data Coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Overview of data sources (cont’d)Aircraft
AIREPS ("manual" reports from pilots)
AMDARs, ACARs, ASDARs: automated (high quality)
parameters: wind, pressure, temperature (NO humidity)
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Overview of data sources: Satellite data
Two different types of space agencies
• Research Agencies
• Operational Agencies
Two ways of looking at the earth/atmosphere
• GEO (GEOstationary satellites)
• LEO (Low Earth Observing satellites)
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
GEOSTATIONARY OBSERVING SYSTEMS(36 000 km from the earth)
Advantages:
Wide space coverage (whole disk)
Very high temporal coverage ( a few minutes)
• Particularly suitable for short-range NWP and Now-casting applications
• Suitable also for meteorological feature tracking– (Atmospheric Motion winds)
• Suitable for applications in which the diurnal cycle representation is crucialDrawbacks:
Spatial coverage limited to the disk (need for constellation)
Unsuitable to observe the polar regions
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Low Earth Orbiting OBSERVING SYSTEMS(400 to 800 km from the Earth)
Advantages:
Cover the whole earth after several cycles (polar orbiting satellites)
More suitable to sound the atmosphere in the microwave spectrum.
Drawbacks:
Moderate temporal sampling (several hours to go back to the same point)
Requires constellation to ensure a reasonable temporal sampling
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
NOAA-15 NOAA-16 NOAA-17
Goes-W Goes-E Met-7 Met-5 GMS(Goes-9)
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Current Space based Observing System
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Overview of data sources (cont’d)AMV - Atmospheric Motion Vectors (formerly SATOB)
geostationary satellites (GOES 9/10/12; METEOSAT 5/7)
Polar orbiting (MODIS on Terra)
Availability on a very rapid increase (higher space and time resolution, new platforms, quality indexes)
Unknown parameter: height!
Raw radiances
HIRS, AMSU (NOAA 15/16/17/Aqua), AIRS (Aqua), METEOSAT/GOES
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Another type of inversion: Polar WV winds from MODIS
Source: P. Menzel, 2003
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Overview of data sources (cont’d)
Scatterometer (Microwave, active)
2 platforms (ERS2 and Quickscat)
parameter: sea wind (+wave heights from ERS altimeter)
DMSP/SSMI (Microwave, passive)
3 platforms (DMSP F-13/F-14/F-15);
parameters: raw radiances (total column vapour water+sea wind)
Ozone
Envisat/NOAA/ERS
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data coverage
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
RESEARCH AGENCIES
NASA: National Aeronautics and Space Administration
JAXA: Japanese Aerospace eXploration Agency
ESA: European Space Agency
…(several other national agencies)
• Research Agencies promote demonstration missions, with innovative technologies
• Research instruments can provide independent information for model and/or other observations validation
• Near Real Time delivery of data is not necessarily a priority
• Research satellites pioneer future operational missions
• In principle, the life time of research missions is short (<10 years)
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
OPERATIONAL AGENCIES
EUMETSAT: EUrope’s METeorological SATellite organisation
NOAA: National Oceanic and Atmospheric Administration
• NOAA-NESDIS-DMSPJMA: Japan Meteorological Agency
Russia, China,…
• Operational Systems inherit from Research demonstration missions
• Operational Satellites are committed to Real Time delivery to end-users
• Operational missions ensure a stabilised long-life mission technology (HIRS instrument onboard NOAA satellites has lasted for ~30 years)
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Operational versus Research AgenciesThanks to a WMO initiative, R&D satellites are now fully considered as part of
the Global Observing System
Should ease the transition from research to operations
Has implications on NRT delivery requirements
Operational centres use pragmatically R&D instruments:
for model validation (POLDER, CERES,…)
for data assimilation (ERS, QUIKSCAT, AIRS,…)
Drawback of using research satellites:
Lack of visibility on the modifications of instrument calibration/configuration
Sometimes “Take it or leave it” approach…
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
ESA
ENVISAT
Heritage of ERS-2
– Multi-instrument platform
– Ozone monitoring:
» GOMOS,SCIAMACHY, MIPAS
– Sea Ocean State monitoring
» ASAR, MERIS, AATSRADM-AEOLUS
– Doppler wind lidar to provide 3D-wind coverage SMOS, EARTHCARE, WALES,…
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
EUMETSAT
Geostationary program
METEOSAT (currently 5 7)
– Infrared window and water vapour
– Visible
» (Atmospheric Motion Winds)METEOSAT 2nd GENERATION (8)
• SEVIRI (Spinning Enhanced Visible and Infrared Imager) – 12 channels (T,q,O3,..)
• GERB (Geostationary Earth Radiation Budget)Preparation of METEOSAT 3rd GENERATION (MTG)
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
EUMETSAT
Polar program
EPS: European Polar System
• Part of the Initial Joint Polar System– will include a NOAA satellite from the USA and a
METeorological Operational (METOP) satellite from Europe
• Variety of instruments– IASI (high resolution interferometer)
– ASCAT (wide swath scatterometer)
– GOME (ozone measurement instrument)
– NOAA package (HIRS/AMSU/AVHRR)
– GRAS (GPS receiver)
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
EUMETSAT Satellite Application Facilityfor Numerical Weather Prediction
Science Plan for NWP SAF deliverables
• User requirements
• ATOVS
• IASI
• MVIRI/SEVIRI (and other geostationary imagery)
• Scatterometers
• SSM/I and SSMIS
• Ozone monitoring Instruments
• Radiative Transfer Modelling
• Preprocessing packages
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data sources for the ECMWF Meteorological Operational System (EMOS).The numbers refer to all data items received over a 24 hour period in May 2003.
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data selection
Used data only
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data monitoring
Methodology
statistics obs-model guess over large samples
exchange of informations among NWP centres
Results
Blacklists
Bias corrections
Feedback to data providers (ECMWF WMO lead centre for radiosondes and pilot data monitoring)
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data monitoring: the AMV case
Comparison against model guess, aircraft and radiosondes
All sources point at an underestimation of the winds by the satellite tracking technique
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
AMV error correlations (Bormann et al., MWR 2003)
observation errors keep non-zero correlation over distances much larger than their nominal resolution
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Data monitoring: Bias correctionBias computed as a function of pressure level and solar elevation for two different sonde types
To be corrected
Not to be corrected
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Profilers
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
ECMWF Web service (http://www.ecmwf.int)
Monitoring information available now:
Data coverage maps (last 24h)
Time series (last 30 days)
Radiances monitoring
Monthly Monitoring report
GUAN
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
http://www.ecmwf.int/products/forecasts/d/charts/monitoring/coverage/dcover
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Use for forecast verification
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Recent progress: Rainfall events distribution
1 10 100 1000
Daily rainfall (mm)
0.01
0.1
1
10
100C
DF
2002-03 SYNOP reports1999-00 SYNOP reports
Distribution of daily precipitation eventsNorthern Extratropics (>20N) Dec.-Jan.-Feb., 1500 stations
4%
Of SYNOP reports exceed
10mm/day
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Use for Forecast verification
1 10 100 1000
Daily rainfall (mm)
0.01
0.1
1
10
100C
DF
2002-03 SYNOP reports18-42h forecasts (DJF 1999-00)1999-00 SYNOP reports
Distribution of daily precipitation eventsNorthern Extratropics (>20N) Dec.-Jan.-Feb., 1500 stations
1999-2000: Too many light rain…
… too few heavy rain events
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Recent progress: Rainfall events distribution
1 10 100 1000
Daily rainfall (mm)
0.01
0.1
1
10
100C
DF
18-42h forecast (DJF2002-03)2002-03 SYNOP reports18-42h forecasts (DJF 1999-00)1999-00 SYNOP reports
Distribution of daily precipitation eventsNorthern Extratropics (>20N) Dec.-Jan.-Feb., 1500 stations
2002-2003: less light rain…
… and more heavy rain events
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Recent progress: Rainfall events distribution
1 10 100
Daily rainfall (mm)
0
0.5
1
1.5FB
I
1999-20002002-2003
Frequency Biais IndexNorthern Extratropics (>20N) Dec.-Jan.-Feb.
1999-2000: T319
2002-03: T511 + changes to convection
ECMWFMet’Training Course- Data Assimilation and use of satellite data (3 May 2005)
Summary
The range of observations that are nowadays available is quite large
Data however are very inhomogeneous in quality, space and time resolution,etc…
Quality control (and bias correction) is crucial
Tools to help interpret the impact of observations on the model still in their infancy