ECMWF MetTraining Course- Data Assimilation and use of satellite data (3 May 2005) The Global Observing System Overview of data sources Data coverage Data.

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