EUMETSATs Network of
Satellite Application Facilities
Lothar Schüller
EARSel Workshop Bern February 2014
Outline
•
EUMETSAT
•
SAF Programmatic framework
•
SAF concept and establishment
•
SAF structure and processes
•
SAF product characteristics
•
Current SAF Network
•
Committed SAF snow products
•
SAF interaction with users
•
SAF phasing
•
Research and/to Operations
Slide: 2
EARSel Workshop February 2014
Bern
Current EUMETSAT satellites
METOP A-B
(LOW-EARTH, SUN – SYNCHRONOUS ORBIT)
EUMETSAT POLAR SYSTEM/INITIAL JOINT POLAR SYSTEM
METEOSAT 8-9-10 (2nd GENERATION)
JASON-2
(LOW-EARTH, 63° INCL. NON SYNCHRONOUS ORBIT)
OCEAN SURFACE TOPOGRAPHY MISSION
METEOSAT – 7 (1st GENERATION)
(GEOSTATIONARY ORBIT)
(GEOSTATIONARY ORBIT)
INDIAN OCEAN DATA COVERAGE MISSION AT 57°5 E
TWO-SATELLITE SYSTEM:
- METEOSAT-9: FULL DISK IMAGERY MISSION AT 0° (15 MN)
- METEOSAT-8: RAPID SCAN SERVICE OVER EUROPE AT 9.5°E (5 MN)
EUMETSAT ground segment overview
EUMETSAT geostationary systems
Meteosat satellites
Jason-2
Optional programme for ocean altimetry
Data acquisition
and control
EUMETSAT Polar System
NOAA satellites
Metop satellites
Flight Operations
Other satellites
Pre-processing
EUMETSAT HQ
Data Centre
EUMETSAT HQ Darmstadt
data sources
Meteorological product extraction
EUMETSAT HQ Darmstadt
Centralised processing
and generation of products
Satellite Application Facilities
within the Member States
Decentralised processing
and generation of products
Data dissemination
via EUMETCast
Applications
Ground
Segment
Users
Slide: 4
EARSel Workshop February 2014
Bern
Third-party
What is a SAF?
• SAF = Satellite Application Facility
• part of the EUMETSAT application
ground segment
• complement production of standard
meteorological products at EUMETSAT
central facility
Satellite
Application
Facilities
SAFs
• providing products and services to users
• specialised on topics and themes
• located at Weather Services in
EUMETSAT Member and Co-operating
States
• developed and operated by consortium
of partners
Slide: 5
EARSel Workshop February 2014
Bern
What is a SAF?
• SAF is developed and operated in
a consortium
Partner
Institute
Partner
Institute
• involving partners in EUMETSAT
Member and Co-operating states
Partner
Institute
Leading
Entity
Partner
Institute
• National Meteorological
Services
Partner
Institute
EARSel Workshop February 2014
• Other operational
organisations
• Universities and Research
Institutes
SAF
Slide: 6
• lead by a Leading Entity (a
National Meteorological Service)
Bern
How do SAFs work? Providing data products
Some SAFs
provide
data products
For each pixel
do
derive data
done
end
User
SAF
Slide: 7
EARSel Workshop February 2014
User site
Bern
How do SAFs work? Providing Software products
Some SAFs
provide
software
products
For each pixel
do
derive data
done
end
For each pixel
do
derive data
done
end
User
SAF
Slide: 8
EARSel Workshop February 2014
User site
Bern
Nature of SAF products
The objective of SAFs is to provide operational products.
“Operational” means:
Continuity of product provision
Continuity of product improvements
Continuous quality monitoring
Committed user services
Validation and review before official release
Complete Documentation of Products, Algorithms,
Validation Results
• ...
•
•
•
•
•
•
Slide: 9
EARSel Workshop February 2014
Bern
EUMETSAT’s 8 Satellite Application Facilities
1
Support to Nowcasting and Very Short Range Forecasting
2
Ocean and Sea Ice
3
Climate Monitoring
4
Numerical Weather Prediction
5
Land Surface Analysis
6
Ozone and Atmospheric Chemistry Monitoring
7
Radio Occultation Meteorology
8
Support to Operational Hydrology and Water Management
SAF Consortium Member
Slide: 10
EARSel Workshop February 2014
Bern
Hydrology SAF
 SAF on Support to Operational Hydrology and
Water Management
 Leading Entity: Italian Meteorological Service
(USAM)
 SAF products focuses on
 precipitation
 soil moisture
 snow parameters
 utilisation of these parameters in hydrological
models and NWP
Slide: 11
EARSel Workshop February 2014
Bern
Currently available snow products from H SAF
“operational”:
• Snow detection (snow mask) by VIS/IR radiometry
• Snow water equivalent by MW radiometry
“pre-operational”:
• Effective snow cover by VIS/IR radiometry
• Snow status (dry/wet) by MW radiometry
Slide: 12
EARSel Workshop February 2014
Bern
Snow detection by VIS/IR radiometry
•
•
•
•
•
•
Binary map of snow / no-snow situation.
Using Meteosat/SEVIRI observation
Different methods used for flat/forested and mountainous regions
Daylight product, output result every 24h
Probability Of Detection (POD): Flat / Forested areas: 85 %
Mountainous areas: 70%
False Alarm Rate (FAR): Flat / Forested areas: 15 %,
Mountainous areas: 20%
Slide: 13
EARSel Workshop February 2014
Bern
Snow status (dry/wet) by MW radiometry
•
•
•
status of the snow mantle, whether it is wet or dry and, in time
series, thawing or freezing.
Multi-channel MW observations SSMIS on DMSP (middle
frequencies), and the algorithm is based on thresholding.
Requirements:
• Hit Rate (HR): 80 %
• False Alarm Rate (FAR): 10 %
Slide: 14
EARSel Workshop February 2014
Bern
Effective snow cover by VIS/IR radiometry
•
•
•
•
Multichannel (VIS, NIR, IR) analysis
AVHRR on NOAA and Metop satellites
Requirement: 65% overall accuracy
Validation against snow observing stations
Slide: 15
EARSel Workshop February 2014
Bern
Snow water equivalent by MW radiometry
•
•
•
•
•
Maps of snow water equivalent derived from MW measurements SSMIS on
DMSP sensitive to snow thickness and density.
Algorithm based on assimilating MW brightness temperatures of several
channels at frequencies with different penetration in snow, into a first-guess
field built by the (sparse) network of stations measuring snow depth.
Requirements:
Flat / Forested areas: 20mm
Mountainous areas: 25mm
Slide: 16
EARSel Workshop February 2014
Bern
Interaction with Users:
Services provided to users
Documents
• ATBD
• Product User Manual
• Validation Report
Near-Real-Time Products
EUMETCast
For each pixel
do
derive data
done
end
Software and
SAF
off-line products
Internet
(FTP, WEB)
EUMETSAT
Data Centre
Slide: 17
EARSel Workshop February 2014
Bern
User
Interaction with Users:
Feedback provided to SAFs
For each pixel
do
derive data
done
end
User
SAF
User
Helpdesk
Workshops
Beta user
programme
Slide: 18
EARSel Workshop February 2014
Bern
SAF Network Phasing
1997
2007
Developments and Initial Operations
SAF
2012
CDOP
2017
CDOP-2
2022
CDOP3
NWC SAF
OSI SAF
O3M SAF
NWP SAF
CM SAF
ROM SAF
LSA SAF
H-SAF
MSG
Slide: 19
EARSel Workshop February 2014
Bern
EPS
MTG
EPS-SG
Synergy: Research -> Operations
• SAFs embedded programmatically in the EUMETSAT mandatory
programmes, currently secured until 2022.
• Current proposal for a EPS Second Generation Programme foresees
funding for SAF development and operations until 2032.
• The purpose and commitment of SAFs is to:
• deliver operational products addressing requirements of the
EUMETSAT member states
• develop new products implementing mature retrievals validated by
Research (transition from Research to Operations)
• Next SAF development and operations phase (CDOP-3 for 20172022): proposals expected end of 2015.
Slide: 20
EARSel Workshop February 2014
Bern
SAF implementing mature research algorithm
Own
Algorithm
Development
Implementation
in processing
chain
Operational
Data
products
Slide: 21
EARSel Workshop February 2014
Bern
SAF implementing mature research algorithm
Own
Algorithm
Development
External
Algorithm
Development
Implementation
in processing
chain
Operational
Data
products
Slide: 22
EARSel Workshop February 2014
Bern
Operational
Data
products
Summary
 SAF = Satellite Application Facility
 providing products and services to users on an
operational basis with a long-term perspective
 specialised on topics and themes
 located at Weather Services in EUMETSAT
Member and Co-operating States
 developed and operated by consortium of
partners
 part of the EUMETSAT application ground
segment
 complement production of standard
meteorological products at EUMETSAT central
facility in Darmstadt
 Data and software are usable free of charge
Slide: 23
EARSel Workshop February 2014
Bern
BACKUP SLIDES
Slide: 24
EARSel Workshop February 2014
Bern
Why SAFs ? Facts and Objectives
Facts:
• New generation of Meteorological Satellites (MSG and Metop as well as MTG
and EPS-SG) have much wider areas of application
• specific expertise available in EUMETSAT's Member and Cooperating States
Objectives:
• SAF concept encourage the utilisation of existing skills and infrastructure in
Member- and Cooperating States for developing geophysical data products
and services
• Facilitating cost-effective exploitation by ensuring services are distributed in
the most appropriate way
• SAFs improve the ability of EUMETSAT Member States to exploit satellite
data
• Fostering development of cooperation with non-Member States and other
organisations
Slide: 25
EARSel Workshop February 2014
Bern
Nowcasting SAF
“Support to Nowcasting and Very Short Range
Forecasting” (NWC SAF)
 established to utilise the new data from
Geostationary satellites (Meteosat and others) and
the polar platforms (Metop, NOAA, Soumi NPP) for
enhancing Nowcasting
 Development of Software packages for the
operational extraction of products relevant to
Nowcasting and for local installation
 Leading Entity is the Spanish Meteorological Agency
AEMET in Madrid
 Regular updated software packages since October
2004
Slide: 26
EARSel Workshop February 2014
Bern
Rapidly developing Thunderstorm Product
Ocean and Sea Ice SAF
Sea surface temperature
 Ocean and Sea Ice (OSI) SAF routinely
produces and disseminates products
characterising the ocean surface and the
energy fluxes across the sea surface
 Operationally produces information on the sea
ice characteristics (extend, concentration, …)
 Leading Entity is Météo-France in Lannion
 OSI SAF distributes near real-time products
based on NOAA, MSG, Metop, Oceansat,
DMSP and GOES data
Sea ice concentration
Slide: 27
EARSel Workshop February 2014
Bern
Ozone SAF
 SAF on Ozone and Atmospheric Chemistry
Monitoring (O3M SAF)
 developed for the processing of data on
ozone, other trace gases, aerosols and
ultraviolet radiation
 Emphasis on the Global Ozone Monitoring
Experiment (GOME-2) and IASI on EPS
(Metop)
 Leading Entity is the Finnish
Meteorological Institute FMI, Helsinki
 First release of products in Summer 2007
Slide: 28
EARSel Workshop February 2014
Bern
Climate Monitoring SAF
 SAF on Climate Monitoring
 generates and archives high-quality data-set for specific
climate application areas
 Currently concentrates on:
 cloud parameters
 radiation budget parameters
 atmospheric humidity
 Leading Entity is the German Weather Service DWD,
Offenbach
 NOAA-AVHRR based data operationally produced since
November 2004, MSG based data from October 2005,
Metop data used since 2009.
 Climate Data Records: 20 years of SSM/I Water Vapour
information released in 2009.
 NOAA-AVHRR based 20 years of homogeneous data
record (clouds, surface radiation) released 2012
 Upcoming Climate Data Records based on SEVIRI,
ATOVS, SSM/I
Slide: 29
EARSel Workshop February 2014
Bern
NWP SAF
 SAF on Numerical Weather Prediction (NWP
SAF)
 aims at increasing the benefits to Met.Services from Numerical Weather Prediction
(NWP)
 develops advanced techniques for the
effective use of satellite data
 development and maintenance of RTTOV
radiative transfer model, ATOVS and AVHRR
Pre-processing Package (AAPP)
 Leading Entity is the UK MetOffice, Exeter
Slide: 30
EARSel Workshop February 2014
Bern
ROM SAF
 SAF on Radio Occultation Meteorology
 GRAS: Global Positioning System (GPS) Receiver for
Atmospheric Sounding flown on EPS/Metop satellites
 near real-time and offline:
 sounding data (temperature, pressure, humidity)
 corresponding validation products, and
 assimilation software
 The Leading Entity is the Danish Meteorological Institute DMI,
Copenhagen
 Software packages released since 2007, first NRT product
dissemination in October 2008
Slide: 31
EARSel Workshop February 2014
Bern
Land Surface Analysis SAF
 SAF on Land Surface Analysis (LSA SAF)
 established to increase the benefit from
MSG and EPS data related to land, landatmosphere interaction and biospheric
applications
 Generates operationally data services
related to Surface Radiation, Vegetation
and wild fire
 Leading entity is the Portuguese Institute
for Meteorology IM, Lisbon
Slide: 32
EARSel Workshop February 2014
Bern
Hydrology SAF
 SAF on Support to Operational Hydrology and
Water Management
 Leading Entity: Italian Meteorological Service
(USAM)
 SAF products focuses on
 precipitation
 soil moisture
 snow parameters
 utilisation of these parameters in hydrological
models and NWP
Slide: 33
EARSel Workshop February 2014
Bern
SAF Development and Service Model
Initial and
new User
Needs
Needs for Changes
and Improvements
User Provided
Operational
Feedback
Validated
Products
Initial
Objectives
SAF Research
and Development
SAF Operations
Delivered SAF
Services
Beta-User
Tasks
Slide: 34
EARSel Workshop February 2014
Bern
CDOP
review logicand Service Model
SAF Development
Committed operations baseline
Previous Phase
Operational
Readiness
Reviews
Development
Life Cycle
of new operational chains
RR
Requirements
Reviews
PCR
New/enhanced
Products
Definition
Slide: 35
EARSel Workshop February 2014
ORR
Products
Distribution
updated
baseline
Operations
Reporting
User Support
Provision
Product
Consolidation
Reviews
Products
Improvement
Implementation
Bern
OR
updated
baseline
User Workshops
and Consultation
Operation
Reviews
CDOP-2: Second slice of the Continuous
Development and Operations Phase
SAF Network after 2012:

funding for 5 years of CDOP-2 and 5 more years for a CDOP-3 foreseen
within the MTG budget

CDOP-2 started March 2012

Operational Continuation of existing products

Development of new products and services (list of committed SAF products
soon to be available).

Preparation activities of Meteosat Third Generation (MTG) based products

Exploitation of synergies within the Application Ground Segment (SAF
Network + Central Facilities)

Reprocessing and homogeneous data set generation (e.g. for climate
applications)

Integration and interfacing with other initiatives: GMES, WMO SCOPE-CM,
EUMETCal, GHRSST, GODAE, ESA CCI, etc.
Slide: 36
EARSel Workshop February 2014
Bern
Examples for new Developments in CDOP-2
Long-term satellite data records:
• CM SAF: Meteosat since 1982
• O3M SAF: GOME-2 since 2006
• OSI SAF: Sea Ice since 1978,
Sea Surface Temperature
since 2002
• ROM SAF: Reprocessing of all
GNSS-RO missions (since
1995)
• LSA SAF: reprocessing of
SEVIRI products (since 2002)
Slide: 37
EARSel Workshop February 2014
Bern
Application: PV GIS
Country-scale and Europe / Africa Maps
Slide: 38
EARSel Workshop February 2014
Bern
Application: SMA Solar Checker
www.sma.de
→ Products
→ Plant Planning
Slide: 39
EARSel Workshop February 2014
Bern
Examples for new Developments in CDOP-2
Bringing external developments
into operations:
• O3M SAF: IASI products
developed at LATMOS
• LSA SAF: NDVI (vegetation
index from VITO, Belgium)
• NWC SAF: Precipitation
based on cloud microphysics
(KNMI)
Slide: 40
EARSel Workshop February 2014
Bern
V2013, GEO-CRRv4.0 outlook
CTMP Function: Based on the one developed by Roebeling and Holleman*
Cloud Top Microphysical Properties used by this algorithm:
- Phase (Ph)
- Effective radius (Reff)
- Cloud water path (CWP)
Liquid water path (LWP)
Ice water path (IWP)
Two steps:
1.- Delimitation of the precipitation area
2.- Assignment of rain rates
(*) Roebeling, R. A. and I. Holleman, 2009: SEVIRI rainfall retrieval and validation
using weather radar observations. J. Geophys. Res., VOL. 114, D21202.
Slide: 41
EARSel Workshop February 2014
Bern
41
PREPARATIONS FOR V2013 –
DAY-TIME ALGORITHM
Example over Spain
22th August 2008 at 14:00 UTC
CRR (3D Matrices)
CRR (CTMP function)
Radar (PPI)
Slide: 42
EARSel Workshop February 2014
Bern
42
42
PREPARATIONS FOR V2013 – DAY-TIME ALGORITHM
Comparison: 3D Matrices vs CTMP function
Dataset:
Spain: 46 days, May-September 2008, 10:00 – 14:00 UTC every 30 min
Accuracy Statistics
Categorical Statistics
2,5
90
80
2
70
60
3D Matrices
CTMP Function
1
(%)
1,5
50
3D Matrices
40
CTMP Function
30
20
0,5
10
0
0
MEAN
ME
MAE
RMSE
FAR
POD
CSI
PC
RMS error decreases with CTMP function because precipitation maxima are better located and precipitation
pattern is more accurate.
False alarms increases 1% while probability of detection increases 24% for this dataset.
Slide: 43
EARSel Workshop February 2014
Bern
43
43
Examples for new Developments in CDOP-2
New satellite sensors:
• OSI SAF: OSCAT based wind
products
• H SAF: GPM data
• NWC SAF: Soumi NPP VIIRS,
additional GEO imager
• ROM SAF: cosmic
• Preparation for MTG and EPS-SG
Slide: 44
EARSel Workshop February 2014
Bern
Development of PR-OBS-2
Precipitation rate at ground by MW cross-track scanners
CDOP-2
CDOP-1
NN trained with
NN trained
global Neural Network
algorithm trained with a global CRD: not optimized
Initial
VersionwithH02
Cloud Dynamics
and Radiation
Database
Cloud Radiation DatabaseCRD
for Europe/Mediterranean
area; using
different CRM
than H01CDRD
Operational
New Version
• Correction of corrupted channels [i.e., MetOp-A, AMSU-A Channel 7
Ver. 1
H02
(54 GHz)]
• Generation of a new version of the neural network algorithm,
trained by the same CDRD as H01:
-Representativeness for Europe/Mediterranean basin
Rome,
-Harmonize products from cross-track and conical
MW Italy
scanners
• Improvement of the screening procedures (under
particularly
Flash
Flood
unfavourable backgrounds (e.g., snow cover, coastlines)
20 Oct 2011
• Quality flag and indication of phase (under verification)
Under
verification
Operational Dec.
2012
8:20 UTC
MetOp-A
AMSU/MHS
Ver. 1
H02A
H-SAF Area
Ver. 1
H02B
Extension to Full Disk Area (exp. op. 2014)
Ver. 2
H18A
H-SAF area: introduction of SSI index. (exp. op. 2017)
Ver. 2
H18B
Full Disk Area (exp. op. 2017)
Slide: 45
EARSel Workshop February 2014
Bern
45
Concept of the Hydrological Validation
Slide: 46
EARSel Workshop February 2014
Bern
Test catchments of the Hydrological Program
Slide: 47
EARSel Workshop February 2014
Bern
Examples for new Developments in CDOP-2
New product developments:
• NWC SAF: Convection Initiation, Extrapolated Imagery,
precipitation approach
• O3M SAF: trace gases (Formaldehyde, BrO, ..) and aerosol
properties from GOME-2
• LSA / CM SAF: long-wave surface radiation from SEVIRI
Slide: 48
EARSel Workshop February 2014
Bern