ECMWF/GEO Workshop on Atmospheric Reanalysis

1
ECMWF/GEO Workshop on Atmospheric Reanalysis

• Held at ECMWF, 19-22 June 2006
• Programme for the meeting was developed in
  liaison with WOAP and AOPC
• Followed
• 2005 NASA/NOAA/NSF Workshop on the Development of
  Improved Observational Data Sets for Reanalysis
• 2005 ECMWF/NWP-SAF Workshop on Bias Estimation
  and Correction in Data Assimilation
• 2005 ECMWF Workshop on a potential European
  Regional Reanalysis project (EURRA)
• 2006 SCAR/CliC/ICPM Workshop on High Latitude
  Reanalysis

2
ECMWF/GEO Workshop on Atmospheric Reanalysis

• 21 participants from outside ECMWF (including
  Kent, Lorenc and Trenberth from WOAP)
• 29 presentations covering
• Experience and plans of reanalysis centres
• Status and needs for reanalysis user views
• Developments in data assimilation for reanalysis
• Observations and boundary forcing fields
• Other aspects
• Presentations may be viewed at
• http//www.ecmwf.int/newsevents/meetings/workshops
  /

3
Plenary Discussion User Aspects

• The user base
• The number of users of reanalysis products is
  considerable
• User base is broad, and extends well beyond
  climate research community
• Specific user requirements
• Full range of user activities was not
  represented, but workshop discussed
• Dynamical processes
• High-latitude processes
• Analysis quality for pre-satellite era
• Chemical transport modelling
• Long-term continuity and balance
• Horizontal resolution
• Additional output products
• Precipitation and surface fields

4
Plenary Discussion User Aspects

• General user requirements
• Comprehensive documentation of reanalyses
• Mechanism for exchange of information originating
  from users
• Measures of expected accuracy or uncertainty
• Statistical correction of products
• Continuation in CDAS mode
• Remote access to data assimilation system to
  perform OSEs
• Involvement in User Advisory Groups
• Period for future reanalysis
• For many users, satellite era is sufficient and
  improvements are still needed here
• Some interest in data back to 1940s or 1930s
• Reanalysis further back in time based on analysis
  of only surface data is feasible

5
Plenary Discussion Observations and boundary
and forcing fields

• Observations for assimilation in reanalyses
• Working Group on The On-going Development of
  Improved Observational Data Sets for Reanalysis
  supported
• Distributed data centres working to common,
  managed standards and purpose, including feedback
  from reanalysis centres
• Satellite requirements essentially covered by
  supplement to the GCOS Implementation Plan
• Data recovery needs to continue
• Data from field experiments or reference
  observation sites are needed, preferably in
  near-real-time for monitoring and use in CDASs
• Also considered
• Homogenisation of radiosonde data
• Sea-surface temperature and sea-ice distributions
• Vegetation
• Snow cover
• Precipitation
• Composition

6
Plenary Discussion Data assimilation

• Long-window, weak-constraint 4D-var
• Error estimation
• to provide information for users
• for adaptive specification of analysis statistics
• Adaptive correction of
• model biases
• observational biases
• Monitoring
• Coupled assimilation
• ocean waves, ocean circulation, land surface,
  CTMs

7
Plenary Discussion Organizational issues

• International coordination
• Role of WOAP and AOPC
• Issues are essentially as discussed in last
  years WOAP paper
• Healthy level of collaboration between reanalysis
  teams (and with data providers) at the working
  level
• Data policy issues were not discussed
  systematically, but points raised have been built
  into report, drawing on policy from ten-year
  implementation plan for GEOSS
• Need for long-term institutional funding for
  reanalysis
• Lifetime of a particular reanalysis can be a
  decade or more
• Expertise and experience should be maintained
• Ongoing advocacy for reanalysis from the
  committees, panels and working groups of GEO,
  GCOS and WCRP is still necessary