Note: When you click on links below you may be leaving the CLIMAR-II website.
Recommendations from CLIMAR-II, Brussels 2003.
E. Metadata and Homogenisation
Digital availability of the entire record of the WMO ship catalogue (Publication No.47; 1955-), in a format suitable for use in association with both operational and climate data, should be made a priority. Editions for 1955-72 and 2005 are not yet available in digital form.
Status: Good progress
Notes: WMO Pub. No. 47 is now available in machine-readable form for the full period 1973 to the end of 2004 (as of July 2004). Editions for 1955-72 have been imaged by NOAA's Climate Database Modernization Program (CDMP) and await digitisation in 2005. More information can be found on: icoads.noaa.gov/etmc/ (modern metadata: document 5.1; historical metadata: document 5.3)
Observing practice literature, both national and international, is an import
ant aspect of climate metadata. Two of the more important decisions recorded in this literature were the historical WMO/Commission for Marine Meteorology (CMM) decisions which improved VOS data and the Marine Climatological Summaries Scheme (MCSS). To document the evolution of observing practice, a procedure for identifying, archiving and distributing this type of metadata should be developed. The archive should be updated through JCOMM and its ETMC, without destroying the older entries, when observational practice is updated. Eventually, the archive could also link to the results of instrument validations and comparison studies.
Status: Some progress
Notes: The ETMC have investigated whether all versions of the WMO Manual on Codes (WMO-No. 306) and information on IMMT formats documenting the history of the marine ship codes and exchange formats are available. More information on icoads.noaa.gov/etmc/, document 4.2. Many relevant documents have now been scanned by JMA, and through additional efforts by the NOAA Climate Database Modernization Program (CDMP), made available as searchable PDFs at this location: http://goos.kishou.go.jp/ws/ETMC/code_task/ (availability from CDMP is also planned).
An archive of metadata for moored and drifting buoys, and other ODAS (e.g. offshore platforms), should be filled by Members, with WMO coordination, as soon as possible with information on both current and historical deployments.
Notes: A web based buoy metadata collection scheme is being developed at JCOMMOPS thanks to EGOS funding. Developments started in January 2004. Implementation is planned for early 2005 to mid-2005. Collected metadata will be publicly available and submitted to the JCOMM ODAS metadata database. The database will be hosted in China (see JCOMMOPS website for more details: http://wo.jcommops.org/cgi-bin/WebObjects/JCOMMOPS)
If possible, a given buoy should have a unique identifier. The re-use of identifiers (buoy numbers) for different buoys can cause erroneous application of metadata. If buoy numbers must be reused, the metadata should include sufficient features (e.g., timestamps) so that they can be correctly applied.
Metadata, including information on homogeneity adjustments applied, should be clearly linked to data.
Status: Some progress
Notes: A subset of Publication No. 47 metadata has been linked to ICOADS individual ship reports for the period 1973 to 2002 using the metadata attachment to the IMMA format. The metadata subset has also been prepared and matched to real-time data for 2003-2004, which we plan to make available in IMMA format by late 2005 as part of an ICOADS update for that period.
It remains essential to acquire data from independent platforms (e.g. VOS, buoys, research vessels, satellites), to allow independent validation and homogenisation of records. The important VOSClim data validation and improvement project should be continued.
Notes: The decline in the numbers of observations from VOS was highlighted at the 3rd meeting of the JCOMM Ship Observations Team and will be the subject of a poster at the JCOMM Scientific Conference, Operational oceanography and marine meteorology for the 21st century, 15-17 September 2005 and at the JCOMM-II 19-27 September 2005, both in Halifax, Canada. See also D2.
There is a need to investigate the best way of applying wind homogenization techniques in the absence of adequate metadata.
Proxy data (e.g. coral-based SST estimates) should be carefully matched with instrumental data, following the GCOS Climate Monitoring Principles. Error-adjusted annual fields may help in this process.
Continue efforts to make QC of data more consistent and effective, including documenting and homogenising the methods used as much as possible.
For further information, or to provide status updates, contact: