GCOS GOOS WCRP/OOPC IX/3
The VOSClim dataset is a valuable resource for climate researchers. Its wide use must
be encouraged. There are three distinct phases in the scientific analysis for VOSClim. Firstly,
the data collected must be thoroughly assessed, using the metadata to determine bias and
scatter in the observations and how they depend on observational practice and environmental
conditions. The second phase will use this information to develop a strategy for the
production of a high quality dataset from the data collected by the VOSClim ships and
promote good practice amongst the VOS by feeding back the results to the VOS operators.
Only then can the third stage, the use of high quality data for climate science, begin. At
present, the only scientific input has come from the VOSClim Scientific Advisors at the UK
Southampton Oceanography Centre. It is desirable that a wider range of scientists have input
at each stage. Expanding the user base should be actively encouraged to promote a diversity
of research and ensure that this dataset is used to its full potential5. It is further desirable to
promote the wider analysis of error and bias in historical VOS datasets such as ICOADS
(Diaz et al. 2002, see Section 7.3 of GCOS 2003).
Action: AOPC should encourage the wider scientific exploitation of VOSClim dataset.
In recent years the practice of making observations has been made easier by the
introduction of automated reporting software such as TurboWin, SEAS or OBSJMA. This has
reduced the burden of coding reports on the ships officers and provides help, for example,
with making cloud reports by including pictures of different types of cloud. This type of
software should have reduced errors due to the incorrect calculation of true winds from ship
relative winds. However the use of TurboWin logging software has resulted for the first time
in the implementation of a WMO directive to correct winds to 10 metres height at source
(Shearman and Zelenko, 1989). SOT II recognised this as a problem, particularly for climate
research, as there is no metadata to show which reports have been corrected to 10 metres and
which have not. SOT II has set in motion the process to revoke the WMO height correction
directive, and in the shorter term the TurboWin developers will remove the height correction
from the next version (although leaving height correction for fixed platforms reporting in ship
code following consultation with representatives of NWP centres). An interim mechanism
using footnotes in WMO Publication 47 should allow the identification of some of the reports
that have been height corrected at source from metadata. With pressure of time on ships'
officers, there has been a move towards fully automated systems by some countries, notably
Canada and France. A good quality automated system, with the facility for manual input of
some parameters such as cloud types, has the potential to produce climate quality reports. A
further advantage is the reliable delivery of frequent observations in severe weather
conditions. However, care needs to be taken that automatic systems are introduced in line
with the GCOS Climate Monitoring Principles6 (GCOS, 2003) and ensure that enough ships
still record the full range of variables required for surface flux calculation.
Note for the OOSDP: the lack of progress on the SURFA project has removed one potential early
customer for the data set.