The aims of this study, which are to be undertaken by a consultancy, are:

1. To compile a consolidated account of the existing UOT data base, using WOCE, Levitus and whatever other data bases that are available. The attributes that we seek to quantify are:
   
   • The sampling as a function of space and time, extending back to at least the early 1980s. These statistics should be arranged by longitude and latitude bins (2? x 2?) as well as by traditional lines/regions (as defined in TOGA/WOCE). The statistics should give some measure of the vertical sampling and depth.
   • A measure of the continuity for specific lines/regions, including the relative information content of low-density, frequently repeated and high-density lines.
   • A measure of the quality (level of QC that the data have been subjected to).
   • Availability - delayed mode versus real-time; gaps in delivery systems; public versus private; etc. permanent archives
   • Availability of metadata
   • Logistical considerations with respect to particular tracks
   • Any other political/technical attributes which might impact/qualify the attributes.
   The analysis should include all forms of sampling (broadcast, high-density, frequently-repeated, ad hoc, etc.)
2. Produce consolidated "maps" of information level/content based on the dominant scales of climate signals. The raw distribution statistics do not take account of the actual information content so it is useful to seek some consolidation of the information in (1) even if it does depend on certain assumptions. The SOOP contribution should be identified.
3. Document the existing practices for assembling, quality control and distribution of upper ocean data, working from existing material of GTSPP, WOCE UOT/DPC and IGOSS SOOP.
4. Document to the extent possible the "value adding" of thermal data process chains, be they automated assimilation (e.g., NCEP, BMRC, UKMO, ECMWF), quick-look/semi-automated quality control (e.g., GTSPP, AOML) or higher-level scientific quality control and assembly (e.g., CSIRO, AOML, Levitus, NODC).
5. Provide quantitative assessment of all SOOP lines. This should include an assessment of relevance/impact against scientific objectives including seasonal-to-interannual prediction, environmental/ocean prediction, improved climatologies and climate change monitoring, scores against key attributes (continuity, quality, etc.), notes on extenuating circumstances and the existence of proxies in the event of gaps/discontinuities in the lines. The broad-scale sampling should also be assessed as a precursor to Argo with a view to maintaining the temporal and spatial integrity of resolved signals such as the global ENSO wave, the Antarctic Circumpolar Wave, decadal variability, etc.
6. On the basis of (5), provide a renovated SOOP plan including broadcast and high-density strategies, taking account of, as far as is practical
   
   • The existence (or potential) of other direct sampling networks (e.g., TAO, Argo);
   • The indirect information available from remote sensing, particular altimetry; and
   • The indirect information available from models, e.g. wind-forced equatorial.
7. Produce a report (through a consultancy) based on the above which will form the background for a Workshop to be convened in the 3rd quarter of 1999. The Executive Summary from this process will constitute a key paper for the OCEANOBS99 Conference, 18-22 October 1999. The Report will be published jointly under the joint auspices of the GCOS/GOOS/WCRP OOPC, IOC/WMO IGOSS SOOPIP and CLIVAR UOP.

• Background
• Objectives
• Method
• Atlantic Ocean
• Indian Ocean
• Pacific Ocean
• Workshop