Friday, July 30, 2010

White paper 16 - Interactions with other activities

The interactions with other activities white paper is now available for discussion. When making posts please remember to follow the house rules. Please also take time to read the full pdf before commenting and where possible refer to one or more of section titles, pages and line numbers to make it easy to cross-reference your comment with the document.

The recommendations are reproduced below:
1) The development of a land surface temperature databank should allow for the extension to multivariate datasets and compatibility with global databanks from the outset.

2) Digitization activities should take a multivariate approach and always incorporate all metadata.

3) Emerging links between land and ocean dataset developers and researchers should be fostered and facilitated by the development of compatible databanks, data products and joint research projects.

4) Appropriate linkage to the activities supported by space agencies aimed at the generation of long-term climate data records addressing the GCOS Essential Climate Variables should be established when developing the surface temperature databank. The experience of GHRSST in the combined management and use of satellite and in situ data should be exploited.

5) Research is needed to improve multivariate analysis methods and to develop techniques to produce consistent global data products.

6) The important role of reanalysis in providing global multivariate analyses with wide application including quality assessment should be recognized. These products will form an essential part of a successful surface temperatures project providing both a set of estimates and a wealth of metadata regarding the data quality.

7) Funding agencies should recognize that an internationally coordinated and sustained approach to the development, maintenance and improvement of climate databanks and derived data products will have wide benefits. This most logically includes a CMIP type portal for climate data records from all observing platforms with common formats and strong naming conventions to enable ease of intercomparison.


  1. One potential interaction that is important to keep in mind is the next version of World Weather Records. As IPCC stated (, WWR has been the core of all global surface temperature data sets since the first version was initiated following a 1923 resolution by the predecessor of WMO. For every decade since then, WWR has complied monthly values of temperature, precipitation and pressure observations from many thousands of stations around the world. WMO should send out a request for the new decade of WWR in January or February. This is an opportunity to (a) try to make sure all countries are involved and (b) perhaps expand the contributions to more stations, time periods or even from monthly to daily as well.

  2. Activities for assembling regional data should be formally incorporated. When we made daily grid precipitation data for northern Eurasia, the NCDC9813 dataset was inevitable. Traceability and reproducibility are important but it is not practical to seek it in a system. So, I recommend to store original data on 1) gts-based, 2) pre-compiled (in a group) and 3) input from NMHS. It should not be a problem to switch on/off of using each data. The other activities which produce(d) 2) should be involved. I know ASEAN compendium climate database, which contains lots of temperature/precipitation data over South Asia. It was made by tremendous efforts but did not updated after 2001 or so. The specialists who were involved in this project were mostly retired from the NMHS, and the strong restriction policy remained. (Monthly database can be redistributed but daily database should not be redistributed). For the early success of creating high-resolution temperature database, activities to get permission with this kind of group should be important. In addition, not only digitizing efforts, but also for this kind of negotiation, we should take a multivariate approach and always incorporate all metadata, if situation allows.

  3. Activities to assemble database should be cooperated. Pre-compiled dataset are very important. To negotiate about pre-compiled dataset, we should take a multivariate approach.
    (I wrote long comments on this but it seems not have been successfully posted. )