To deliver climate services for the benefit of society we need to develop and deliver a suite of monitoring products from hourly to century timescales and from location specific to the global mean. Society expects
openness and transparency in the process and to have a greater understanding of the certainty regarding how climate has changed and how it will continue to
change. Necessary steps to deliver on these requirements for observed land surface temperatures were discussed at a meeting held at the UK Met Office in September 2010 attended by climate scientists, measurement scientists, statisticians, economists and software / IT specialists. The meeting followed a submission to the WMO Commission for Climatology from the UK Met Office which was expanded upon in an invited opinion piece for Nature. Meeting discussions were based upon white papers solicited from authors with specialist knowledge in the relevant areas which were open for public comment for over a month. The meeting initiated an envisaged multi-year project which this website constitutes the focal point for. It was also written up in a meeting report available as Open Access through the Bulletin of the American Meteorological Society.
For more details see:
Thorne, Peter W., and Coauthors, 2011: Guiding the Creation of A Comprehensive Surface Temperature Resource for Twenty-First-Century Climate Science. Bull. Amer. Meteor. Soc., 92, ES40–ES47. doi: http://dx.doi.org/10.1175/2011BAMS3124.1
As work continues both this site and the accompanying moderated blog (used primarily to disseminate news items and solicit comments) will form the central focal point of the effort.
The Initiative is overseen by a Steering Committee who hold regular teleconferences which are minuted. The over-arching implementation plan written by the steering committee provides a road map outlining a planned timeline for advancement to the completion of the first benchmarking and assessment cycle. As described in the BAMS article there are four principle strands to the Initiative. Each of these is expanded upon below.
1. Global Land Surface Meteorological Databank
The envisaged process includes as its first necessary step the creation of a single consolidated international databank of the actual surface meteorological observations taken globally at monthly, daily and sub-daily resolutions. In the first instance the focus has been on creating a monthly resolution temperature holdings release. Subsequent versions will consider daily and sub-daily data and additional meteorological elements as resources permit. Version 1 of the monthly resolution databank holdings was released as a beta version for a three month community review in October 2012. It consists of a number of stages as indicated by the figure above:
Stage 0: Images or hard copy of original data where available - in many cases this is a link to a third party resource or physical location.
Stage 1: Digital data records in the data format provided to the databank curators
Stage 2: Data converted to a common format with data provenance and tracking flags attached
Stage 3: Collated stage 2 sources based upon an automated merge decision algorithm. Several variants are supported so that the implications of uncertainties in how to undertake this merge can be explored by users and dataset developers.
All code used in conversion to Stage 2 and collation to Stage 3 is publicly available along with freeware compilers where necessary. At time of writing (December 2012) the method is in the process of being written up for a formal paper. The beta release recommended merge version (off which the benchmarks detailed below are envisaged to be derived) consists of over 34,000 stations.
More details on the databank construction including provenance and version control and efforts to promote data rescue can be found at the databank page. The databank itself is currently accessible through an ftp site at World Data Centre A and mirrored at an ftp site at World Data Center B.
Data submissions are strongly encouraged at any time and can be of data at any time resolution and are encouraged to contain all available meteorological elements (not just temperatures). The version 1 release holdings are now all in, but we will make periodic updates so it is not too late to submit for inclusion. Details of how to submit can be found in the data submission guidelines document. There exists both a standard data request cover letter and a certificate of appreciation countersigned by the Chairs of GCOS, WCRP and the WMO Commission for Climatology to aid efforts to facilitate data exchange.
The databank working group are supported by a data rescue task team who work with other organizations interested in data rescue issues to further this effort. By a conservative estimate as much data exists prior to 1950 in solely hard copy or image form as is currently available digitally. Rescuing this data would undoubtedly help to improve understanding of historical climate variability and change.
2. Creation of new data products
From this databank it is hoped that many investigators will attempt to create surface temperature data products. A range of approaches is required both to quantify the range of plausible estimates of how climate has evolved and also because different approaches will have different strengths and weaknesses and therefore be more or less useful for a given application. Efforts will be encouraged from outside the climate science community. Beyond taking all reasonable efforts to encourage the development of new algorithms and approaches the success of this aspect largely depends upon broader community buy-in. Experiences in other areas of science and even climate science show the value of having multiple independent teams of analysts look at the same piece of data to enhance our understanding.
3. Benchmarking and assessment of submitted data products
"All animals are equal but some animals are more equal than others" - George Orwell, Animal Farm
The real world climate system evolved along one single trajectory. Because of inhomogeneities within the station records and the inadequate and time-varying spatial sampling it is inevitable that multiple, somewhat distinct, estimates will result from running different algorithms on the recommended merge version of the databank. The question then is whether there is a away to ascertain which estimates may be more plausible. Sadly, we do not know the single real world solution (and if we did this Initiative would not exist and we would be working on other things). What we can do is produce analogs to the real world which have similar spatial and temporal characteristics and identical sampling. Through constructing a number of such analogs under different assumptions as to the structure of inhomogeneities we can begin to ascertain which data products may be closer to the real world when applied to the real world databank observations.
To ascertain their fundamental quality a common benchmarking exercise will be undertaken that data product creators partaking in activities associated with the Initiative will be invited to submit their efforts to. This benchmarking process is envisaged to be double blind and cyclical with each cycle lasting approximately 3 years. It is undertaken under the auspices of an international working group. The envisaged mechanics are given in the figure below:
4. Serving products and advice to end-users in an appropriate manner
Finally, the project will aim to create a set of user tools and provide these value added products back to the data providers and to society to enable critical decision making. Provision of such tools and advice is key to realising the scientific benefits that this program will deliver to society more generally. Work on this aspect can only really begin once the databank is available and products are starting to be created from it to populate the holdings.