Thursday, July 24, 2014

The WMO Commission for Climatology meeting and developments on the WMO front



The World Meteorological Organization’s Commission for Climatology had its four-yearly meeting in Heidelberg, Germany, from 3-8 July, preceded by a Technical Conference from 30 June – 2 July. The Commission is the central body for climate-related activities in WMO, and has a major role in establishing international standards and setting international work programs in the climate field, particularly through setting up networks of Expert Teams and Task Teams to work on particular issues. Its  President (re-elected at the meeting) is Tom Peterson of NCDC, who will be well-known to many of you. The International Surface Temperature Initiative was set up as the result of a resolution of the last Commission for Climatology meeting, in 2010.

I made a presentation to the Technical Conference on the current status of ISTI. By happy coincidence, this presentation was scheduled for the morning on 1 July, a few hours after the release of the first version of the ISTI databank. The presentation appeared to be well-received; there were few direct questions or follow-ups, but the pile of leaflets we brought describing ISTI (once they got there, after a couple of bonus days enjoying Berlin with the rest of my luggage) was a lot smaller at the end of the week than it was at the start. One particular reason for targeting the Commission audience is that many of the attendees at Commission meetings are senior managers in their national meteorological services (often the head of the climate division, or equivalent), and so potentially have more influence over decisions to make data available to projects such as ISTI than individual scientists would.

Slow progress is also being made in two other areas of WMO of interest to ISTI. The inclusion of at least some historic climate data amongst the set of products which countries agree to freely exchange has been a long-standing goal of ours. The key decisions on this will be made at the full WMO Congress, which will be held next year, but progress to date (including through the recent WMO Executive Council meeting) is encouraging. There are also moves to include the month’s daily data in monthly CLIMAT messages, which are the principal means of exchanging current climate data through the WMO system but currently only contain monthly data. This will be very useful for the ongoing updating of data sets, as it will make daily data available which can be assumed to be for a full 24-hour day and is likely to have received at least some quality control (neither of which is necessarily true for the real-time synoptic reports which are the primary current source of recent daily and sub-daily data). Considerable technical work remains to be done, though, to implement this, even once it is formally endorsed.

Data rescue and climate database systems continue to be a high priority of the Commission, with several initiatives outlined at the meeting. Among them are proposals for an international data rescue portal, which (among other things) would potentially facilitate crowd-sourced digitisation. It is, however, an indication of how much work still remains to be done in many parts of the world that, according to results of a survey reported at the meeting, 25% of responding countries still stored their country’s climate data in spreadsheets or flat files, and 40% had a climate database system which was not fully functioning or not functioning at all.

The Commission also agreed to establish a new Task Team on Homogenisation. The full membership (and chairing) of this group are not yet clear but I will almost certainly be part of it. This team will be working closely with ISTI, but will also have a major focus on supporting the implementation of homogenised data sets which contribute to operational data products nationally and internationally.

Also of interest to ISTI is a new WMO initiative to formally recognise “centennial stations”, which, as the name implies, are stations which have existed with few or no changes for 100 years or more. Countries are to be asked to identify such stations, whose data will clearly be of considerable value to ISTI, if not already part of our databank. Free access to data and relevant metadata are among the recommendations for centennial stations.

And one advantage of holding an international meeting during the World Cup: it provides an instant conversation-starter with delegates of almost any country. (Perhaps fortunately for the Brazilian delegation, the meeting finished just before the semi-finals).

(Update 5 August: the resolution which came out of the WMO Executive Council meeting is available at


Monday, July 21, 2014

Later talks from SAMSI / IMAGe workshop

Following up to the previous post there were two further recorded talks at the event.

http://video.ucar.edu/mms/image/samsi2014_jonathan_woody.mp4 - Jonathan Woody gave a talk on analyses of snow depth.

http://video.ucar.edu/mms/image/samsi2014_bo_li.mp4 - Bo Li provided a talk on model selection in the use of palaeodata analyses.

We hope to have a meeting report out within a matter of days to weeks. We will post this here.

Overall there was a lot of active participation and many new directions to be taken in the analysis of surface temperatures. Our thanks go out to both SAMSI and IMAGe for facilitating this meeting and to all the participants for being active. More details to appear soon ...

Friday, July 11, 2014

Talks from SAMSI / IMAGe workshop on International Surface Temperature Intiative

We are currently in-situ in Boulder at a workshop organized with SAMSI and IMAGe. Work is ongoing and a formal write up will follow after completion. Most, but not all, of the talks have been streamed and are available for viewing.

http://video.ucar.edu/mms/image/samsi2014_richard_smith.mp4 - Richard Smith provided an overview of the SAMSI program and their expectations for the workshop.

http://video.ucar.edu/mms/image/samsi2014_peter_thorne.mp4 - I provided an overview of the ISTI program and progress to date

http://video.ucar.edu/mms/image/samsi2014_peter_thorne_jared.mp4 - I deputized for Jared to provide an overview of the databank process

http://video.ucar.edu/mms/image/samsi2014_kate_willet.mp4 - Kate Willett provided an overview of progress with creation of benchmarks and remaining challenges.

http://video.ucar.edu/mms/image/samsi2014_lucie_vincent.mp4 - Lucie Vincent provided an overview of typical inhomogeneities found in station timeseries and some of their likely causes.

http://video.ucar.edu/mms/image/samsi2014_jaxk_reeves.mp4 - Jaxk Reeves provided an overview of at most one changepoint techniques.

http://video.ucar.edu/mms/image/samsi2014_colon_gallagher.mp4 - Colin Gallagher provided an overview of fitting regression models.

http://video.ucar.edu/mms/image/samsi2014_robert_lund.mp4 - Robert Lund provided an overview of multiple changepoint techniques.

http://video.ucar.edu/mms/image/samsi2014_enric_aguilar.mp4 - Enric Aguilar and http://video.ucar.edu/mms/image/samsi2014_victor_venema.mp4 Victor Venema provided an overview of several state of the art climate homogenization techniques.

http://video.ucar.edu/mms/image/samsi2014_matt_menne.mp4 - Matt Menne provided and overview of the Pairwise Homogenization Algorithm and Bayes Factor Analyses by NCDC and their benchmarking.

http://video.ucar.edu/mms/image/samsi2014_peter_thorne2.mp4 - I provided an overview of uncertainty quantification in climate datasets.

http://video.ucar.edu/mms/image/samsi2014_colin_morice.mp4 - Colin Morice provided an overview of the HadCRUT4 uncertainty estimation techniques.

http://video.ucar.edu/mms/image/samsi2014_doug_nychka.mp4 - Doug Nychka provided an overview of spatial statistical aspects.

http://video.ucar.edu/mms/image/samsi2014_jeff_whitaker.mp4 - Jeff Whitaker provided an overview of comparisons between surface temperature products and dynamical reanalyses driven solely by observed SSTs and surface pressure measurements.

http://video.ucar.edu/mms/image/samsi2014_enric_aguilar2.mp4 - Enric Aguilar gave a talk on problems in some typical data sparse non-N. American / European series.

http://video.ucar.edu/mms/image/samsi2014_finn_lindgren.mp4 - Finn Lindgren provided a talk on spatial statistical aspects.