I’ve just spent three days in Exeter at a Met Office initiated exercise on issues associated with constructing a new surface temperature working programme.

I’m never going to write it up in any detail, but for a flavour, here are my tweets using the #climateobs hashtag (plus two that add some value):

  • At the surface temperature workshop http://www.surfacetemperatures.org/ #climateobs. See also my delicious tags at http://is.gd/eYNw (Tue)
  • Steve Worley on ICOARDS: entire dataset 90GB, issues of data publication, citation, format, software & project sustainability #climateobs (Tue)
  • #climateobs Jay Lawrimore experience at NCDC (world data centre for climate) on gathering and providing access to historical & current data (Tue)
  • #climateobs … surprised to discover that daily precip and max/min temp are not required by WMO … and therefore difficult to obtain (Tue)
  • #climateobs Albert Klein Tank: eg why daily extremes to be recorded (and extracted historically) Moscow July 2010, 31 days > 25C, cf av 9.5 (Tue)
  • #climateobs Will some aims of this project founder on national met services restricting access to data to protect commercial exploitation? (Tue)
  • #climateobs Now getting short presentations on the white papers (at the website where we introduced the hashtag) … (Tue)
  • #climateobs As usual I’m struck with the antiquated approaches in the met/climate data community wrt data modelling and data curation. (Tue)
  • #climateobs … but fossils have their role, and at least they don’t change (much). Fossils are much much better than missing links. (Tue)
  • In breakout group on provenance and version control: too busy engaging to tweet.#climateobs (Tue)
  • metoffice Watch conference organiser present the aims and aspirations of the global temperature workshop at http://ow.ly/2AtMt #climateobs (Tue)
  • When you have a hammer everything is a nail. My hammer is RM-ODP http://bit.ly/azsSsY #climateobs could do with decomposing accordingly (Tue)
  • kevingashley @bnlawrence Are the antiquated approaches because they’ve been doing it longer (so practice ossifies) ?
  • @kevingashley Existing practice certainly a factor. But part is that key influencers don’t have the available attention span to look out (Tue)
  • #climateobs Tension between extensible data formats, and defined formats so that folks know what is wanted. This is such a solved problem! (Tue)
  • #climateobs Oh no, someone is seriously proposing redefining a new METAR format and constraining it to transmit via BUFR. BUFR? Dinosaurs! (Tue)
  • #climateobs So I commented from the floor about BUFR. Apparently 180 countries are in the process of moving to BUFR and fear more change. (Tue)
  • #climateobs … and I rest my case about dinosaurs. How in the 21st C did we get to a lossy binary format with external tables that change? (Tue)
  • #climateobs day 2: concentrating on existing and required efforts to homogenise data to produce consistent records (Wed)
  • #climateobs importance of metadata: some stations moved from city centres to airports around ww2 (or did they)? Impact on trend analysis! (Wed)
  • #climateobs Matt Menne: Climate Hippocratic Oath: Do not flag good data as bad. Do not make bias adjustments where none are warranted. (Wed)
  • #climateobs Scary: correlation is all statistics and not physics? Radiosonde analyses without assimilation? Network not really dense enough? (Wed)
  • #climateobs (I should say the last tweet is my take, no one else has said that.) (Wed)
  • #climateobs Interesting initiative, about to be rejuvenated: An Informed Guide to Climate Data Sets; http://bit.ly/9yZKyP (Wed)
  • #climateobs Need a distinction between scientific assessment criteria & testing strategies (including using held back and/or synthetic data) (Wed)
  • #climateobs day3 Christy: our science is now in court … that changes everything. Beyond provenance: rules of admissable evidence matter! (Thu)
  • #climateobs what worked for scientific uptake of datasets: update regularly, easy web access, clear explanations of changes (Thu)
  • #climateobs google.org relevance? Can they help? Only if they are in it for the long term - worrisome history (google science data) (Thu)
  • #climateobs Google tools good! But worry about aim to host data in cloud; what happens when they move on & institutional resources lost? (Thu)
  • #climtateobs Important standards: open provenance model, provenance markup language, iso19156 observations and measurements, netcdf/cf etc (Thu)
  • #climateobs Yet another community who need to use a lot of words rather than use the word metadata: far too ambiguous (http://is.gd/f2fWm) (Thu)
  • #climateobs Should data products be allowed “in” (whatever that means) that use data inputs that are not publicly available? (Thu)
  • CameronNeylon: @bnlawrence As long as license agreements don’t contaminate downstream products or access to other data I don’t see a problem #climateobs (Thu)
  • #climateobs Breakout groups back in plenary, my take: governance & ongoing impetus is going to be difficult (like all commons projects) (Thu)
  • #climateobs: Future ECMWF reanalyses will provide feedback timeseries compatible with the input station net on web in tbd databank format (Thu)
  • #climateobs Many institutes depend on data sales. If WMO res40 was renegotiated (how)? It would cause extra financial pressure at a bad time (Thu)
  • #climateobs And the room bifurcates: Should we use all the data or only the open data? Does traceability and openness trump correctness? (Thu)
  • #climateobs And now I have to leave Exeter … discussion continues, but you wont hear about it from me. (Thu)
  • #climateobs Postscript: Feedback about my comments re BUFR. More than a dozen folk came up over next two days to agree: BUFR is a disaster. (Fri)