Data from the 20 UK CLIMAT meteorological stations

Posted: October 9, 2012 by tchannon in Analysis, data, Dataset, Surfacestation

A WMO resolution requires release of some climatic data in the form of CLIMAT data messages. For the UK there are 20 stations, few given the land area. Most are from aerodromes, all are synoptic stations which have already been briefly covered as part of the surfacestation project.

I’ve obtained and decoded the raw data into .csv files intended for spreadsheet import.

[Important update]

[update 9th Oct. I've now found what looks like the formal WMO CLIMATE document

Search on this http://www.wmo.int/gsearch/gresults_en.html?q=climat&submit.x=0&submit.y=0

Points to "Handbook on  CLIMAT and CLIMAT TEMP Reporting (2009 edition), WMO/TD-No. 1188"

/update]

Runs from at earliest 1996, to date. Is fundamentally monthly.

A first for the Talkshop, in a zip here.

Not everything is decoded but if anything useful is omitted, discuss. Eg. no wind data, monthly is of limited use. Decoded via a script which I had to write, pretty much automatic now.

I’ve included latitude and longitude complete with ready to use Bing and Google links for a web browser. (whether a spreadsheet will allow this, ah, could copy it)

Possibily of particular interest is inclusion of month min and max figures with date, unless it is tie, in which case the date is for the first instance.

I’ve barely looked at the data myself

One I have examined is Manston, Kent about which a part 2 post is mostly prepared. (have lots not published).
This station moved from human read min/max in 2000, apparently no bridging to AWS but a bit of mystery exists, may have been as many as four different sites until the present one.

There isn’t enough data to be clear on change min/max to AWS, at least I prefer obvious effects, not statistic, but it looks likely there is 4C under some conditions, I assume because the old site was poor. Got the location.

Very interestingly the hottest day was 10th August 2003, an infamous case of many records… but herein starts a tale.

Brogdale, also in Kent, 33km away is accepted as 3.9C hotter than Manston. I don’t buy this as sensible for competent sites. I still have reservations about Manston (eg. palleted instruments, a log fence, quite close to a massively wide runway) but it is reasonable.

I note the Met Office compare Brogdale with a wide area of stations but omit Manston, the only climatic station in the region. When I went down their list, oh dear, cherry picking poor sites. (might show some of the incidentals eventually)

As it happens I’ve spotted a new problem at Brogdale, perhaps Tim’s equivalent of the Anthony Watts “air conditioners are attracted to thermometers”, something else. You’ll have to wait.

Comments
  1. Lance Wallace says:

    Looking at the first few cases, it appears that Tmax and Tmin are provided to one decimal point only, and then Tavg is rounded up in all cases, leading to an upward bias. This is the same convention adopted by the US CRN. When I complained to them about this, they informed me that they are “working on” a different approach, which might bring them into line with the ordinary physics approach of rounding up or down depending on the next-to-last digit of the sum of Tmax and Tmin (e.g., up if odd, leave if even). Just a small error that would not affect the trend, but still seems sloppy. Who knows how long it might take to replace this convention.

  2. tchannon says:

    I’d not come across the formal specification for climat. Enough was available to figure it out, the most useful being a WMO presentation style brickbat aimed at authors of climat data, ie, indirect complaining many got it wrong.

    What some of the data actually means, pass, can only guess.

    I’ve now looked more carefully and found what seems to be the full document.

    Article is updated.

    I’m struggling with so many loose ends so it probably looks and is sloppy.

    Anyone adding commentary and insight, excellent.

    CLIMAT stations are all around the world but being able to get a rough look at the location is a whole different game since you will have to find the thing. Mileage will vary. For the US the surfacestations project might have all you need.

  3. [...] recent post talks about 20 UK CLIMAT stations, which are used to report to GHCN. It might be interesting to see where those stations are located [...]

  4. Trev says:

    FWIW, I live near Benson and the conditions in that area always seem to throw up colder temperatures.