honestlyreal

Icon

A flurry of #uksnow

Just before I get down to putting the lunch on, I glance at Twitter. @davebriggs says it’s just started snowing. A few seconds later, @stephenfry also reports a few flakes. @paul_clarke isn’t cool or dedicated enough to know Mr Fry’s precise whereabouts, but I’ve got a vague idea of Norfolk or north London, and I have a reasonable idea where Dave is.

This gets me thinking about a wave of “it’s starting to snow” tweets that I know are going to be appearing. What might that look like on a map? As a historical picture of advancing snow fronts, or even a live, dynamic view of where the flakes will land next. Powered by the people actually experiencing it. On a scale and level of detail unachievable by any formal organisation (that I know of).

So I thought about a hashtag: #uksnowstarts or something like that. Another glance at the mobile, and I see #uksnow. Ah, beaten to it – oh well – so what’s this about then? Turns out #uksnow is starting to bubble with reports from weather-obsessed Brits about whether they have (or indeed have not) seen floaty white stuff in the air.

Fresh from yesterday’s UK Government Web BarCamp, #ukgc09 (worth looking at ukgovweb.org), my mind was pretty open to data sets and what could be done with them. Except the contents of #uksnow wasn’t data – at least nothing that could be readily useable.

Time for a suggestion on standards. I’d put my own tweet out about weather conditions in RH1, so I quickly came up with a format: #uksnow [1st part of postcode] [n/10]; as a simple way of getting the holy trinity of mashable data together. Place, condition and time (from Twitter’s own logging). Bingo – three things that could readily be parsed by someone with the time and inclination to do so. And then analysed, mapped, mashed, shared, whatever…

That someone isn’t me, by the way. Well, perhaps using coding techniques popular in the late 80s, but all that fell behind me years ago. But the one thing that has been drummed into my head is that if the data sets are openly available, and there’s use to be had from them, and there’s some sort of standard to them, good things will follow. From someone, somewhere.

So now, having just put the spuds on, I’m trying to promote an instant standard for a crowdsourced data stream to track some current and impending weather events. Great.

I sensed a ‘perfect storm’ moment: this was a simple suggestion, with hopefully no privacy/security issues (through using a partial postcode). On a subject which Brits love to talk about. With news warnings that something juicy was on the way. And on a Sunday, with people hanging around Twitter and not much to do.

Stuttering take-up at first, but a few re-tweets of the standard got it moving. A question comes in on privacy. Another on whether n/10 was a good scale. I thought by now it probably wasn’t. Was it explicit that 0/10 would be a good way of saying “it’s not snowing” – in itself useful and valuable data for the exercise? Would anyone use 4,6,7 and 8 (questions familiar to survey geeks). Wouldn’t it be better to have a 0-5 scale with more fixed meanings like “it’s a few flecks” to “it’s sticking” to “white-out”?

Nice ideas all, but I stuck to my gut feel about a subjective 0-10 rating. After all, subjectivity is something we can all do – and thinking about the data averaging and errors that would be present anyway, surely swirling clouds of grayscale 0-11s washing across the country like a starling murmuration would tell the story quite adequately? Unused rating points could simply be merged or skipped completely in the analysis.

We’ll see – it’s Sunday evening now. I’ve been out doing family things and not tracking #uksnow. I’ve no idea if anyone’s taken up the mantle on a mashup. Perhaps there’ll be something there tomorrow as a data stream that could usefully do the job. Perhaps not. Perhaps everyone’s bored of it by now. Perhaps the mighty blizzards will start later and it’ll all go nuts again. Who knows?

I wonder if my message “without standards, it’s just noise” was transmitting onwards. Had anyone felt the urge to keep retweeting the format in my absence? It might have changed. Did I care? Not really, all part of the fun, and the learning.

Either way, this is the closest I’ve been to a Twitter meme, and it hasn’t half made me think. And given the day job, and trying to stimulate real innovation in public information, that’s probably quite a good thing. I mean, if this did go anywhere, I can think of quite a few players in my working world who would take it Very Seriously Indeed. And that is a hint, by the way ;-)

UPDATE: 16 December

The story of how the map was created unfolds in the comments below this post.

UPDATE: 18 December

If you’ve come to this page via the TechCrunch link, there’s a more up-to-date take on the “IP” discussion for #uksnow here.

Category: Other

Tagged: , , , , , ,

27 Responses

  1. peeebeee says:

    It was great to see it taking off – my brief thoughts at http://peeebeee.wordpress.com/2009/02/01/crowdsourced-snow-reports-on-twitter/

  2. Mulqueeny says:

    Brilliant Mr Clarke, well good to watch even if you overanalysed a bit here!!!!… brief plea, please lose the white on black: 1. deeply inaccessible, and 2. does my head in (more importantly :))

    Am not normally this grumpy, ’tis Sunday and can see Monday

  3. megov says:

    Hey PC
    Now you’re a “Content Generator”.

    Well done!

  4. Mulqueeny says:

    G’Lawd… what a difference!! brilliant… (I am talking about the new blog look)

  5. paulclarke says:

    …and as of about 22.40 tonight, I spot a mashup, http://www.benmarsh.co.uk/snow/, exactly as the theory suggested would happen.

    Got an idea that will be of interest? Found a way to propagate it? Can you structure the data to some sort of standard? Is everything freely accessible and re-usable?

    Four ticks? Bingo: the power of information will make something interesting happen.

    Sustainable? Repeatable? Perhaps not – certainly not guaranteed. But a nice end-to-end case study for today.

    Now waiting for any fallout along the lines of derived data… those partial postcodes perhaps… ;-)

  6. paulclarke says:

    Well, I didn’t spot it myself, as such. H/T to @lexij for being networked enough to beat me to it!

  7. kirsty says:

    I was coming up with a map here… http://nocto.com/uksnow/ …but Ben is doing a much nicer job and I need to get some sleep! Really interesting data to play with though. I’ve enjoyed it.

  8. [...] he describes in his blogpost, Paul Clarke came up with a quick ‘n’ dirty format to log the progress of the snow [...]

  9. tombasham says:

    Hi Paul,
    Thanks for the comment on my post. Hope you’re not snowed in like I am this morning! :)

  10. [...] and savvy snow spotters like Paul Clarke are considering more ways to make crowdsourced data mashed, mapped and meaningful. If we can apply this technique to unfolding events and crisis situations instead of just the [...]

  11. Phil says:

    Hi Paul,

    Apologies, it didn’t realise my blogpost would leave that on your comments!

    I’m planning on having a bit of a brainstorm and coming up with other potential applications of this sort of distributed data capture, then I’ll pop them down on my blog.

    I’d appreciate your thoughts!

    http://phildonohue.wordpress.com/

  12. [...] Within minutes of posting this (and informing Twitter) it comes to my attention that Paul Clarke came up with the standard of #uksnow [postcode prefix] n/10 and started the meme off. As part of the UKGovWeb lot he’s [...]

  13. Gavin Wray says:

    Hi Paul,

    I felt the urge to keep tweeting the updates after you mentioned the standard structure. I wasn’t sure what was going to happen – and those following me on Twitter who hadn’t heard about this mashup must’ve been thinking I’ve turned into a robot. However, as a ‘web and data’ person I was really interested to see how crowd-sourcing a dataset would actually work in practice. Great idea to latch this onto weather, an issue that UK residents love to talk about.

  14. [...] Here’s the story about the snow map. SHARETHIS.addEntry({ title: "As snow hits the UK the Twitter mashups storm [...]

  15. [...] la zona devono solo inserire il codice postale) le parti del Regno Unito colpite da nevicata: qui racconta come gli è venuta questa [...]

  16. [...] la zona devono solo inserire il codice postale) le parti del Regno Unito colpite da nevicata: qui racconta come gli è venuta questa [...]

  17. It was exciting to be a part of this and watch it unfold. A fun proof of concept.

  18. [...] I was heading across the river towards Waterloo anyway that afternoon, and kept my eyes open for a child as described. Probably many others did, or gave it some thought. Crowdsourcing at its best – unorganised, viral, organic, with a unifying purpose, but nothing else by way of structure to get in the way… Remember #uksnow? [...]

  19. [...] A flurry of #uksnow [...]

  20. John P says:

    Good piece. Made me think of Snowblind and snowsight, which led happily to this:
    http://video.google.co.uk/videoplay?docid=6210815799745676252#

  21. Social comments and analytics for this post…

    This post was mentioned on Twitter by amcunningham: RT @paul_clarke: how the #uksnow reporting meme came about (for those who didn’t see in Feb) http://bit.ly/78ybEP

  22. [...] Paul Clarke, photographer and public sector consultant begs to differ. [...]

  23. [...] goodbyes at work, another successful government barcamp, and a fourteen hour journey through terrible snow storms to get here and be reunited with my [...]

  24. EthrDemon says:

    XKCD did a comic about this phenomenon with respect to earthquakes:
    http://xkcd.com/723/

  25. Paul says:

    ha! I remember that one well. Thank you :)

Leave a Reply

Flickr Photos

Dare_23Sept_070

Dare_23Sept_069

Dare_23Sept_068

Dare_23Sept_067

Dare_23Sept_066

Dare_23Sept_065

Dare_23Sept_064

Dare_23Sept_063

Dare_23Sept_062

Dare_23Sept_061

More Photos