Fuck, it's cold.

[paolo]

Slightly off-topic I think, but here's a climate change question for yous...

This year the Met Office has predicted a long hot summer and a mild winter (for non-UK residents, these were both way off). Why should I believe them when they say what the weather will be like in 50 years? Is it because these are just short term irregularities but in the long term the temperature is going up and up because of increased C02 and what have you in the atmosphere?

Padraig - Damn, I thought my school was bad :slanted:

 

[scottdisco]

Is it because these are just short term irregularities but in the long term the temperature is going up and up because of increased C02 and what have you in the atmosphere?

correct.

 

[sufi]

taking us further off topic...

normally in uk summer is a bit dry, right?
so where does all the water go??? does it vacation down to the southern hemisphere or something????

(sufi - expelled from geography dept well pre o'level)

 

[Mr. Tea]

normally in uk summer is a bit dry, right?

Um, you do live in Britain, right?

 

[jenks]

Not sure if this is OT or not but...

they just closed my school at lunchtime - kids hooting and howling all the way home.

Their smiles were as nothing compared to staff's beaming grins.

 

[Mr. Tea]

 

[sufi]

Um, you do live in Britain, right?

but where does the water go?

 

[Mr. Tea]

I guess it just rains more somewhere else. I mean, our summer is Australia's winter, right? Think someone mentioned this already. (edit: it was you! there you go then)

 

[mixed_biscuits]

This year the Met Office has predicted a long hot summer and a mild winter (for non-UK residents, these were both way off). Why should I believe them when they say what the weather will be like in 50 years? Is it because these are just short term irregularities but in the long term the temperature is going up and up because of increased C02 and what have you in the atmosphere?

The test of a scientific model is its predictive power - unlike mid-term forecasting, the test for the predictive capabilities of a long-term forecasting model can be deferred indefinitely, just like the final whistle in a Man Utd home game.

 

[Mr. Tea]

In contrast to whoever said they'd heard a forecast of a mild winter, I remember hearing in the summer or autumn of 2008 that some climate scientists had predicted two consecutive cold winters (in Britain/NW Europe, I mean), and after than an open-ended rise in average temperatures. Prediction looks pretty good so far from here...

 

[mixed_biscuits]

There may be a logical problem with buttressing the general predictive power of climate models per se with the post-hoc choice of one successful instance.

After all, my rudimentary climate prediction device - a coin labelled 'winter colder' on one side, 'winter warmer' on the other would have performed better than most climate models, giving a 25% chance of two consecutive colder winters.

 

[scottdisco]

you should patent that, there's a mint to be made from the guys at Accuweather dot com, and Christopher Booker would be all over it

 

[zhao]

fucking minus 15 or some shit during the day here for the next 2 weeks. fuck.

 

[scottdisco]

of precisely zero consolation to Zhao, but the following was interesting

The current big chill is a result of high pressure over the polar region, which has pushed cold air out of the Arctic towards much of northern Europe, parts of Asia and the US. Winds from the north and north east, rather than the south and south west, have brought freezing temperatures to the UK...However, while parts of the world suffer freezing temperatures, the seesaw patterns mean other areas are warmer than usual, including Alaska, northern Canada and the Mediterranean.

Airport chaos as icy weather grip northern Europe

 

[mixed_biscuits]

you should patent that, there's a mint to be made from the guys at Accuweather dot com, and Christopher Booker would be all over it

I'm no sell-out, my model is open source, so that everyone can see exactly how the predictions (for which I give exact probabilities) are made - just like all the professional climate models erm aren't.

 

[scottdisco]

I'm no sell-out, my model is open source, so that everyone can see exactly how the predictions (for which I give exact probabilities) are made - just like all the professional climate models erm aren't.

that's excellent, kudos.

i have something similar going on wrt the Israeli/Palestinian conflict and likelihood of peaceful compromise in the future along two state model lines, i don't see why all these mediators, diplomats, think tanks, etc, waste their time on complex stuff when my coin trick is just a phone call away

 

[mixed_biscuits]

I think my original point was that just because a prediction is correct doesn't mean that anything meaningful brought about the correctness - a climate model would need to be consistently correct in a meaningful way to be useful (ie. Mr Tea's example would be useful if it was 'this climate model predicted correctly because it accounts for A, B and C, that the Met's model doesn't).

I guess that this close inspection by laymen is impossible, as good models' results are worth money to big business (this was why the Met was reluctant to release large amounts of data). Even bad models are of use, as they are either/both a) potentially good models or b) examples of what the good models aren't (and so may inform by negative instruction).

The problem with the long-term models is that one has to play a long-term waiting game to see whether they are actually working (with 'working' meaning predicting future real-world events rather than fitting past events) and they haven't been around long enough to have proved themselves. And even if one does prove itself, one has to assert that it didn't do so by chance (someone was bound to be correct) or by accident (it got the outcome right but the process wrong).

 

[zhao]

thanks scott...

However, while parts of the world suffer freezing temperatures, the seesaw patterns mean other areas are warmer than usual, including Alaska, northern Canada and the Mediterranean.

yeah i hear Brazil is full of bikini clad babes and in the full swing of constant massive beach parties right now...

 

[scottdisco]

I think my original point was that just because a prediction is correct doesn't mean that anything meaningful brought about the correctness - a climate model would need to be consistently correct in a meaningful way to be useful (ie. Mr Tea's example would be useful if it was 'this climate model predicted correctly because it accounts for A, B and C, that the Met's model doesn't).

I guess that this close inspection by laymen is impossible, as good models' results are worth money to big business (this was why the Met was reluctant to release large amounts of data). Even bad models are of use, as they are either/both a) potentially good models or b) examples of what the good models aren't (and so may inform by negative instruction).

The problem with the long-term models is that one has to play a long-term waiting game to see whether they are actually working (with 'working' meaning predicting future real-world events rather than fitting past events) and they haven't been around long enough to have proved themselves. And even if one does prove itself, one has to assert that it didn't do so by chance (someone was bound to be correct) or by accident (it got the outcome right but the process wrong).

good points all, and agreed. (close inspection by laymen would be pretty damn hard, i'd guess, because of the complexity of the models involved, also; i can't get too far into the EPA's climate change site before i'm hollering for a dictionary.)

i think you can forgive me my earlier playfulness, given i was really responding to some of your imperatives from the climate change thread (my bad, i just love the biscuits semantic parlour games)

on topic again (well, sort of), my favourite procrastination device of today (no doubt at least some of you will use this from time to time, but a new one on me)

go to www dot google dot com

and type in "temperature X" where X is a city of your choice - you'll get the result in funky icon style and ongoing forecasts if the place is big enough (or has enough stations nearby i assume, i can get the unremarkable Manchester suburb where i grew up, but not bigger suburbs right next door)

hours of, er, well, if not fun, then, procrastination :slanted:

hey, Zhao, it's 38 celsius in parts of Mexico right now, fancy a norteña party?!

 

[sufi]

There may be a logical problem with buttressing the general predictive power of climate models per se with the post-hoc choice of one successful instance.

After all, my rudimentary climate prediction device - a coin labelled 'winter colder' on one side, 'winter warmer' on the other would have performed better than most climate models, giving a 25% chance of two consecutive colder winters.

Yah,
i am 100% confident that i could do at least as good a job of weather predictions as the met office, especially since they seem to constantly hedge towards worst case scenario ("what if there's a hurricane and we fail to warn everyone?") i mean they struggle to come up with 3 days in the future

unfortunately however i lost a quid tuesday putting money where the mouth is, betting against snow that evening - but that doesnt disprove my weather forecast skilz, just i'm a crap gambler as any eejit could have got that right.....

now back to my question pls:

normally in uk summer is a bit dry, right?
so where does all the water go??? does it vacation down to the southern hemisphere or something????