24 replies to this topic

[Roger J Smith]

I have the optimistic feeling that we are only a few years away now from a situation where long-range forecasting becomes perceptually more reliable and starts to "back into" the accuracy range now staked out by the conventional models at day 4-6. In other words, I'm saying that in the rather noisy and cluttered field that is long-range forecasting, there probably already exist some approaches, some workers, some recent track records, that would encourage one to say "look there instead of at FI" for guidance in the time frame beyond seven days.

Ian, in my humble opinion, has fallen victim to a situation where a plausible anchoring for long-range forecasts in one particular region (western Europe) has dissolved faster than most had expected. That plausible anchoring was the long-term height anomaly and temperature regime witnessed in broad general terms from 1988 to (April) 2007 that seemed to thwart even the efforts of a modest block in winter 2005-06 to deliver anything more than feeble wintry synoptics. Given the combination of this persistence ("the trend is your friend," as they say) and the claimed results of global warming studies, it probably affected everyone's range of expectation to some extent, and there's no saying that this background warming is not real and affecting even this winter's rather robust blocking and cold anomalies.

More reliable long-range forecasts do seem to be coming along, I think people on Net-weather would generally agree that if you take the whole range of them available in past years, there has been an improving trend. I know there has been for Fred and myself -- in fact, I should mention that Fred was overly kind perhaps in his remarks, I feel that if he had gone solo this winter he might have appeared to have nailed this cold winter regime more so than our combined effort which tended to be filtered through my growing conservatism in such matters (just because I didn't totally trust that the signals for this 1962-63 style blocking could be sustained over the warmer North Atlantic without interruptions of the sort we saw last January).

I would imagine that there are going to be very mild winters in the near future, and this whole process of returning to a paradigm of a full possible range of outcomes may once again fall out of favour once the quiet sun episode fades from experience and memory (if it does). So, hopefully, these advances in LRF will lock in at some point and more credible agencies than isolated amateurs or mets dabbling in LRF in their spare time will be making, issuing and taking credit for the forecasts. That way, it won't be just an academic exercise that has no win situation, just a variety of lose and draw situations, all inevitably terminable by human frailty of some kind or another.

It's only a discovery if other people say so, until when, it's a personal obsession.

I'm basing all this in part on UK + Ireland, and in part on U.S. experience, as I see many different approaches being used, and some encouraging track records (anyone can fluke one LRF by chance, a track record is necessary for credibility). The challenge is going to be for somebody (and I'm on the case, hopefully not alone) to investigate all the promising techniques and try to get a handle on why they work independently or if they are perhaps different ways of looking at the same problem. This sort of approach requires that one be resistant to, although not totally opposed to, clamours for enshrining global warming as a huge factor in all forecasting. A more successful approach is going to require maintenance of balanced assessment of past climatic performance and not any tendency to throw out all records from before a certain date. At the same time, there is probably a new base line in existence, perhaps it is now descending back into a more historic mode, but different regions will respond differently to these recent changes, for example, North America seems not to have been affected very much if at all.

[The PIT]

Theres a few bust forecasts. Anyone remember the mild end to Dec and mild start to Jan.

Nothing wrong in getting it wrong even the pros get it wrong. As long as you admit it went tis up everything is fine.

GP is doing very well this winter and was forecasting a very mild Feb after a cold jan. Will it happen ???

[Pennine Ten Foot Drifts]

Roger J Smith, on 02 January 2010 - 21:40 , said:

It's only a discovery if other people say so, until when, it's a personal obsession.

I'm basing all this in part on UK + Ireland, and in part on U.S. experience, as I see many different approaches being used, and some encouraging track records (anyone can fluke one LRF by chance, a track record is necessary for credibility). The challenge is going to be for somebody (and I'm on the case, hopefully not alone) to investigate all the promising techniques and try to get a handle on why they work independently or if they are perhaps different ways of looking at the same problem. This sort of approach requires that one be resistant to, although not totally opposed to, clamours for enshrining global warming as a huge factor in all forecasting. A more successful approach is going to require maintenance of balanced assessment of past climatic performance and not any tendency to throw out all records from before a certain date. At the same time, there is probably a new base line in existence, perhaps it is now descending back into a more historic mode, but different regions will respond differently to these recent changes, for example, North America seems not to have been affected very much if at all.

I don't want to clutter things up by repeating the whole thing, but I have to say that that RJS is a beautiful example of the kind of thoughtful, informative and fascinating posts that you have become noted for on Netweather, and is one of the reasons I stick with this forum as my number one port of call. Great analysis, excellently written, balanced and informative. 10/10 !!!

[BLAST FROM THE PAST]

The PIT, on 02 January 2010 - 22:17 , said:

Theres a few bust forecasts. Anyone remember the mild end to Dec and mild start to Jan.

Nothing wrong in getting it wrong even the pros get it wrong. As long as you admit it went tis up everything is fine.

GP is doing very well this winter and was forecasting a very mild Feb after a cold jan. Will it happen ???

Very very well PIT, its on our LRF thread. If you spent Xmas down here...yep it was fairly mild, and yes the cold hung on to the north more than mentioned. Thank you though for your imput and nice that many parts have stayed cold [for cold lovers].
By the way, GP MAY well update his Feb thoughts by mid Jan...just a little heads up.

BFTP

Edited by BLAST FROM THE PAST, 05 January 2010 - 12:05 .

[LomondSnowstorm]

Hey, most of the LRFs on here are bust, particularly mine, which went for a quite period at the end of December, a December CET of 5.5C, and a very mild spell to start January!! Nowhere near right at all! Long range forecasting in such detail takes real guts, and having read over many of your previous seasonal forecasts I think you are one of the most accurate long range forecasters out there.
It has been a surprising winter for all of us though, even those who went for cold didn't think this cold this early!
LS