johnholmes

how many have been -2 or below? Can anyone post the years please?

re 'covered all bases' or 'vague', just what on earth do you expect anyone writing a long range forecast to do? It is impossible at 15 days to do more than suggest the weather type so how on earth at 3 months can anyone do more than suggest a weather type? Indeed anyone who does start putting dates of severe frost and heavy snowfall is, best I don't use the word, but it is as much use as a chocolate fireguard. IF any of it turns out correct it is simply luck no more.

for skiers=winter predicted to arrive, at altitude say over 4000ft, during next week. Wengen/Lauberhorn forecast as an example below http://www.meteocentrale.ch/en/europe/switzerland/weather-wengen/details/S067281/ all based on GFS output of course.

summer continues! A low of 13.4C, patchy shallow Cu and sunshine with a very light SSW breeze

I think you are correct Steve, it shows just how much folk have learnt since then.

my summary of NOAA this evening and EC-GFS is below The most interesting development is the ridge and +ve heights in the Alaska border region. This was first tentatively shown on Wednesday, has been amplified and increased since and is now starting to show on the 8-14 outlook. No real idea what effect this may have but it must have some effect I would have thought. Will keep watching it. Links to 3 main outputs below. Note that NOAA does not mirror the ridge and +ve heights over northern Europe shown on both EC and GFS. This difference has only occurred today so for now I would discount EC-GFS in that area. http://www.cpc.noaa.gov/products/predictions/814day/500mb.php note that EC does show a similar idea re heights and ridging on the Alaskan border. NOAA make no comment about this feature, rather surprisingly? http://mp1.met.psu.edu/~fxg1/ECMWF_12z/hgtcomp.html Nothing in any of these charts over the past few days suggests that any major cold outbreak is likely in the 15-17 day time scale, beyond that I do not pretend to know.

I was out of my depth then, I think Steve M was also part of it, no doubt Paul will find the video when he gets time.

2009 showed -3.15 I thought?

Just missed a 'full' 20C at 19.8C but almost summer like at times into the early afternoon when it clouded over completely.

apologies all round-the old fella and his eyesight again, the 2012-13 was 4.8, 3.5, 3.2 NOT what I previously quoted. In fact this mistake was drawn to my attention yesterday, will make the necessary alterations. I have deleted my previous post to try and avoid confusion!

yes very mild again with a low of 14.0C, dry, cloudy with no wind

thanks sk, has anyone got the NAO values for those years, East or West based, or are they the same sign as the ENSO? Daft question perhaps so sorry about it.

you know more about that pitfall than anyone on here Jo other than Michael

currently 14.2 with Td=12.7, it has been much lower than this during the 'summer' months at this time of night.

If we look at day 6 for all models then all show a dip recently and the latest 6 days are missing UK Met, and others, so not sure why the models are missing http://www.emc.ncep.noaa.gov/gmb/STATS/html/new_acz6.html good to see and hope his bosses feel the same way!

as you say Jo, 'cheeky' but a possibility perhaps?

looking again at the data sk gave in reply to my original post this evening. We still are left with a mix of QBO and ENSO. So any link, regardless of a deeply -ve OPI is not clear. It is highly interesting though and who knows, as |'ve just commented, it may be another link in the jig saw of trying to predict with reasonable accuracy what the winter may turn out to be.

very true, a 'mild' winter by CET standards does not rule out a cold spell of 7-10 days, and more than once in the December to end of February. But it is the possible link with OPI and other possible links which is the interesting feature.Maybe OPI will be proven to another part of the jig saw. At the moment we do not know this and will not, irrespective of what the final OPI figure is this month or the subsequent winter values. There is still a lot of work to be done as the folk involved with the OPI openly acknowledge.

thanks for that sk, especially the ENSO and QBO info which is what I was asking for. Now all we need are the bits re the NAO, west or east for the years most close to any link?

ta-I'll check that out when I have time

A less than scientific look at this latest phenomena Is the OPI index reliable when comparing CET data with values for the OPI of lower than -1.7? This is a very un scientific take on the possible link between the OPI and ‘cold’ CET winters. For cold I have taken winter months with 2 or more below average months. Winter being taken as December, January and February. Temperatures in degrees C. The years of OPI below the figure quoted are:- 1976=-1.75 winter 1976/77 1978=-1.9 1978/79 1984=-1.8 1984/85 1985=-1.9 1985/86 2009=-3.15 2009/10 2012=-1.76, so -1.7 to the first decimal point. 2012/13 The CET data is that available on this site http://www.personal.dundee.ac.uk/~taharley/centralengav_temperat.htm And has the average based on 1971-2000 December=5.1, January and February both at 4.2; this gives a mean of 4.5C. 1976/77=2.0, 2.8, 5.2 1978/79=3.9,-0.4, 1.2 1984/85=5.2, 0.8, 2.1 1985/86=6.3, 3.5, -1.1 2009/10=3.1, 1.4, 2.8 2012/13=8.5, 5.7, 6.2 can be discarded as no month fell below average Before examining the above data we also need to see when the above criteria for cold winters CET values were satisfied but did not show an OPI of -1.7 or lower. The following winters fell into this category 1981/82=0.3, 2.6, 4.8 mean=2.6 1982/83=4.6, 6.7, 1.7 mean=4.3 so requirement not really met with the mean 1990/91=4.3, 3.3, 1.5 mean=3.0 1995/96=2.3, 4.3, 2.5 mean=3.0 1996/97=2.9, 2.5, 6.7 mean=4.8, so again the mean makes this winter not relevant in spite of 2 ‘cold’ months. 2010/11=-0.7, 3.7, 6.4 mean=3.1 And the OPI very positive=1992=1.75, winter =1992/93=3.6, 5.9, 4.6 mean= So tidying the above up leaves 1981/2, 1990/91, 1995/96, 2010/11 with CET fitting the criteria used, with one year 2012-13 when a very negative OPI did not lead to any month being below average in the following winter. A total of 3 years when the OPI did not come anywhere near -1.7, the actual values, in the same order as the CET years given above was:- -0.4,0.55, -0.65. We therefore have 5 years with the OPI and CET values agreeing with the theory of OPI indicating cold winters and 3 winters with cold CET values (as defined above) with the OPI giving no clue to this occurring. One year with a reasonably low OPI (-1.76 with no single cold winter month following. Putting the OPI in order of negativity with the winters alongside gives 1 2009=-3.15 2009/10=3.1, 1.4, 2.8 mean=2.43 3 2 1985=-1.9 1985/86=6.3, 3.5, -1.1 mean=2.90 2 3 1978=-1.9 1978/79=3.9,-0.4, 1.2 mean=1.57 1 4 1976=-1.75 1976/77=2.0, 2.8, 5.2 mean=3.33 4 If we compare these figures with those years when the OPI did not give an indication, we have 1981/82=0.3, 2.6, 4.8 mean=2.6; this would rank 2 in the above winters 1990/91=4.3, 3.3, 1.5 mean=3.0: this and 1995/96 would rank joint 4th 1995/96=2.3, 4.3, 2.5 mean=3.0 2010/11=-0.7, 3.7, 6.4 mean=3.1; even this one shows a mean winter temperature below the 4th in the table above. None of the above had all 3 winter months below average whilst 2 of the winters in the above table did so. Without doing any statistical tests on either of the above results it seems to me that the expectation, by some, on Net Weather, of this being the key to predicting winter are somewhat off the mark. As I suggested at the beginning it is a non scientific explanation and please feel free to shoot this post down if you wish. One other point is that the period of data is, relatively, small for predictive purposes. A minimum of 30 years is used for climatological records but for something of this type, an attempt to use a predictor much longer periods are the norm. Maybe someone could discover in the winters I have highlighted what the QBO etc showed as well as this might refine the results? Were the winters that supported the theory west or east based NAO. Equally what about those winters with below average months what was the QBO and NAO showing? I remain, sceptical of any claim that this will give a reliable indication ON ITS OWN, but it may be another indicator to use with others already in use. One thinks of the SAI and Tropospheric influences to name just two. Interesting to play around with the data though.

much the same at 6700ft at the station above Wengen, Change trains here for the Jungraujoch or to go down into the other valley to Grindelwald. The Eiger in the background http://www.bahnhof-scheidegg.ch/default.asp?itemid=107〈=e the local prediction is for it to change at height anyway come next week, but of course it only as accurate as GFS!! http://www.meteocentrale.ch/en/europe/switzerland/weather-lauberhorn/details/S067291/#trend

re comments about bias if mod/admin will allow, to clear this up I will e mail UK Met and ask; maybe someone might like to do the same with ECMWF and GFS?

Dry, overcast, no wind after some rain earlier, another double digit low=10.2C

As to how accurate then detail out to 5 days and decreasing detail from then on, much as you see from UK Met 1-3 days outlook to 5 days then general ideas of above/below average/rain/wind etc for the main areas of the country, out to day 30. Met now show a Fax chart out to 120h, when these first started, hand drawn, they were for 12 hours ahead then 24 and that lasted for maybe 8-10 years other than in house testing.