Crocodile23

A huge leap from -2.75 to -1.89!! That is completely dependent, and the reason this has happened, on the GFS 12z versus 06z forecasts of the next 9 days. So simple!

Not in full details, not even in small details. Only the original link with the complete paper of 28 pages that gave free yesterday for some time contains the whole method. I've managed to grab it in time but unfortunately i guess it's not legal to provide it as the authors don't provide it. I guess when it gets publiced it will be in English so it will get known and examined by experts and non experts much better....

I'm surprised that weather forums aren't flooded with discussions about the new OCTOBER PATTERN INDEX. Perhaps some of you don't know what this even is. And myself just learnt about it before 2 days. So here is some read(unfortunately it is in Italian so you will have to use a translator whatever that means): http://www.meteogiuliacci.it/articoli/october-pattern-index-opi-un-nuovo-indice-altamente-predittivo-per-la-stagione-invernale.html http://forum.meteonetwork.it/meteorologia/150667-per-i-curiosi-disponibile-paper-opi.html http://www.centrometeotoscana.it/forum/index.php?topic=7356.0 http://www.centrometeotoscana.it/forum/index.php?topic=7364.msg http://forum.meteonetwork.it/meteorologia/150683-opi-october-pattern-index.html The original link with the complete researce paper(which is 28 pages and it is also in Italian) was available for downloading but now it isn't. The reason if i understood correctly from the crappy google translation, is that the 3 authors will publish their research to a scientific journal and in order to do so, a free available copy is forbidden to exist. The deal is simple. But its results(if correct) ARE AMAZING! EXTRAORDINARY! It is based on a research job of Cohen with his snow advance index(SOI) where he managed to create this index to correlate it with AO(Arctic Oscillation) with very good results. But the October Pattern Index(OPI) goes many steps further and manages to predict the winter AO(allegedly, by all the data the authors give in their paper and anyone can repeat and see by themselves) since it is an index we obtain fully at the end of October and then we can predict the next winter's AO with an amazing accuracy! They talk about 90% accuracy for the years(if i recall correctly) 1979 to 2012. And i repeat, that we can verify these results by applying the method the authors give in their paper(that is now not available, but some people had managed to get it in time) to search for possible mistakes. Note that the authors are not a joke, they are scientists and hobbyists at weather/meteorology, the one is astrophysicist for example. So any thoughts? I find the result highly annoying! Because in a nonlinear chaotic system like the weather we should not be able to predict the next 4 months of October, or at least such a critical index of the next 4 months(like Arctic Oscillation), by just doing some "statistical investigations" on October. We shouldn't be able to do that! Furthermore there is the logical flaw i find with this amazingly high correlation(between the October OPI and the Winter AO), that the solar activity on the winter has only a minor, if at all, role on how the winter will go. If find it amazingly hard to believe that just an index, a statistical index which is rather simple to find or conceive the idea, by knowing its value at the end of October, is able to predict the whole winter circulation of the atmosphere, to be able to predict how the massive amount of different things that drive the climate like solar activity, QBO, AO, NAO, ENSO, SSTs, AMO, etc, etc, will combine with each other. This is utter nonsense if you ask me. It can't be done! What the authors say with all these(if i understood the google translations good enough), is that the OPI at the end of October gives a distinct special stamp of the state of the atmosphere and its OPI will lead to a specific AO value/winter state. And this is ridiculous to say since it assumes an EXTREME OVER-OVERSIMPLIFICATION of the way the climate works. But the problem is the data supports them! So please help....

I don't get it since: •Which are the 2 low pressure systems you are talking about? •It doesn't matter if there are 2 or 3 or 400 low pressure systems somewhere for determining what the wind will be. The only thing we care and the only thing it matters to derive the direction of the wind is to see(for example in a pressure diagram of 1 mb contours spacing), in the point we are interested in finding the wind direction, where the 2 adjacent isobars of pressure are! And in the example i gave the isobars are the 1011 and 1012 mbar lines. So the wind will start to blow from 1012 to 1011 and after a while with Coriolis force and friction it will stabilize to a ESE direction, the opposite of what GFS wind has! I guess it's a mistake of the pressure of GFS data but i can't be sure.

Hi, i know it's a bit off topic but this is somewhat serious about GFS output: I have noticed a HUGE incosistency in the pressure versus winds charts of the GFS as shown by the www.weatheronline.co.uk site. In the following map(a part of Greece which is on the northern hemisphere) it is the SURFACE wind as it is predicted by the GFS model. Look at the area noted by the red circle that i found the incosistency(among the other areas also that also have this incosistency): (click on the image to zoom in since forum makes the images appear smaller than their actual size) The incosistency occurs looking at the corresponding SURFACE pressure diagram for this area: Where we see that the predicted wind(from this surface pressure profile), is not the WSW(260° approximately) wind that the previous wind map had, but should be a completely opposite direction ESE(100°). And this is logical since wind blows from high pressure area to lower and with coriolis turning it right it would have the SE direction and with friction "turning it slighly to the lower pressure" it would stabilize to ESE direction. Or more easily with Buys Ballot's law it should go something like this: So what is going on???? Has GFS screwed things up? If you try zyGrib with the GFS data you get the same result so it's not Weatheronline.co.uk to blame. Both 2 maps(wind and pressure) are from the 01-08-2013, 06Z GFS output and correspond to 12 UTC time for today. You can see both images here: http://www.weatheronline.co.uk/cgi-bin/expertcharts?LANG=en&MENU=0000000000&CONT=grgr&MODELL=gfs&MODELLTYP=1&BASE=-&VAR=pslv&HH=6&ZOOM=0&ARCHIV=0&RES=0&PANEL=0&LOOP=0&INFO=0&WMO= http://www.weatheronline.co.uk/cgi-bin/expertcharts?LANG=en&MENU=0000000000&CONT=grgr&MODELL=gfs&MODELLTYP=1&BASE=-&VAR=uv10&HH=6&ZOOM=0&ARCHIV=0&RES=0&PANEL=0&LOOP=0&INFO=0&WMO=

QC?

And what about this? http://www.3bmeteo.c...valide----55900 (It's in Italian.) Any official news about this? From NOAA for example that refers to. Here is a super crappy translation from GOOGLE: The GFS model maps from Elaborate and ECMWF are not valid perched vitiated by error of data initialization. May seem like a joke, my hand is pure truth, so much so that NOAA Reed announced yesterday, through an official note, the disadvantage occurs. Blame some of oceanographic buoys position in the Atlantic, that failure would have provided incorrect data about the current state of the atmosphere. Well everyone knows how apologetic develop a Weather Search more simular it correctly a complex system How long can the atmospheric-land, Should Think So what if I can get is to have many approximations we introduce a cast of a reed error such as wrong initial conditions . In other words we start with a state that is already wrong, by making a forecast that over time can only amplify this error. The validity notice issued yesterday for all partners in the weeks and could explain the blatant mistakes made in the final period by two models, especially in the medium and long ends, resulting in the often unreliable and ready to change it to run in a run. We expect all if it systems, refer Went Appropriate checks next week, with the hope that not only has to settle my model Let nature take it's reed due course, in winter, giving us worthy of his name.

Just because the word bias is included in the page, it doesn't mean it's the same bias you were talking about in the first place.

I have one question about GFS main run data: I saw today the following: Having downloaded the data for a location for the 0-192 hours dataset of GFS and for the 192-384 hours dataset of it(it has them as different downloads) for the 18Z run the following data occurred: 1st dataset: DATA INITIAL TIME: 13 FEB 2011 18Z& CALCULATION STARTED AT: 13 FEB 2011 18Z& HOURS OF CALCULATION: 192 & Hour MSL-PRESSURE/hPa TEMP-850 hPa/°C DEW POINT-2m/°C TEMP-500 hPa/°C THICKNESS-500 hPa/DM HEIGHT-500 hPa/DM PRECIPITATION/mm +0. 1015.2 7.9 6.2 1.3 540.9 553.9 0.00 +3. 1015.7 8.2 6.5 1.2 540.5 553.7 0.00 ........................................................................................................................................................ ........................................................................................................................................................ ........................................................................................................................................................ +189. 1015.4 5.7 3.8 -1.3 536.1 549.1 0.32 +192. 1017.4 3.7 1.9 -1.4 535.9 549.8 0.00 2nd dataset: DATA INITIAL TIME: 13 FEB 2011 18Z& CALCULATION STARTED AT: 21 FEB 2011 18Z& HOURS OF CALCULATION: 192 & Hour MSL-PRESSURE/hPa TEMP-850 hPa/°C DEW POINT-2m/°C TEMP-500 hPa/°C THICKNESS-500 hPa/DM HEIGHT-500 hPa/DM PRECIPITATION/mm +0. 1017.4 4.3 2.3 -2.1 535.5 549.9 0.68 + 12. 1019.9 0.7 -0.7 -3.6 533.0 549.2 2.41 + 24. 1021.9 3.2 1.8 -3.2 534.8 552.8 0.37 ................................................................................................................................................ ................................................................................................................................................ Obviously the last date and time of the first dataset, is the same with the first date and time of the last dataset and it is 21 February 18Z. And yet the had different data! Precipitation, temperatures etc are all different. Why is that? I thought every date on each run(example 18Z like this), gets the data of the predecessor's prediction, i.e the +3 hr takes the data from the initial 0 hr data, the +6 hr takes the data from the +3 hr prediction etc. But it seems this is not currently occurs as the +192 hr data, of the first chunk of data(1st dataset) that NCEP provides for GFS is different from the +0 hr of the second chunk of data(2nd dataset) of NCEP's GFS, which corresponds to the exact same date and time. So from where the +192 hr takes its data on the second dataset?:unsure: What do i'm missing here?