forecaster

Likewise. The October 13th pattern stuck out as a potentially snowy synoptic if it happened in winter. Now this EC run, beyond reliable range, would be occurring in late autumn, and it would perhaps be the wrong side of marginal. Nonetheless, were it to occur it would be a proper "forecasting nightmare" with the fine detail (only reliable at short range) determining who would get snow and who would be in very cold rain!

The last few frames of that ECMWF would be cyclonic and wet, not dry and frosty and foggy. It looks like an October 13th repeat pattern, but 5-7C colder. Could potentially be hideous for low lying areas!

The ECMWF det looks strange, would not say "wrong" obviously, but the behaviour of the 500hPa low to the south is very "specific" and has ramifications on where the high goes. I would not bank on the low behaving in that precise manner nearer to the time.

With the 500hPa low cut off to the south, the high is not going to sink.

Easterly is still zonal I agree it is quite astonishing to see that, despite being at unreliable range. People seem to like printing various arbitrary perturbations, but the control run should be one of the most skillful members of the ensemble, hence pretty interesting to see this outcome!

Perhaps most important will be tomorrow's EC32, to see if it continues the suggestion of positive heights to our north or northwest for the latter part of the month and early December.

The post-"modernwinter" era?

I prefer to use the normalised SD shown on the ECMWF website. Although the spread is high, the context provided by the normalised product shows that it doesn't stand out as being seriously unusually high for that part of the world in recent weeks. Flicking through the runs up to the end of next week, I'd argue that the most interesting spread appears to our south, associated with the possible troughing forming over Europe, which is relevant to how the high might eventually orient itself if it does flatten somewhat.

I don't think reports, or lack of, from a few scattered people is important. It was a very warm Australian winter (for max temps, second warmest on record), a warm autumn, not to mention:

Maybe this would have been one? http://www.wunderground.com/history/airport/EGAA/2011/11/11/MonthlyHistory.html Only three days in the entire month with maximum temperature of less than 10C.

The joke is that Queenslanders for some reason worry about the impact that daylight savings could have on their curtains There are some tropical countries in the SW Pacific that use daylight savings. I think Fiji might be one of them, can't remember. It does seem a bit strange, but there you go!

I imagined the SAI as a change with respect to time....so would you approximate it by taking the (d/dt) gradient of the snow cover extent from Rutgers or elsewhere?

Yeah, this is not entirely right. Turbulence forecasts have been issued for a long time. Along with turbulence SIGMETS (a warning of significant weather for aviation). Although of course forecasting of it is really tricky. You can use pattern recognition to identify areas which are known to be conducive for severe turbulence (vertical wind shear being one good example). Tricky thing about turb is that it's so deeply non-linear; turbulence happens at microscopic levels and can also work its way up to levels large enough to impact aviation. The microscopic stuff must be parameterised, it can't be forecast explicitly, so this introduces assumptions from an early stage. However, you can still use global models for it - with the Ellrod Index (a combination, IIRC of vertical and horizontal wind shears) being pretty common. That's for synoptic turbulence - low level turbulence (mechanical, convective) is different again. And then there's mountain waves....

What annoys me about places like iceagenow is that they just run around the internet desperately grabbing anything that mentions cold weather, with no sense of context or climatology, and of course they completely ignore any warmer weather happening. Recent example would be a post about potential record overnight minima occurring in November for Canberra. The contact said "and Australia is supposed to be warming up!". The implication being that this event is somehow evidence that it's not!It's beyond belief that they can smugly toot out such nonsense. After such an absurdly hot year throughout Australia in general, a few cold nights in early November is meaningless!

I thought that the index itself is proprietary (of AER) and hence is not published, but is inferred.

I noticed today that the newest EC seasonal run has done a fairly dramatic flip in ditching the signal for northern blocking over winter, and going for higher than normal heights across the temperature Atlantic, just to our south, and lower than normal heights to our north. Now looks very similar to the MetO seasonal forecast from a few weeks ago.

The newest SST-CA model run is in. If nothing else, it's consistent with its previous run:

50% of the members represented by the mid-Atlantic ridge outcome in the GEFS: http://www.weatheronline.co.uk/cgi-bin/expertcharts?LANG=en&MENU=0000000000&CONT=euro&MODELL=gefs&MODELLTYP=2&BASE=-&VAR=cslp&HH=348&ZOOM=0&ARCHIV=0&RES=0&WMO=&PERIOD=

I don't think there is anything too exciting in the sub-surface. Those anomalies should become apparent nearer the surface as time goes by, so perhaps we'll end up with a pretty vague SST pattern in the end....

I see what you mean. I found lots of broken links there. Even if you google "NASA seasonal forecast" the first link that comes up gives an error. Try this http://gmao.gsfc.nasa.gov/cgi-bin/products/climateforecasts/GEOS5/index_atmos.cgi This is odd. I clicked it and got an error. Hit refresh and it worked. http://gmao.gsfc.nasa.gov/cgi-bin/products/climateforecasts/GEOS5/index_atmos.cgi

Appears that way. It is difficult to demonstrate skill over the mid-latitudes, that's for sure! You can see the H500 anomalies here, very bland almost everywhere: http://gmao.gsfc.nasa.gov/cgi-bin/products/climateforecasts/GEOS5/atmos_monthly_masked.cgi?var1=oct&var2=2013&var3=glb&var4=h500&var5=sea&var6=October Monthlies are here: http://gmao.gsfc.nasa.gov/cgi-bin/products/climateforecasts/GEOS5/atmos_monthly_masked.cgi?var1=oct&var2=2013&var3=glb&var4=h500&var5=mon&var6=October Some signal for a blocked February, but with the masking applied this does look less-so. Interesting stuff anyway.

I'd like to know how good the model is. Unfortunately it seems Joe B favours posting cold outcomes. OK it looks like he posted the raw output. If you look at the one with the masking (i.e. removed anomalies where the expected skill is too low) it looks much more bland:

Yes, I currently have mixed thoughts regarding the OPI. On the one hand, the glimpses at the theory and approach we've been afforded do not seem completely bananas. On the other, we have to be very careful with anything in science that cannot be replicated (for whatever reason). In the meteorological sense, it would surprises me that some patterns set up in October could dominate the winter so. Does the atmosphere really become *that* heavily preconditioned? What about November, which is typically a rapid cooling month?So at the moment I am watching with interest, but prepared to be disappointed!

The best runs to look at for a "quick glimpse" of the CFS is surely the 10 day chunks, which have been pretty consistent for weeks in showing more mild areas than cold areas: I like BFTV's interesting approach to the CFS though.

I don't agree. Comparing the deterministic with the ensemble mean should be part of the approach. See the (not-unquestionable) Bible here: http://www.ecmwf.int/products/forecasts/guide/user_guide.pdf