Glacier Point

After an indifferent Sunday for GFS strat forecasts, a definate trend being retouted by the GFS today. I'm actually thinking we have two warming episodes going on here, one in the lower middle stratosphere centred over Canada and a second more focussed in the upper stratosphere. The lower / middle warming continues to be evident within 10 days and reaches peak insentity around the end of the Year. The second is manifested with a really strong uptick in temperatures in the upper stratosphere (check out 1 hpa values). This one doesn't get going until New Year. If the 12z GFS is to be believed, which was characterised by a succession of deep upper lows (not all going east to west !!), we have a series of further wave 1 episodes down the line, and all also that toxic combination of a Euro / Russian ridge - Aleutian trough around 24th December which argues for a wave 1 and wave 2 response around 13th January (as well as a possible earlier disturbance arising from tomorrow and impacting the upper vortex by 7th January). The timing of the mid January wave 1 / 2 disturbance of obvious interest if the MJO is heading into the western hemisphere at this time.

Again, another fascinating GFS forecast. Something we may be underplaying here is a strong warming trend in the middle and lower stratosphere, particularly over the Canadian Arctic. This is modelled at day 10 onwards with a strong ridge centred over Alaska and Canada. This would lead to a marked rise in temperatures modelled at day 10, especially so at 70 hPa. Note also the displaced vortex at 30 hPa undergoes another split. 30 hPa 70 hPa Around the end of the month we see a surge of warmer air at 10 hPa and 5hPa into the core of the vortex consistent with a surf zone intrusion and strong ridge over Alaska. So are we looking at a split vortex event or Canadian warming hybrid in the lower and middle stratosphere preceeding the major warming at 10hPa ? Tropospheric feedbacks will be interesting to monitor given these potential timing complexities. 30 hPa jump of temperatures by around 15C 25th December > blocking signature 3 weeks later on > 14th Jan, possibly sooner given lower stratospher being equally if not more agressive warming. 10 hPa jump forecast at 31st December > initial fast propogating wave response 3-8th January, then second response 20th January, more sustained end January onwards. The cumulative net impact of these warmings throughout the different layers is the conundrum. Will this bring forward the wave responses through faster downwelling -ve zonal wind anomalies ? If so, January looks like a lot of fun if you like a strongly negative AO.

Steve I disagree fundamentally with your assessment of the winter forecast to date that it is somehow way off of the mark on current anomalies and the somewhat inward and parochial approach to medium, long and seasonal range forecasting that is being put forward here. Here is the forecast MSLP anomaly for December and alongside it the observed H5 anomaly pattern and the H5 and MSLP composite analogue packages used so you can see where I was coming from. Forecast> Observed> Analog H5> (note the Scandinavian / Siberian anomaly depicted) Analog MSLP> Broadly, these look a very good fit across all of the key anomalies, with the one around southern Greenland being the least impressive, but the European context with troughing depicted very well. This is demonstrated in the surface temperature anomalies. Again, forecast and observed anomalies to date with the composite analogue for reference alongside. Forecast> Onserved> Analogue> That demonsrates a very good verification to date, especially in terms of the temperature distribution suggesting the core of the cold in Europe. Obviously, the data over the next 7-10 days is going to be not consistent with this forecast, although speculating ahead too far looks problematic given model variability. For interest, I've attached ECM day 1-5 and GFS days 6-10 and 11-15 H5 ensemble mean height anomalies to give an illustration of where we might end up. I would suggest on this basis, a retained +ve height anomaly across the northern latitudes and troughing over Europe, which essentially will verify the pressure forecast, much of which was put together in October. Both the Forecast and the Preview made great play on a negative phase of the AO and NAO. So far, and projected, we have seen both indices in a negative state, the AO definately more so than I would have anticipated. How useful are any of these indices for a relatively small surface area such as the UK ? Well, to answer this, it's worth understanding where we are coming from. There is a wealth of model data coming at us, both medium, long and seasonal range. In applying some degree of thought process, using an understanding of the dynamical processes in the atmosphere, we are attempting to spot a good run from a bad one. In the words of Ed Berry, this will give an edge on those using NWP alone on average. I must stress this on average as clearly there will be instances that I and others have been and will be wrong. This doesn't undermine the approach at all, far from it. It will be very easy to pick examples of where it didn't go right (last week's failed easterly for example). However, I suspect due to the complexity of the subject matter, rarely do we see examples of the numerous successes being cited. For example, the idea that increasing angular momentum picked up in late October would lead to a major vortex disruption (and strongly negative AO phase) in late November and December deserves special mention, as does the identification of precursor patterns and timings for the vortex disruption and upcoming warming event well in advance of any NWP. These continue to strengthen the belief that we are actually winning here. Both the mechanism and outcome were well forecast at a hemispheric scale. So how do we reconcile the small surface area of the UK and use of such predictive macro scale techniques ? The use of the AO and NAO strike me as a particulalry good example here. The AO and NAO can be negative and the UK might end up warmer than normal but on a European scale, they might be particularly good representations. Certainly from some of our commercial clients, especially those involved with the energy sectors, the European context is vital given the interconnectivity of the gas markets. Within long range forecasts, the use of AO / NAO forecasts is now becoming pretty much mainstream. Overall though, and taking say 10 rolls of the dice, a -NAO / -AO pattern will deliver cold for the UK. Probably eight or nine times out of ten. This is the key point underpinning all of the analysis and forecasts, a probablistic basis is premised on the various dynamical processes delivering by a loading of the dice in one particular favour. On a related note, last week's failed easterly was attributed wrongly IMO to a microscale feature cutting off the supply of cold air and preventing blocking from forming. The key aspect that was incorrectly modelled was the anomalous ridge that developed in the sub-tropics and removed any notion of a trough between the Azores and Iberia. In fact, the reverse happend. This ridge developed in response to about as macro scale feature as you can get. Anomalous westerly winds developed in two discrete Rossby wave trains emerging from the tropical Pacific and Atlantic. These are clearly visible in the OLR anomaly plot from 13/12/12. These had the net impact of pumping the sub-tropical ridge and making it impossible for any westward extension of the ridge to our east. Interestingly, I suspect that the sharp increase in solar activity four weeks ago was the route cause of this. High levels of solar activity are thought to increase the height of the tropopause thereby strengthening the Hadley and Walker circulations. The increasing strength and northward projection of the sub-tropical ridge are certainly consistent with this, as is the angular momentum anomaly for the tropics which suddenly dropped from postive to negative in this timeframe. This will be worth monitoring because solar activity has since plummeted and now stands at very low levels. Will the lagged impact of a reduction in solar activity work it's way through the system in the next 3-4 weeks ? This will be very interesting in the coming weeks because of the advertised warming of the stratosphere (see Strat thread) and also the emergence of an increase in +ve frictional torques: There was a dual peak in frictional torques in November which preceeded the sudden increase in tendency of relative angular momentum at the end of November. If the frictional torques begin to deliver increasing angular momentum, then we will effectively be building the foundations of the next cold pattern, at a distance of 3-4 weeks.

Period between Christmas and New Year holds a modest level of interest, around 10th Jan I think things begin to get interesting.

.. still around, and yes, patience a virtue, or at least allows for some cool, level headed analysis. The final half / third of the month was always in my book least likely to have a broader scale interest (tropospherically) as we enter a low angular momentum base state for the time being as the initial wave flux generated in late October and November ebbed away. I'll try and post up some time some analysis on last weeks shanaligans - and a possible link to solar activity spurt at the end of November. Interestingly, angular momentum likely to just nudge back up (but not excessively so) in the next few days which lends support for the ECM op type scenario of modest height rises centred around Iceland - so the potential for some colder incursions to the northern half of the UK as per UKMET extended and if we get enough ampliifcation some link up with the Russian High but odds definately favour a more cyclonic type pattern. The broad scale indices over the polar field remain negative, so over the longer term, we still have the dice slightly loaded into cold's favour overall.

18z GFS gives us a projected 85C rise in 10hPa temperatures over central Siberia from start to finish. Final frame looks to me to be point where the polar night jet is undergoing wavebreaking, meaning a rapid intrusion of the warm air anomalies as the jet suddenly deccelerates. Might just be worth keeping tabs on extended GFS ops as there may be a fast response wave induced down through the troposphere as well as the more traditional day +20 downwell. PS. 5hPa temps off the scale and a more impressive (if that's possible) near 100C rise in temps in this area from start to finish.

Finger print activity in the upper stratosphere suggesting that wave 1 activity is likely to increase in the next 20 days. GFS again picking up the idea of a strengthening stratospheric ridge extending into Alaska and Northern Canada, now at t348. This extends all the way up to 1hPa but not so discernible below 30hPa which does suggest wave 1, as does the total displacement of the pv. Still looking very possible for a major warming going into the New Year. GFS currently bounding around in outputs but last week of December forecast to be well above average at the North Pole, something akin to current values which would be 10-15C above normal. Still a few days before this hits ECM timeframe, outputs around 13th December should be of interest to see if we get some idea of cross model agreement although the causal effect (and pattern right now) looks to be sound in terms of timing and spatial modelling.

General thoughts as summarised in the other thread: http://forum.netweather.tv/topic/75105-model-thoughts-summaries-27th-nov-onward/page__st__60#entry2436405 Still a great deal of hemispheric uncertainty regarding the emptying of the pv from Canada (and also a rapid transfer of energy across the North Pacific). Supported again by 06z GFS. Don't think I will be going to Cheltenham Races on Friday and Saturday at this rate. An intriguing thought this,, if we get into a protracted cold block / Atlantic low scenario, the emptying of the PV behind the low starves the upper low of energy. Cold comes back, although not as severe as before because we are getting to the point where a top of 'fuel' is needed. That isn't due until early / mid jan so we are left running on the auxillary tank but they may well be big enough to keep us running below average for the foreseeable. And, another GFS run with a Canadian warming at its end. Definately starting work on the foundations of the next sever cold phase.

Ensemble mean projections look reasonably in line with the overall trend in the 10 day timeframe. The blocking centred north of the UK (not suprised that we see the main core of this centred around Iceland) forecast to remain until at least day 10. The shifting pv leaves enough residual energy to phase with a strong jet streak over the eastern US coast as the sub-tropical ridge strengthens over Bermuda. That brings us into slider territory and the potential for snow. Although note that the movement of the pv and how much residual energy is still yet to be resolved - which will have ramifications on how much energy is sent south-east, which will have an impact on the longevity of the deep cold, which will have established by this time. The GEFS mean temperature anomalies depict nicely where I think we stand in the next 2 weeks, which maybe understand the persistence of surface cold: Day 10 GEFS H5 anomaly also a good representation of where we stand at the edge of the mid range: Where we go after that ? Some waning of the cold pattern is inevitable but note the position of the jet modelled in any format well into the future - well south of the UK and low heights associated in the Atlantic and over Europe. That still probably translates to below average temperature for the UK and much of Europe (continuation week 2 GEFS). Proibaly a 10 day peridiocity to the cold waxing and waning. Longer term, the extended GFS continues to advertise a building stratospheric ridge into Alaska and Canada. That is the precursor to a major mid winter warming which still looks very much on for very late December. So December already running around 2C below average no prospect of any milder weather and the likelihood of further dropping temperatures and some snow. Happy with that.

Cool, Atlantic blocking structure at day -15/-16 leading to increased Wave 1. GFS reacting to this ? November meridional heat flux was also good in that respect: If so, that will be a sustained wave 1 attack in 15-20 days time given this structure is likely to be in place for a good few days - takes us to around end December for MMW displacement type.

Thing to note about both GEFS and ECM ensembles beyond day 10. The mean is in something of a no mans land. Both suites show a clustering heading back down in the extended range.

Too far out to answer that one but overall set-up at least brings that into play.

06z GEFS H5 anomaly has us under an Omega block at day 10 [cold & snowy] Interesting that we have a clear development of a wave 2 pattern across the NH (and look at the low heights right across the sub-tropical latitudes). These tend to retrogress with time and are important precursors to stratospheric forcing. It's a fair bet that December will advertise a very strong wave 2 pattern overall given what we've had in the last week, and what is programmed over the next 10 - 15 days. Trend from ensemble means is for blocking to remain and retrogress and week 1 temperatures around 4-5C below average, moderating a touch to a around 1-2C for week 2, with surface cold likely to be prevelant. Low pressure systems likely to pass close to the SW. High snow risk.

06z GFS .... best strat forecast of the winter, admittedly at the very extended range but it demonstrates the evolution of the cut off lobe which is now reliably being modelled, as is the shift of the main body towards Siberian sector. 70 hPa 30 hPa 10 hPa A very displaced vortex and we have a full on Canadian warming forecast there, which could very conceivably shift northward into the Pole and trigger a MMW. Based on this GFS output, MMW centred late December, tropospheric feedback mid jan with a possible earlier fast response similar to what we've seen in the last 2-3 weeks. This is the GFS we're talking about, and there has been a great deal of uncertainty in these outputs w/r/t the eastward development of the warm pool in the middle and lower stratosphere, but the building blocks being put in place around the middle timeframe are worth keeping tabs on from run to run because this could be highly signifiacnt in terms of weather ramifications in January. Judah Cohen must be pleased.

..very plausible Steve given overall direction of travel in GEFS and shifting pv towards Siberia (advertised yet again here). This may wax and wane but the potential for long term attacks and significant cold pool developing into Russia in the extended time frame cannot be ignored. If the next 10 days comes anywhere near as programmed, the zonal and meridional surface fluxes over the Atlantic will begin to feedback and lock down the -NAO / -AO for the winter.

As if NWP didn't have enough on its plate with the disturbance to the PV, MJO forecasts consolidating on a phase 2 projection. That will expose the GFS in particular (it has a large bias to tropical convection in the longer range) to greater pressure to turn the jet northward over Iberia. This will be a false bias given that the GWO is in phase 8/1 - and this is the superior measure. That lobe of the pv cut off over NE Siberia also rolling around like the preverbial loose cannon. GEFS H5 day 10 a reasonable depiction however (huge snow potential btw): Substantive ridge between Greenland / Iceland / Svalbard. Tonight's CPC 8-14 dayer will pretty much look like this. That will keep the jet well to the south and only when the ridge pulls fsr enough west will it be able to come up from the south. Always the risk however of shortwaves, but these likely to maintain our cold pool in situ by advecting the cold air westwards. Stick with your ensemble means!!

Just monitoring those GEFS means through t192, and the daily ops w/r/t 30 hPa heights, you do get the sense this a very dynamic situation. If the day 8/9/10 forecasts are near the mark, we will have gone full circle from where we were at the end of November. That's not long ago but feels like an eternity. I suspect many more rapid evolutions around this core theme of a -NAO, until at least we start to see the stratospheric vortex settle a bit, probably towards Siberian sector.

Eye popping -NAO developing on the 12z GEFS. If you're mithering about GEM, GFS and UKM, consider this trend developing for retrogression signal... yesterday t192 GEFS H5 anomaly >> today >> that contradicts the idea of a slightly stronger sub-tropical ridge extending northward and developing that flow in the Atlantic.

Some operational GFS runs have maintained this as a discrete feature cut off from the main body of the PV and taken it into the SW States. This has forced the retrogression further westwards. Adding to the last post, I suspect the background factors will sharpen the ridge development around Iceland and maintain this, meaining the retrogression (if and when) will have an axis around a quasi Greenland-Iceland-N.Sea block, possibly extending into Norway and dropping through Finland (and then hopefully SW though northern Europe).

To add to the positively cold vibe going on.... The trend for vortex shift towards Siberia gathers apace. This should assist in retrogressing the ridge over Scandinavia towards Greenland in the extended range. GEFS mean hight anomalies depicting an amplification of the North Pacific ridge which teleconnects to height rises over southern Greenland. The only complicating factor remains the modelling of a cut off section of the pv over NE Siberia and its movement across the North Pacific.

I was looking today around 7th December for a forecast anomaly pattern to fit a precursor wave 1, however day 7 looks a bit more plausible, particularly if the next +AAM spike fits that timescale. This is interesting as it teams up a possible wave 1 and wave 2 strike on the pv towards the end of the month.

Anthony Great to have your views here, and welcome ! Stewart

So stratospherically, we are running very similar to 1968. A like for like comparison of 1st October through 1st December of both years at mean flow 30 hPa. 1968 ----------------- 2012 Obviously this year we have seen a stronger westerly component at 70N, but the key here is to note the almost identical reduction in easterly zonal winds over the tropics over the time period, and the date of the 1st disruption (sudden easterly zonal wind) across the high latitudes, taking place late November and early December (the most recent not impressive on the chart just yet but due to delay in data will look more like 1968). Remembering that 1968 is potentially a very good analogue being a second cold season east QBO relative to the solar cycle. What might we learn from 1968 going forward? Plot below carries 1968 forward from 1st October through January 31st 1969. There were four subsequent disruptions which took place, roughly 2 weeks apart, the biggest occuring during January 1969. Each time there was an attempted rebuild of polar westerlies prior to the sudden reversal / weakening westerlies. So, perhaps in the coming weeks we should see a pattern of returning westerlies and reverals to easterlies or weakening westerlies. The easterly wind flow at 70N during mid January really catches my eye. I think that one is our bigee.

Iin the extended range we have a nice signature for mean jet flow displaced well to the south allowing stable blocking structures to take root. Spreads at day 14 show little variability on a flat trajectory around the Med. GEFS H5 mean height anomaly depicting a retrogressive signal: Note in the last one the anomaly over the Arctic showing good agreement on shifting of the core pv towards Siberian sector.

very tidy GEFS mean at day 8: Sub-tropical ridge held well in check over Bermuda, -ve height anomaly over the Azores and +ve height anomaly over much of the NE Atlantic. Solid stuff from the ensembles with not too much scatter.