Summer 2007 Outlook

[Glacier Point]

Summer 2007

This will be my fourth seasonal forecast and I won't try to pretend that any of the three previous ones have been close to the mark. What I have learnt through error is that long range forecasts for the UK and western Europe are essentially influenced by a surprisingly small number of variables with each one vying for dominance.

In order of importance, I would consider the following as being key to a winter or summer outlook:

1) the polar pressure anomaly, more particularly at the 30 hPa to 100 hPa level which is essentially a representation of the Arctic Oscillation (AO) and temperature of the polar stratosphere;

2) the climatic trend for warming;

3) the relative state of El Nino Southern Oscillation (ENSO);

4) Atlantic Sea Surface Temperature Anomalies (SSTAs).

This summer poses a particular dilemma as the most influential factor (polar pressure anomaly) is likely to be a non player whilst there are considerable uncertainties about the timing and strength of any La Nina.

Looking at the most recent polar temperature anomalies from surface to stratosphere, we see that there is no overall trend emerging for either warm or cold and, as a result, we see no particular height anomalies - especially at the critical 100 hPa - 30 hPa level where most of the influence for upper level zonal winds and broad weather patterns takes place.

Note how unusual this is in terms of the winter and spring which were marked by big anomalies (mostly -ve height anomalies across the polar region driving the AO rampantly +ve. Now with the final subsidence of the polar vortex and shift to an easterly phase Quasi Bienniel Oscillation (QBO) in April the weather patterns have returned to something a little more like a little rather than a lot above average.

All the indications are for this situation to continue. There are no warming or cooling events showing up in the upper atmosphere whilst the QBO can have an impact on polar pressure anomalies but does not according to seasonal correlations for the June July August period.

Key assumption (1): The upper level polar temperature and pressure anomalies will be minimal allowing other factors to influence the summer pattern.

Speaking of the QBO, note the correlation of the -ve or easterly phase QBO with (blue) height anomalies across from Bermuda to the Azores and Iberia - (remember this is a linear correlation map so a blue colour infers a reverse correlation). April's QBO showed a switch to the easterly phase with an index of -5.18. Current upper equatorial zonal wind anomalies depict a building of the easterlies indicative of a strengthening -ve QBO but recent westerly winds in the western pacific have hindered this making a summer with a weakly -ve QBO the call.

http://www.cdc.noaa.gov/Correlation/qbo.da...

The climatic trend for warming is really self evident. Over the last decade, the string of above average months is mightily impressive and statistically overwhelming. This doesn't preclude average or below, neither does it imply warming to an anomaly in excess of +2 as other factors such as the polar stratosphere have enhanced any warming dramatically (note how the excessive warming trend of March and April suddenly disappeared during May when the polar anomalies shifted). Bottom line is that the -warming trend is likely to add +0.25 to +0.5 to seasonal averages on a highly probable frequency - as well as shifting pressure patterns northwards.

Key assumption (2): The climatic trend for warming will more than likely favour an above average summer in terms of temperature but not of the magnitude +2 or +3 on its own.

Moving on to La Nina. Most of the models have picked up on the large body of below average sub-surface waters across large parts of the equatorial Pacific, particularly off the coast of Peru and within the ENSO region 3.4 (often regarded as the key to determining the strength of ENSO events).

Currently the Pacific SSTAs show neutral ENSO conditions with the hint of cold water anomalies at the surface beginning to gain strength based upon latest weekly SSTA analysis:

http://www.cdc.noaa.gov/map/images/sst/sst.anom.gif

With the developing easterly trade wind regime associated with the -ve QBO, these cold water anomalies at depth should be dragged to the surface or up-welled favouring a developing La Nina although the situation in the western Pacific remains unclear with westerly winds tending to suppress any cold water development at the surface.

The excellent IRI site summarises the position neatly regarding La Nina and some of the uncertainties about timing and strength of La Nina.

http://iri.columbia.edu/climate/ENSO/currentinfo/update.html

http://iri.columbia.edu/climate/ENSO/currentinfo/SST_table.html

Taking a view on La Nina is important as can impact quite considerably on our weather patterns although not as much as an El Nino event. The impact of the ocean-atmospheric coupling of a La Nina is not easy to identify in western Europe. This reanalysis for linear correlation for June-July-August for the Multivariant ENSO Index (MEI) shows no real impact of either ENSO state:

Many strong La Nina summers have, however, tended to feature mid Atlantic ridges and low pressure centred over western Europe during July and August.

Currently we are in ENSO neural conditions.

http://www.cdc.noaa.gov/people/klaus.wolter/MEI/

The majority of models suggest a steady decline in the Multivariant ENSO measure (MEI) indicating a slow but steady decline towards a weak to moderate La Nina event by the late summer or autumn. This seems very probable given the state of the easterly trade winds and sub-surface cold water anomalies.

However, the overall change in SSTAs during May compared to April was highly marginal and the difference between mid April and mid May's SSTA across the tropical Pacific tells the story nicely.

April ENSO:

May ENSO:

Difference:-

Considering the MEI for April was 0.088, this makes May's MEI also around this mark. The Southern Oscillation Index (another measure of ENSO) is also falling making La Nina less certain. An MEI of around 0 for May would make any of those models predicting a La Nina (MEI <0.5) for the Summer far from being on track.

Key assumption (3): ENSO neutral conditions will affect the early part of the summer with a decline towards weak La Nina by August - I.e. no major climatic impact from this factor.

This leaves the Atlantic SSTA as the main driver for the summer over the UK and western Europe.

In order to identify how the pressure patterns will be affected by the SSTAs, we need to establish the most likely base period and lag effect. There is much divergence on just how long it takes for a surface anomaly to influence the atmosphere (for the mid latitudes). Some research suggests one month, others more subtly over a longer period (seasonal SSTA lag). It makes sense to identify a trend if any and taking the period March - May gives an indication of both the shorter term and longer seasonal trend. If the trend and the more recent pattern are similar, then more confidence can be given.

Looking at the SSTA for the Atlantic during the period March - May we see the following:

- cold water anomalies up and down the western Atlantic;

- warm water anomalies in the central and eastern Atlantic;

- warm water anomalies the Davis Straits;

- developing warm water anomalies around the UK.

Taking the first three sets of anomalies, we can see 14 broad matches since 1952 which might help to predict how the summer pressure patterns might look if influenced by the SSTA alone. Removing years where there was an El Nino or strong La Nina which might distort the atmospheric response, we have a reasonable sample of 8 years.

This is the composite reanalysis of all those years compared to the March-May 2007 output:

Historic:

Current:

The reanalysis for May to date (when the polar anomalies faded) using these analogues provided a reasonably good explanation of the experienced pattern:

Analogues:

May 2007 pressure pattern:

Tellingly, this hints at the climatic changes over the last decade or so which suggests the high pressure belts to have migrated northwards making lower pressure further north over Greenland much more likely.

Rolling forward, the JJA outlook for 500 hPa pressure anomaly pattern based on the analogues looks like this - and I think we will not deviate too far from this type of pattern given one or two allowances for climatic trend:

Notably, there is a considerable inter-month variance with June and Augusts witnessing a similar pattern with July the month most likely to see a more unsettled picture.

June:

July:

August:

Analogues should however be only used as a guide. We also need to look at the current SSTA and the influence of the climatic tend. For example, taking the last three most recent analogues (and an early warming trend) suggests July and August to look like this…

July: (upper low shifted west)

August:

For what it's worth, frame this pressure anomaly pattern for August as the blueprint for the overall Summer.

In terms of interpreting the likely impact of the current SSTA, my first point of call for predicting summer patterns from SSTAs are cold (or neutral) water anomalies in the sub-tropical or tropical Atlantic. Time and time again we have seen these generate displaced high pressures to the north and east of the anomaly centre. This year we see two such anomalies - around Bermuda which has shown up in the March-May SSTA pattern and a more recent development in the area off the west African coast.

Sandwiched in between the two is a warm water anomaly around the Azores. Taken together the three should influence a substantial and complex if not rather flat sub-tropical high pressure belt with a +ve long wave height anomaly east of Bermuda extending across the Atlantic to the Azores and Iberia. Handily, this also conforms with the seasonal correlation for the QBO in the sub-tropical Atlantic.

For the UK and western Europe, the analogues suggest high pressure to be in control during June and August, and with the warm SSTAs, I think a repetitive high pressure block will be located over or close to the UK - perhaps just to the west during June. With a strong Azores High, this will necessitate some sort of upper level weakness around Biscay and mid Atlantic trough.

The analogues imply July to see something of a pattern change with low pressure in the near Atlantic. However, the current SSTAs also suggest high pressure to be in control but shifted east a little which could help draw some very warm air northwards. The climatic trend also backs this shift of the July upper low westwards.

Key assumption: The impact of Atlantic SSTAs will be for a strong sub-tropical high pressure belt with +ve anomalies east of Bermuda and between the Azores and Iberia, with a secondary ridge over NW Europe.

In terms of the summer forecast, the absence of any lead from polar pressure anomalies, no La Nina within the early and middle part of the season, and the SSTA there is a high probability of an above average summer.

With the QBO entering a -ve phase, there will be periods of high latitude blocking which may serve to allow the ridge to partially break down but this should be a persistent feature so long as La Nina remains weak and the polar stratosphere does not warm. A cooling stratosphere will open the flood gates for even greater warmth however.

The predicted temperature departure is between +0.75and +1.5 C for southern England which is perhaps surprisingly low given the forecast pressure anomaly patterns. Remember though the distorting factor of the polar stratospheric anomalies - which are not present and the likely pattern for June which is unlikely to deliver a record month given the pressure anomaly pattern. Rainfall is predicted at below average for the entire UK throughout the period although July is the most likely month for a return to more average rainfall totals.

So a summer more along one of the vintage traditional lines given the influence of the Azores High but with the added factor of warm SSTAs around the UK and the climatic trend to shift pressure patterns around 5degrees north, something very pleasant in store I think.

[summer blizzard]

1) the polar pressure anomaly, more particularly at the 30 hPa to 100 hPa level which is essentially a representation of the Arctic Oscillation (AO) and temperature of the polar stratosphere;

2) the climatic trend for warming;

3) the relative state of El Nino Southern Oscillation (ENSO);

4) Atlantic Sea Surface Temperature Anomalies (SSTAs).

I'd like to start by sayin that this is a very good forecast and i agree with your prognossis about a small number of factors influencing a forecast mainly, though i would argue that the PDO should hold weight in there as well.

I agree with your thoughts about no major warmings or coolings in stratospheric temperatures, though i think that as the summer progresses and the strength of the QBO increases, August is likely to record an AO value of -1> in my opinion, helped along by a weak La Nina developing by that time, while not enougth to influence the AO on their own, together i think they will be strong enougth to force a stratospheric warming event during July, which will affect August.

In summary, aside from a dry but cooler August than you expect, i think your forecast is pretty sound.

Incidently, i will also be releasing my own forecast in the next few days.

[Polar Continental]

Thanks GP, obviously a great deal of effort put in to your Summer forecast, makes for some very educational and interesting reading, although I admit, some of this gets a little to in-depth for me, but a vintage summer eh.. sounds ok, as long as it is followed by a classic vintage winter from,....now lets see, 1962-3 seemed like good years. :unsure:

Cheers GP, keep up the good work.

Paul

Edited May 28, 2007 by Paul Carfoot

[johnholmes]

I would echo fully what Paul has posted GP, heck of a lot of work there, some of it is beyond my working knowledge, but another good summer would be nice, fingers crossed.

[acbrixton]

Many thanks GP...will be intersting to see how things pan out...

Regards

ACB

[beng]

Great effort GP - as others have said you must have put a lot of time into this forecast.

ECM is now showing hints in its extended run of high pressure settling to the West of the UK during June - so your June suggestion of a high to our west is looking very possible.

[kold weather]

Yep a great effort there GP and I find your ideas are rather close to my own---more unsettled in July the the other two months but not like May but thanks to a lot of thundery outbreaks.

As for June, i agree with your ideas, here is what i said on 20th of May, very pleased with the models so far!!

A pretty easterly month I suspect (though depending on where the blocks we may see some southerlies with it, esp later in the month.) with another strong block forming somewhere nearby, exactly where determines the CET not sure it'll be that above average unless the block slides into Europe which it may do at one point

[Glacier Point]

Thanks all for your comments.

Just one or two updates before the forecast 'goes live' tomorrow.

The polar cross sectional temperature anomalies show on balance the slight suggestion of a cold anomaly across the upper atmosphere which, if this persists, will marginally favour warmer conditions to develop earlier in the season with the upper zonal winds stirred up once more. That could shunt the projected high a little further east to sit slap bang over the UK. AS things currently stand, there is no impact showing on the crucial polar height anomalies around 30 hPa so the general outlook remains as stated.

http://www.cpc.noaa.gov/products/precip/CW...hgt.ao.cdas.gif

The upper zonal winds across the equator are showing quite reasonable -ve anomalies so May's QBO is likely to be into the moderate strength category. This shouldn't have too much direct impact but it will serve to strengthen the subtropical ridges.

Edited May 31, 2007 by Glacier Point

[BLAST FROM THE PAST]

Based on certain cyclical factors I think this summer will not be anywhere near a record breaker and indeed the overall set up pretty different to last year. I think there will be different positioning of HP cells producing periods of warm to very warm weather just as the approaching period is being shown on the models but no major heatwave and no record heat. I think a period will occur where a Scandi trough is responsible for unsettled conditions as mentioned above possible mid June and mainly in August. This is when I think La Nina will be in play and will lead us to a wet end of summer and a wet autumn...totally different to our expected Indian summer scenario.

June

The first half of June will be dry and fairly warm the 2nd-12th being main good period. Although warmer weather may arrive in the first few days, temperatures will cool again as we approach the mid month. I think a period of unsettled squally thundery shower activity is likely around mid month for a week but the expected upward change to summer temperatures won't be felt until about June 21st onwards. So overall slightly warmer than average but no scorcher but with some hefty downpours.

July

This will be the sunniest and warmest month. The warm up in late June will carry through into July with a breakdown period around the second week although I think it will be shortlived and of the thundery activity as HP resets itself. I think a prolonged period of settled and very warm weather will then prevail latter half of July into early August. Above average and warmest month, drier than average

August

This is when I think La Nina will be evident and a breakdown to unsettled and [relatively] cool conditions. This for me will lead us into a wet autumn and bring summer to an early end....this is probably in comparence to recent long hot summers when I say an end to summer so I anticipate a more [good old British summer]. Close to average and wetter than normal.

BFTP

Edited June 1, 2007 by BLAST FROM THE PAST

[Glacier Point]

With June nearly at an end, time to review the first month and some thoughts on how the forecast has faired....

June 500 hPa pressure anomaly so far

June has seen a strong -ve height anomaly to the SW and significant +ve height anomalies to our north which has accounted for the very unsettled weather so far, and, more latterly, the cooler end to the month.

The polar height anomalies have been dominated by a mid to upper anomaly, which shows signs of reinvigorating....

http://www.cpc.noaa.gov/products/precip/CW...hgt.ao.cdas.gif

Interestingly the upper level between 100 and 30 hPa has seen little in the way of anomalies.

The June 30 hPa zonal wind anomaly shows no real anomaly across the northern hemisphere which accounts perhaps for the slow moving low pressures and blocking from warming of the mid layers over the Pole (no supression of this by anomalously strong upper zonal winds).

Turning to the forecasts, I'll use my four key assumptions as a framework for analysis:

1) the polar pressure anomaly, more particularly at the 30 hPa to 100 hPa level which is essentially a representation of the Arctic Oscillation (AO) and temperature of the polar stratosphere;

2) the climatic trend for warming;

3) the relative state of El Nino Southern Oscillation (ENSO);

4) Atlantic Sea Surface Temperature Anomalies (SSTAs).

Although the 30 - 100 hPa level has shown no real anomaly, the mid to upper layers have been dominated by blocking resulting in a -ve AO for most of June. This aspect clearly isn't going to plan and I think the state of the QBO may well explain a lot of this variance. The linear correlation for June / July easterly QBO is pretty consistent with the unsettled pattern experienced so far with high latitude blocking a feature.

Getting this assumption wrong has resulted in the expected June pressure anomaly pattern and rainfall prediction wrong.

In so far as the climatic trend, things are a little better. My range for the summer was +0.75 to +1.5 C deviation and June is likely to be band down the middle. So far so good.

In so far as ENSO is concerned, despite the UKMET's declaration of La Nina (where did they get that from ?), we are currently under ENSO neutral conditions with no immediate prospects of La Nina developing with just the possibility of an August weak La Nina. So far so good in this respect.

Turning to the SSTAs, the predicted pattern for June was wrong but I think the polar height anomalies have masked this (after all, it was the identified as the most improtant factor). July's predicted pattern however looks a lot more realistic with a deep upper low centred over or to the west of the UK.

If we sift through the analogues (based on SSTA pattern), then select those years with a -ve QBO, we are left with 1968 and 2001. The June- July pressure anomaly pattern for these years looks horribly familiar....

Forecast update:

I continue to expect this summer to fall in the above aveage category, with the range +0.75 to +1.5 C.

With almost two months rainfall 'in the bank' and an unsettled pattern for July predicted, an above average rainfall must be the revised call.

The predicted pressure anomaly pattern for July will resemble this:

There is some uncertainty for August's pressure pattern. Although August is the month most likely to see a -ve AO, the centre of the low pressure anomaly tends to move throughout Summers and I think there is already an indication of the position shifting further west which could open up the prospect of a small mid Atlantic trough.I'll keep to the orginal August analogue forecast for pressure anomaly at this stage.

GP