Author Archives: Simon Lee

About Simon Lee

PhD Meteorology student at the University of Reading.

January 2020: Memories of January 2007

In the UK, it’s currently mild – very mild. Provisional data through the 9th shows the mean Central England Temperature (CET) is running at 7.5°C, which is 3.9°C above the 1961-1990 average. The large-scale pattern during the month so far (Fig. 1) has been characterised by a ridge extending from the Azores to Europe, and low pressure near Greenland – creating a strong flow of mild, subtropical Atlantic air (the “tropical maritime” airmass). This is the positive phase of the ubiquitous North Atlantic Oscillation (NAO), which describes the primary mode of large-scale variability in the North Atlantic-European region. The NAO has been in a positive state since 28 December 2019, likely influenced by the strong stratospheric polar vortex.

Mean temperatures at the University of Reading have been well above the 1981-2010 average for every day of the month so far, with the 9th January a staggering 6.1°C above normal at 10.8°C!


Figure 1: 500 hPa geopotential heights and anomalies for Jan 4-10, 2020, from

The mild weather reminds me of January 2007, which was the UK’s second warmest on record (in terms of mean temperature) in the series since 1910 (5.9°C) behind only 1916 (6.3°C). January 2007 had a mean CET of 7.0°C, placing it 5th in the series since 1659, while in Reading it was the warmest January in a series since 1908 with 7.6°C (by contrast, 2020 runs at 8.0°C so far…)

As this was before the colder, snowier period of 2009-2013 (and I had yet to live through a particularly snowy winter despite being in North Yorkshire), I remember thinking that the unusual warmth must have been due to climate change (it followed hot on the heels – pun intended – of many UK heat records in 2006). Certainly, January 2007 was a hot one for most of the globe.  It was the first month in NASA’s analysis to be more than 1.0°C above the 1951-1980 average (Fig. 2), something which we only saw again in October 2015 (and which has since become “normal” with 16 subsequent months breaching that threshold).


Figure 2: Global temperature anomalies for 2007 with respect to 1951-1980. NASA GISS analysis (

You can see from the 500 hPa geopotential height anomalies for January 2007 (Fig. 3) that it was not particularly reminiscent of January 2020 thus far, with a much less zonal Atlantic. January 2007 did not have an extremely positive NAO – in fact, it turned negative from the 20th of the month, and it was relatively cold for a few days in the final third of the month. Without this colder spell, it would have almost certainly taken the crown off 1916 for the UK’s warmest.


Figure 3: 500 hPa geopotential height anomalies for January 2007 from NCEP/NCAR reanalysis (

It’s clear, therefore, just how much small circulation changes or persistence can have on these monthly-level anomalies in the heart of the winter and on the receiving end of volatile Atlantic dynamics – introducing a lot of noise into the climate signal for the month. Generating extremes always needs a “perfect storm” – in this case, a persistent, unchanging pattern bringing warm air for as much of the month as possible, superimposed on background global warmth.

Nevertheless, January has warmed in the UK (Fig. 4), though this is perhaps most noticeable in the absence of cold rather than the presence of warmth – likely in part particularly due to warming seas, but also the rapidly warming Arctic meaning cold air outbreaks are warmer than they once were.


Figure 4: UK January mean temperatures since 1910, expressed as departures from the 1910-2019 mean. Data source:

January 2020 may go on to be a record breaker. Equivalently, it may not, thanks to the sensitivity of the UK’s temperatures to fine changes in circulation patterns. But thanks to global warming, it’s increasingly likely that 1916 won’t hold the record for much longer if the right weather patterns do occur. On that note, Charlton-Perez et al. (2018) found a slight increase in persistence of the positive NAO regime under strong stratospheric vortex conditions. With the forecasts suggesting a strong stratospheric vortex for the remainder of the month, we might just be in for a record-breaker.


Figure 5: CFSv2 forecast from 10 January 2020 of 10 hPa 60°N zonal-mean zonal winds.

There’s more to the UK winter than the stratosphere

The title might seem rather obvious – or a surprising statement coming from me! But what I seek to achieve in this short blog post is provide a simple reminder that historically cold weather has indeed happened without a sudden stratospheric warming (or significant stratospheric polar vortex disruption).

The role of the stratosphere can sometimes be overstated. We saw in winter 2018-2019 how a major SSW doesn’t even necessarily spell dramatic cold weather for the UK since some stratospheric events do not couple strongly downwards. The state of the stratospheric vortex is a “boundary condition” to which the troposphere responds (to varying degrees). That’s what makes it so fantastic – we can gain predictive skill from it for several weeks. But it is not the only boundary condition.

However, for us in the UK on the receiving end of an “Atlantic radiator” of warmth, deep cold does need something significantly unusual to happen. A breakdown of the mean westerly circulation in the stratosphere can do that (as we saw in Feb-March 2018) but other drivers can provide similar impacts. The troposphere is a noisy place.

Figure 1 shows that the DJF-mean 10 hPa 60N zonal-mean zonal wind has only a weak relationship with the UK DJF mean precipitation (temperature even less so, but I’ve used rainfall here as it has no clear trend, and colder winters tend to be drier). It’s interesting that the very wettest winters have had a contemporaneously strong 10 hPa vortex, suggesting these are unlikely without stratospheric support – but that’s about as much as this shows. Evidently, strong vortex does not simply equal wetter and stormier.


Figure 1: DJF-mean 10 hPa 60N zonal-mean zonal wind against UK DJF mean precipitation for DJF 1979-1980 to 2017-2018. A kernel density estimate is applied.

Now time for a few case studies to illustrate the point…

December 2010

With a mean Central England Temperature (CET) of -0.7C, December 2010 was the coldest December since 1890 and the first month with a mean CET below freezing since February 1986. It is the only recent winter month which comes anywhere near rivalling the “great winter” months (e.g. January 1963) that have become increasingly rare due to climate change. I often wonder how cold it might have been had it occurred 50 years earlier.

But, while the winter of 2009-2010 saw some of its cold weather associated with a major SSW on February 9th 2010, there was no major SSW in December 2010. In fact, the 10 hPa vortex was around average strength (Figure 2). There was no “dripping paint” from the stratosphere; the anomalous blocking/negative Arctic Oscillation pattern was confined to the troposphere (Figure 3).


Figure 2: 10 hPa 60N zonal-mean zonal winds for 2010-2011 per MERRA-2 reanalysis (via


Figure 3: “Dripping paint” of 65-90N standardized geopotential height anomalies for Oct-Dec 2010. In Dec 2010, higher than average height anomalies (circled) were confined to the troposphere and did not have a stratospheric precursor (via

Evidently December 2010 was extraordinary, so I’ll support this with another example…

January 1987

The easterly outbreak in January 1987 makes 2018 look like the Wimp from the East, and it was generally unrivalled until 2018’s Beast.  Figure 4 shows the 850 hPa temperature field on 12 January 1987, when the cold wave was near peak intensity.


Figure 4: 850 hPa temperature and geopotential height for 12 January 1987 at 00Z, from the CFSR reanalysis via

Some of you may recall there was a major SSW in January 1987, and you’d be correct – but it was on January 23rd, after the cold weather outbreak! In fact, the Scandinavian ridge that brought the cold wave in 1987 may have helped drive the SSW which followed. Lehtonen and Karpechko (2016) pointed out how the tropospheric patterns (blocking) associated with these severe cold waves can act as drivers of SSWs themselves, with the greatest cold preceding the stratospheric event but the best predictability following the event. Something similar may also have occurred in January 2010.

So, as we move into a period where the polar vortex looks to be strengthening – remember that there’s more to the UK winter than the stratosphere…

10 FAQs on “Increased shear in the North Atlantic upper-level jet stream over the past four decades”

Lee, S. H., P. D. Williams, and T. H. A. Frame, 2019: Increased shear in the North Atlantic upper-level jet stream over the past four decades. Nature,

  1. What is the key result of the study?

Our study shows that the annual-mean vertical west-to-east (zonal) wind shear over the North Atlantic at aircraft cruising altitudes (within the jet stream) has increased by 15% (with a range of 11-17%) over the last four decades (1979-2017).

  1. What is vertical wind shear?

Vertical wind shear is the name given to the change in wind speed (or direction) with height. In our study, we consider the change in the speed of the west-to-east (“zonal”) wind with height.

  1. What data were used to get these results?

We used three reanalysis datasets (ERA-Interim, JRA-55, and NCEP/NCAR). These are our best estimates of the past evolution of the atmosphere, based on observations. We use data from 1979 onwards, as satellites have been routinely used to monitor the state of the atmosphere since then, increasing the reliability of the data.

  1. Why has wind shear increased?

On large scales like the North Atlantic, the vertical wind shear is driven by the difference in temperature as you move from the equator to the pole. When this is larger, the wind shear is larger. At aircraft cruising altitudes (around 34,000 feet), this temperature difference has increased, causing an increase in wind shear.

  1. Why has the temperature difference increased at aircraft cruising altitudes?

At low latitudes (less than about 50N), temperatures are warming at this level in response to increasing greenhouse gases. Poleward of 50N, temperatures are in fact cooling. At these latitudes, this level (34,000 feet) is in the stratosphere, not the troposphere where we live. Increasing greenhouse gas concentrations actually cause this layer of the atmosphere to cool. There is also a cooling effect from a decline in concentrations of ozone.

  1. Why should we care?

Wind shear is a key driver of clear-air turbulence (CAT), which is a significant aviation hazard as it is invisible and difficult to predict. Turbulence in general has been found to play a role in the majority of weather-related aviation incidents. Climate models project a large increase in CAT over the North Atlantic under future climate change (Storer et al. 2017). Our study shows observation-based evidence that a key driver has already increased by a large amount, supporting these projections.

  1. The study finds shear has increased. Have encounters with turbulence increased?

Unfortunately, it is very difficult to say this with any degree of confidence. Over the last four decades, airplane designs have changed, flight numbers have increased, and we are now able to identify regions of possible turbulence and avoid them – so it is difficult to produce any consistent record of turbulence observations. However, our study suggests it may have, as a key driver has increased by a large amount.

  1. So, is the jet stream strengthening or weakening?

We find that, in the lowermost ~5 km or so, the temperature difference between the equator and pole is weakening, thanks to the rapidly warming Arctic. This contrasts with the increasing difference at higher levels. We find that these effects are approximately equal and opposite, like a balanced tug-of-war. The overall speed of the jet stream aloft depends on the total effect (the vertical integral), so there has been no significant change to the speed of the jet stream at aircraft cruising altitudes. As our study only looks at the annual-mean, it is possible that different weakening or strengthening trends occur in different seasons, as other studies have found.

  1. Is the jet stream becoming wavier?

We do not look at this in our study. Some previous studies have suggested this, primarily as a result of the decreasing temperature difference in the lower part of the atmosphere. Our results neither prove nor disprove these findings.

  1. What does this mean for the future?

We focus on what has been observed to occur, rather than what is projected. However, most studies suggest the upper-level “tug” will win, leading to a strengthening jet stream, with impacts on transatlantic flight times (Williams 2016). As increasing greenhouse gas emissions continue to increase temperatures in the troposphere and cool those in the stratosphere, we expect shear to continue to increase at flight levels (though large interannual variability is likely), consistent with model projections of increasing turbulence.

Canadian, TX tornado – May 23 2019

May 23rd, 2019. Day 4 of our 2019 chasecation. Only a few days after the infamous and frustrating “high risk bust” in Oklahoma on May 19th (which remains the worst chase day I’ve had), we found ourselves in the Texas Panhandle under a Moderate risk – having driven all the way from Missouri the preceding day, where we had successfully chased the deadly Golden City tornado.

We based ourselves in Perryton, only a few miles south of the Oklahoma border, and in traditional chaser fashion took up residence in McDonald’s (cheap coffee, fast food, WiFi… heaven!).


McDonald’s in Perryton, TX, with a developing severe thunderstorm behind, 2:27 PM.

By around 3:30 PM we were extremely close to a severe thunderstorm, but the visibility was terrible. Anyone who chased during this period will recall just how much haze and dust there was, and we were filled with the same frustration we felt back on the high risk day. Eventually, we gave up on this storm due to an inability to see anything noteworthy. It went on to be tornado-warned shortly after we abandoned it, so we thought we had made a huge mistake…

How close we were to the storm vs. the seemingly quiet grey skies we could see…

However, by 5:30 PM a new tornado-warned cell had developed further south, and we were ready and waiting – along with a huge number of other storm chasers which included the TORUS field campaign and Reed Timmer! We were in a perfect position and very hopeful, but yet again visibility was extremely poor due to the haze. Several possible wall clouds were noted amongst the murk, but the comparison between the NWS warning of a “very dangerous storm” and what we could actually see was stark.

The Dominator, and a healthy looking tornado-warned storm on radar – but not looking at all interesting in reality!

So, we gave up on this storm and headed a bit further south and east, now concerned that the storms would grow upscale into a linear mode – though we were far enough from the KAMA radar that there was quite a difference between the radar reflectivites and the ground truth (always worth remembering the beam height!).

At around 7 PM, we pulled over to look west and re-think our plan. At this point I’d pretty much lost hope, and I was chatting to a fellow frustrated storm chaser about the haze. Suddenly, one of my chase team (thanks Rounak!) suddenly cried out “THERE! ON THE HORIZON!” (make sure you read that in an Australian accent).

Our first view of the tornado (looking W from US Highway 83 N of Canadian) as it appeared on the horizon in the wilderness, and the accompanying radar presentation. Note that the storm was not tornado-warned whilst a large tornado was already in progress, highlighting the need for spotter observations.

Sure enough, a big wedge tornado had dropped to the ground and emerged out of the murk, and out of a region where there were so few roads it was almost impossible to actually be close enough to see the tornadogenesis. After grabbing a photo and quickly tweeting NWS Amarillo (important storm chaser duty), we all hit the road and bolted north to get close to the tornado – filled with awe and adrenaline that there was actually a big wedge on the ground and we were about to get really close to it! *cue classic excited girly noises*

By 7:09 PM we were just NW of Canadian on Highway 83, and now very close to the wedge. This is where the whole experience became surreal, and the reason I’m writing this blog. Here I was, standing under a kilometre away from what must have been, at that time, the most violent atmospheric phenomenon on Planet Earth. Wind and rain were lashing us, and right in front of us was this huge tornado churning along through the high Texan plains. It was making an incredible sound which I can’t quite describe, but I think at the time we described it as “eating”.

A collection of photos of the Canadian wedge tornado taken around 7:11 PM.

Some might have felt fear, others might have felt satisfaction or scientific curiosity.

I just remember feeling completely transfixed. I remember other members of my crew shouting that we should go because we were probably about to become too close, but finding it very difficult to tear myself away from it no matter how close it was becoming.

The tornado felt strangely personable. Now, I’m aware that’s a strange thing to say, and you might roll your eyes reading about a meteorologist having a moment like this. But in that moment I felt a transcendental connection to the atmosphere which I’m sure I won’t ever forget. I am now finding that it is difficult to put it into words.

Maybe it was a rush of knowing that the thing I love most – the atmosphere – was producing a deadly amount of vorticity right in front of me. Deadly, but in a peaceful way – tornadoes aren’t malicious, because the universe isn’t malicious (I’m sure this is almost a quote from Interstellar). Maybe it was the realisation of a life-long dream to get close to a wedge tornado.

It is moments like these why I chase storms. To be right there with the awesome power of nature, and to be all-encompassed by such a tremendous, indescribable feeling. I’m not religious, but I’d like to think this is what seeing God would feel like.

The tornado was rated EF2, but NWS noted that it may have been stronger during its time out in the open fields.

Trump on climate change: what he should have said

If I followed up every time Donald Trump opened his mouth on climate change with a blog post pointing out where he was wrong, I’d have no time left to do anything else – but this one is a bit more special. This week, Trump visited the U.K., and part of that visit involved a conversation with Prince Charles – an advocate of organic farming and fighting climate change. They couldn’t really have more opposing views, so what did POTUS have to say about climate change afterwards?

The following are quotes lifted from a BBC News article (as of 5 June 2019).

“I believe that there’s a change in weather and I think it changes both ways,” Mr Trump told Piers Morgan in an interview that aired on Wednesday.

This would make more sense if he said a “change in climate”, as weather changes all the time. This one sentence encapsulates the President’s inability to grasp the difference between weather and climate. He’s correct in what he says – weather changes both ways! Indeed, some places do have a cooling climate – e.g. the North Atlantic warming hole – whilst it’s possible that global warming may lead to all kinds of extreme weather including cold weather extremes due to changes to the jet stream.

What he should have said: “I accept that there is solid scientific evidence for climate change, which, whilst it doesn’t necessarily rule-out some short-term increase in cold weather extremes, indicates there will be a long-term shift to overall more warmer weather.”

“He wants to make sure future generations have climate that is good climate as opposed to a disaster and I agree.”

Ah yes, “good climate vs. disaster climate”. He says he agrees… but does he? The following quote suggests he considers creating “good” climate to have nothing to do with greenhouse gases…

What he should have said: “Prince Charles, like many, is fighting to ensure future generations, and all plants and animals, can thrive on this planet even more than we do now. I want to join him in that.”

But Mr Trump once again placed the blame on other countries, namely China, India and Russia, for worsening air and water quality while claiming the US has one of “the cleanest climates there are”.

This is a great example of Trump seemingly getting confused between air quality, climate change, and greenhouse gas emissions. No one refers to a “clean” climate. That doesn’t make sense. If this statement was about greenhouse gases, it’s wrong – with US emissions placed 2nd to China globally.

What he should have said: “The US is one of the largest contributors to global greenhouse gas emissions, and we need to do something about that.”

“Don’t forget, it used to be called global warming, that wasn’t working, then it was called climate change, now it’s actually called extreme weather because with extreme weather you can’t miss,” the president said.

[insert WRONG! Trump GIF]

This is a fallacy which Trump keeps repeating. Global warming (the rise in Earth’s average temperature) drives climate change, which is defined over long time periods. It is manifest in an increased frequency of extreme or record-breaking weather events. I actually believe Trump doesn’t understand that, rather than repeating the falsehood for other reasons.

What he should have said: “Global warming leads to climate change, and now we’re starting to see the effects of this with extreme weather events around the world.”

Mr Trump pointed to past examples of weather disasters to refute the idea that “extreme weather” is becoming more common due to climate change.

“I don’t remember tornados in the United States to this extent but then when you look back 40 years ago we had the worst tornado binge we ever had. In the 1890s we had our worst hurricanes.”

Some serious cherrypicking here. And what’s a tornado binge? There’s also some implication in these words that people have suggested the recent US tornado outbreak is due to climate change, and that they have said so purely because of short-term memory. That’s a load of garbage. Moreover, tornadoes and hurricanes are among the more contentious when it comes to the effect of global warming on their frequency. Tornadoes and hurricanes are also very US-centric, suggesting the President doesn’t care about increases in severe weather in other parts of the world…

What he should have said: “I took an Advanced Statistics class at college, and I know that, in order to see whether there is a long-term change occurring, I need to perform many forms of statistical analysis – including significance tests on linear regression and Kolmogorov-Smirnov tests, to determine if my data is showing climate change, and not to randomly pick outliers in the past to prove my agenda. I also need to account for changes in data acquisition and homogeneity over time before making any conclusions. And a good example of this is the catastrophic loss of Arctic sea-ice since the late 1970s, when reliable records began.”

Final Stratospheric Warmings

The stratospheric Polar Vortex is currently at record-strong levels, based on the metrics of 10 hPa 60N zonal-mean zonal-wind and 60-90N average temperature. This is likely to be due to a combination of the timing and duration of the major Sudden Stratospheric Warming (SSW) in January: the duration of easterlies (most of January) shielded the vortex from vertically propagating wave activity for an extended period (these waves cannot propagate into a layer of mean easterly flow) allowing the vortex to redevelop following the SSW, whilst the early-season timing allowed strong radiative recovery in the absence of significant solar radiation. The associated positive stratospheric NAM is well coupled to a positive tropospheric NAM, which seems also to have enhanced North Atlantic storminess. It’s all a far cry from where 2019 began.


10 hPa winds per GFS 12Z analysis March 15 2019. With U1060 at 49.2 m/s, this fulfills the ‘strong vortex event’ criterion of Tripathi et al. 2015 (41.2 m/s). Image credit

However, and not for the first time this winter, our attentions have turned to stratospheric warming! Recent forecasts from the GFS ensemble have shown a large pulse of wave activity with an amplification of the Aleutian ridge stretching and displacing the vortex. Stratospheric temperatures are expected to rise rapidly, and some ensemble members are suggesting [U] 10 hPa 60N to turn easterly again in early April.

So, a rapid rise in stratospheric temperatures, accompanied by a displacement of the polar vortex and a reversal of zonal winds… is this then another major Sudden Stratospheric Warming?

Yes and no.

The major SSWs we talk about are more technically “major midwinter warmings” (MMWs), which are followed by a recovery of the stratospheric Polar Vortex to westerlies for at least 10 consecutive days before April 30 (Charlton and Polvani, 2007). The upcoming warming event is likely to become what is known as a dynamical Final Stratospheric Warming (FSW), also known as a “major final sudden stratospheric warming (“major final warming”, MFW)” (Manney and Lawrence, 2016).

The “Final Warming” refers to the transition of the stratosphere to “summer mode” – when the polar vortex dissipates and easterlies develop which persist until late August/early September when the vortex reforms. This is most simply driven due to thermodynamics – a result of solar radiation returning and warming the polar region. Climatologically, the last day of zonal mean westerlies is April 12th.

However, sometimes a sudden warming driven by the same dynamical processes which cause MMWs can occur so late in the season that the vortex is unable to recover as the Sun returns to the pole. This then becomes a dynamical FSW.

There’s a couple of reasons why FSWs don’t get the same attention as MMWs. They are dynamically different in that the planetary wave behaviour following the event is different (since stratospheric easterlies remain). The late-season timing of the events means the typical cold-weather outbreaks don’t occur with quite such severity, and tropospheric responses are often lost in the seasonal transition (with tropospheric jet streams and storm tracks evolving from winter-mode to spring-mode). There’s also the consideration of the anomaly magnitude – stratospheric zonal winds are weak or easterly during April regardless of whether a dynamical FSW occurred, which is very different to mid-winter. But, largely, the differentiation is just based on statistics, and everything in the build-up and occurrence of the warming is the same – which makes them another source of information about the behaviour and predictability of the troposphere and stratosphere.

Some FSWs are rather special and are “SSWs in hiding”. March 5th 2016 saw a very strong stratospheric warming event which produced the second-strongest zonal-mean easterlies on record (at 10 hPa 60N), but is often forgotten because it became an FSW [interestingly, the 2016 event also decelerated a record-strong vortex, like 2019 is expected to do…]. The oft-mentioned statistic of the February 2018 SSW being the first since January 2013 is indeed true, but it means the March 2016 event – which was an MMW in all but a technicality – is forgotten. There was some lively debate on Twitter recently as Judah Cohen argued the March 2016 event deserved to be included in the SSW Compendium! Indeed, it did have some significant surface impacts with downward propagation of a negative NAM phase into the troposphere. Britain saw a cold and snowy Easter that year, for example.


MERRA-2 data for 2015-16 10 hPa 60N zonal-mean zonal wind. The event in March 2016 saw date-record strong easterlies which did not return to westerlies, and thus was a Final Warming. Image credit NASA.


Time-height cross-section of normalized 65-90N geopotential height anomalies. The March 2016 stratospheric warming event is clearly evident, with downward propagation into April. Image credit NOAA CPC.

You have to draw the line somewhere. Maybe we should have a Final Stratospheric Warming Compendium so these events are given equal treatment.

Returning to 2019, given that we currently have a strongly zonal state, it will be interesting to see whether the upcoming warming event has an effect on that – though discerning any transitions from those which would occur simply because of the seasonal cycle may be more challenging. And, just like sudden warmings, we face the question of whether or not the troposphere is even all that bothered by the stratosphere!


Tripathi, O. P., A. Charlton-Perez, M. Sigmond, and F. Vitart, 2015: Enhanced long-range forecast skill in boreal winter following stratospheric strong vortex conditions. Environ. Res. Lett.

Charlton, A. J., and L. M. Polvani, 2007: A New Look at Stratospheric Sudden Warmings. Part I: Climatology and Modelling Benchmarks. J. Climate.

Manney, G. L., and Z. D. Lawrence, 2016: The major stratospheric final warming in 2016: dispersal of vortex air and termination of Arctic chemical ozone loss. Atmos. Chem. Phys.

A “winter heatwave” in a warming world

The final week of February 2019 has been characterised by anomalously warm, record-setting conditions over NW Europe. The United Kingdom broke its all-time maximum record temperature for February on several occasions and at several stations – the previous record of 19.7C from 1998 was obliterated, replaced with a new record of 21.2C (a huge difference of 1.5C, which were it to be replicated in August would see the UK experience 40C). For the first time, the UK experienced 20C during a winter month, and this moved the date of the first recorded 20C forward from March 2nd to February 26th. This was by all counts a “winter heatwave”, in magnitude and duration, and widely produced temperatures which wouldn’t be out of place in summer.

At the University of Reading, we also saw a new all-time (since 1908) record maximum for February – the previous record of 17.4C (which was first tied on Feb 25th!) was replaced with 19.5C on Feb 26th.


A classic “heatwave” sunset on February 27 from Whitley Wood Road, Reading, after temperatures reached 17.9C. A touch down on the previous day’s 19.5C, but still 0.4C above the old record!

Why was it so warm?

This is a difficult question to answer, but there’s several components which seem to have been required in order to get the atmospheric configuration such that high temperatures were possible over the UK. Here I present a few that I’ve noticed, but there’s likely other finer components, too (these are not necessarily in any meaningful order):

  • Rossby wave train: evident in Figure 1, there is a pattern in the 200 hPa height anomalies suggesting a Rossby wave train propagating out of East Asia and the Pacific has been evident for the last week. This provides the enhanced meridional flow associated with blocking weather regimes. Figure 2 also shows anomalously weak 250 hPa zonal flow in the mid-latitudes, suggesting reduced propagation speeds of weather systems allowing for (and associated with) extended blocking regimes.

Figure 1: 7-day (20-26 Feb) mean 200 hPa height anomalies from NCEP/NCAR Reanalysis. Apparent Rossby wave trains are shown with superimposed black arrows.


Figure 2: 250 hPa zonal-mean zonal wind anomalies from NCEP/NCAR Reanalysis for 20-26 Feb 2019. Note the anomalously weak zonal winds in the N Hemisphere mid-latitudes.

  • Extreme eastern USA jet streak & cyclogenesis: the record-setting jet stream winds seen on Tuesday 19th preceded the development of the blocking ridge. This may be associated through the downstream impacts of such extreme winds (Figure 3) – decelerating an unusually strong jet requires a very active jet exit region, leading to strong (anti)cyclogenesis. A series of deep cyclones (Figure 4) developed in the jet exit region, and when combined with other factors aiding their meridional track, the cyclones likely acted to build the downstream ridge, with positive feedbacks, helping to amplify the pattern. HYSPLIT trajectories also suggest some of the air over the UK originated within the extreme jet streak prior to undergoing strong descent, which may have been aided by its unusually strong nature driving unusually strong descent.

Figure 3: 250 hPa winds on Feb 19th showing a possible downstream impact of ridge amplification over NW Europe.



Figure 4: MODIS view of a 938 hPa cyclone in the central North Atlantic on Feb 20, 2019.

  • Strong adiabatic descent: HYSPLIT back-trajectories shown in Figure 5 reveal the airmass over the UK originated near the tropopause a few days prior, before descending through the depth of the troposphere. This not only adiabatically warms the air (on top of its warm source region), but also dries out the entire column, allowing for strong insolation needed for the sensible heating to generate strong surface warming.

Figure 5: Ensemble of 84 hour backwards trajectories for air at 1500 m AMSL over London at 12Z Feb 25th based on GFS 0.5 degree data.

  • Anomaly persistence: once established, the block lasted for several days. This allowed for further descent of air which also underwent diabatic warming thanks to the intense radiation under cloudless skies – recirculating around the anticyclone (a similar pattern existed during the summer 2018 European heatwave).

These are the weather components which contributed. They describe the prior and contemporary state of the atmosphere. To relate this to the climate, I’ll draw an analogy. You exeperience a car crash. Why? What I have presented so far would be equivalent to saying “You ran a stop sign”. Now we naturally ask, “what about climate change?”. In my analogy, this is asking, “Were you intoxicated?”. Being intoxicated doesn’t mean you will run a stop sign, and you certainly can do so without being drunk, but it will increase your risk of doing so.

There is no doubt that the configuration of the atmosphere during the last week has been extreme, and primed for producing these warm temperatures. However, in a stationary climate we do not expect to break records with the frequency that we are doing, especially given a lengthening record (e.g. Kendon 2014). Now that we have warmed the mean temperatures, an extreme dynamical perturbation to the mean state (e.g. a monster blocking ridge) will produce an even more extreme temperatures than we would have seen beforehand.

This mechanism is supported by looking more closely at the University of Reading’s weather data record (Figure 6 & Table 1). Similar events, even with similar sunshine, have historically produced cooler temperatures. The recent frequency of extremely warm February temperatures is also evident, and you can also see recent cases of very warm temperatures with much less sunshine than older cases that matched the temperatures but only with strong solar forcing – suggesting, as I mentioned earlier, that it doesn’t take as much of a ‘push’ to equal temperatures which were once close to a “theoretical maximum”, such that now we can obliterate those records with sufficiently unusual large-scale anomalies.


Figure 6: Data from Reading University Atmospheric Observatory 1957-2019 showing daily maximum temperatures above the monthly 95th percentile and associated sunshine hours. Red indicates February 2019, grey indicates pre-2000, black post-2000. The 2019 record is shown with a red star. Tmax exceeding 16C is selected for further analysis in Table 1.


Table 1: Data corresponding to the points within the box in Figure 6 plus the 2019 record value.

But sunny, warm weather in February is nice!

Indeed it is – it was my birthday on February 24th, and I never expected to be celebrating it sitting outside! This event didn’t have the same severe impacts as a summer heatwave, but to me it almost felt more disturbing – the knowledge of what this might mean should a similar extreme be generated in the summer months, and that climate change was “eating away” at winter’s very existence. Unusual late winter/spring temperatures mainly impact the natural world which is highly sensitive to temperature and sunshine at this time of year (e.g. Figure 7), and this is why we should care – this could have many wide-ranging impacts on the ecology which supports our existence.


Figure 7: Blackthorn blossom, complete with Honey Bee (if you look closely), on Feb 23rd in Reading. This blossom is more likely in March and April.

NCEP/NCAR anomaly plots credit