The distribution of extremely low temperatures

[JacobWilliams]

So I have just been reading about the extremely cold weather of January 1982 and I'm curious about what the pattern of temperatures across a large area looks like under certain conditions. So far as I have been able to gather, on the night when Newport in Shropshire set the English low temperature record of around -26, temperatures across the midlands, where skies were clear anyway, were around minus 15 to 18 in most locations. At my local station in Oxford we reached -16.6, which is our all time low record. What interests me is firstly what factors determine where the extreme cold temperatures are recorded in such conditions--is Newport a notorious frost hollow, or is there a strong element of randomness, in which case Oxford could theoretically reach -20 under a perfect storm of conditions? Extreme low temperature events certainly seem more complicated than 'extreme' high temperature events, where the whole of the SE can be 35c+ away from immediate coastal areas.

 

Secondly, what is lowest temperature that can feasibly occur in a widespread fashion in England/the UK (I say England because in a mountainous area like the Scottish highlands I imagine it's much much more complex)? Is it, as I assume, around -16 to -18? And how often do these very low temperatures occur--the lowest that has occurred where I live in my lifetime (a mere twenty years) is, I think, about -12, in December 2010, which is an incredible 14 degrees above the all time record for a location with pretty much the same winter averages...

 

So far as I can tell, a very cold spell, in which temps might reach -10 to -15 across a large area is much more common than an extremely cold one in which they might fall below -15, which might be a once or twice in a lifetime event--and the colder the widespread temperature, the greater the variance, hence the extraordinary reading at Newport. But maybe I'm wrong about some of this. Obviously I am somewhat ignoring urban areas here which seem to struggle even to reach negative double figures at all. I was also fascinated to read that the average daily maximum across many areas in Feb 1947 was as low as -2 to -3--this is like seeing a July averaging over 33 degrees!

 

Perhaps this is a somewhat perverse subject to think about for those of us in the south today, but I'd welcome any input. And if anyone has done any analysis of these events or has any data on the probability distribution of very low temperatures and the effect of location on this distribution I'd be incredibly excited to see it. 

 

If this is in the wrong forum, mods please feel free to move it.

Edited April 15, 2015 by JacobWilliams

[4wd]

With good snow cover and long nights like December and January it can get very cold under clear skies and little wind.
If the conditions persist for several days there is little heat from the sun and days can stay well below freezing too as nights get ever colder.
A kickstart to this scenario is pulling in an exceptional cold pool to start with.
In December 1981 we had a run of days where it plunged as low as -12 even before sunset but typically it would creep up later.
There was about 18 inches of deep powdery snow which lasted most of the month.

Although this area can be a rather cold area it isn't a notorious frost hollow, and there's no reason why large areas of England couldn't get this somewhat freakish set up if conditions conspire.
Built up areas these days will always stay considerably warmer than rural though.
The warmth among buildings is very noticeable when motorbiking at night or just watching a car thermometer - It can easily be 5C warmer in town. -

[hillbilly]

So I have just been reading about the extremely cold weather of January 1982 and I'm curious about what the pattern of temperatures across a large area looks like under certain conditions. So far as I have been able to gather, on the night when Newport in Shropshire set the English low temperature record of around -26, temperatures across the midlands, where skies were clear anyway, were around minus 15 to 18 in most locations. At my local station in Oxford we reached -16.6, which is our all time low record. What interests me is firstly what factors determine where the extreme cold temperatures are recorded in such conditions--is Newport a notorious frost hollow, or is there a strong element of randomness, in which case Oxford could theoretically reach -20 under a perfect storm of conditions? Extreme low temperature events certainly seem more complicated than 'extreme' high temperature events, where the whole of the SE can be 35c+ away from immediate coastal areas.

 

Secondly, what is lowest temperature that can feasibly occur in a widespread fashion in England/the UK (I say England because in a mountainous area like the Scottish highlands I imagine it's much much more complex)? Is it, as I assume, around -16 to -18? And how often do these very low temperatures occur--the lowest that has occurred where I live in my lifetime (a mere twenty years) is, I think, about -12, in December 2010, which is an incredible 14 degrees above the all time record for a location with pretty much the same winter averages...

 

So far as I can tell, a very cold spell, in which temps might reach -10 to -15 across a large area is much more common than an extremely cold one in which they might fall below -15, which might be a once or twice in a lifetime event--and the colder the widespread temperature, the greater the variance, hence the extraordinary reading at Newport. But maybe I'm wrong about some of this. Obviously I am somewhat ignoring urban areas here which seem to struggle even to reach negative double figures at all. I was also fascinated to read that the average daily maximum across many areas in Feb 1947 was as low as -2 to -3--this is like seeing a July averaging over 33 degrees!

 

Perhaps this is a somewhat perverse subject to think about for those of us in the south today, but I'd welcome any input. And if anyone has done any analysis of these events or has any data on the probability distribution of very low temperatures and the effect of location on this distribution I'd be incredibly excited to see it. 

 

If this is in the wrong forum, mods please feel free to move it.

Benson just outside Oxford got down to -20deg in February 1917.Kent took the coldest temp in 1947 at -21deg! If you look at some stats you will find it does get cold in the south! leistershire has recorded below -23 deg.

Edited April 16, 2015 by hillbilly

[damianslaw]

A decent snow cover, clear skies and light/no winds are the conditions required for very low temperatures. Local topography is the other key factor, locations in natural hollows or a bowl such as a river valley away from built up features allows cold air to collect are at much greater risk of seeing low temperatures, compared to very open locations, built up areas, high ground and coastal locations etc. 

 

I know Benson is a local frost hollow as well as Rickmansworth. Not sure about Newport.

 

Here there is often a lot of local variance even from valley to valley. We can get very low temperatures in the right conditions despite not being particularly sheltered, but local topography helps with cold air able to filter down from nearby Troutbeck valley and low hills to the east - despite having a Lake to our west and being relatively close to the sea compared to Pennine, welsh marches, midland and central southern England regions. Its a long time though since we have got below about -12 degrees. Last time was Dec 95 when we hit -18 degrees. Local records show Ambleside not too far away dropped below -21 degrees in Jan 1940.

 

A more interesting question is how coastal locations on occasion can see very low temps, Crosby I believe as recently as Dec 2010 got down to -18 degrees I believe.. what caused such low temps, perhaps the sandy soil which I think can have some impacts.

Edited April 16, 2015 by damianslaw

[Midlands Ice Age]

With good snow cover and long nights like December and January it can get very cold under clear skies and little wind.

If the conditions persist for several days there is little heat from the sun and days can stay well below freezing too as nights get ever colder.

A kickstart to this scenario is pulling in an exceptional cold pool to start with.

In December 1981 we had a run of days where it plunged as low as -12 even before sunset but typically it would creep up later.

There was about 18 inches of deep powdery snow which lasted most of the month.

Although this area can be a rather cold area it isn't a notorious frost hollow, and there's no reason why large areas of England couldn't get this somewhat freakish set up if conditions conspire.

Built up areas these days will always stay considerably warmer than rural though.

The warmth among buildings is very noticeable when motorbiking at night or just watching a car thermometer - It can easily be 5C warmer in town. -

 

The 62-63 winter also started with a gradual cooling off under anticyclonic conditions after a cool flow had been established from the north west. Not the lows you are referring too, but it gives us the best chance of a really cold spell. I am discounting a true beast from the East as was experienced in 1947 as it always seems to miss the UK. 

 

MIA

[Interitus]

A decent snow cover, clear skies and light/no winds are the conditions required for very low temperatures. Local topography is the other key factor, locations in natural hollows or a bowl such as a river valley away from built up features allows cold air to collect are at much greater risk of seeing low temperatures, compared to very open locations, built up areas, high ground and coastal locations etc. 

 

I know Benson is a local frost hollow as well as Rickmansworth. Not sure about Newport.

 

Here there is often a lot of local variance even from valley to valley. We can get very low temperatures in the right conditions despite not being particularly sheltered, but local topography helps with cold air able to filter down from nearby Troutbeck valley and low hills to the east - despite having a Lake to our west and being relatively close to the sea compared to Pennine, welsh marches, midland and central southern England regions. Its a long time though since we have got below about -12 degrees. Last time was Dec 95 when we hit -18 degrees. Local records show Ambleside not too far away dropped below -21 degrees in Jan 1940.

 

A more interesting question is how coastal locations on occasion can see very low temps, Crosby I believe as recently as Dec 2010 got down to -18 degrees I believe.. what caused such low temps, perhaps the sandy soil which I think can have some impacts.

 

This is a reasonable assessment, just to add that very obviously a cold, dry airmass and low sun inclination will assist.

Topography such as valleys and glens help cold air drainage and pooling, provides shelter to the surface air and reduced insolation.

The lowest temperatures though tend to be achieved when the deeper a temperature inversion can be created, and over as large an area as possible so the effects are less localised. The increased stability of the air reduces turbulent mixing and erosion of the cold pool at the surface. Topography will still tend to focus the lowest minima such as in the Severn valley in Shropshire at Newport and Shawbury, but the extreme lows are more widespread.

When this cold pool is sufficiently large and in sufficiently calm conditions, the contrast with the sea temperatures creates a land breeze blowing offshore - the reverse of a summer sea breeze - which negates the proximity to the water. This was the case for Crosby in 2010 which at sea-level also became a low point for air drainage, from a mainly flat rural area to the east, with a deep cover of snow. Sandy or dry soil and their temperatures has an effect but this is reduced the greater the depth of snow.

[damianslaw]

This is a reasonable assessment, just to add that very obviously a cold, dry airmass and low sun inclination will assist.

Topography such as valleys and glens help cold air drainage and pooling, provides shelter to the surface air and reduced insolation.

The lowest temperatures though tend to be achieved when the deeper a temperature inversion can be created, and over as large an area as possible so the effects are less localised. The increased stability of the air reduces turbulent mixing and erosion of the cold pool at the surface. Topography will still tend to focus the lowest minima such as in the Severn valley in Shropshire at Newport and Shawbury, but the extreme lows are more widespread.

When this cold pool is sufficiently large and in sufficiently calm conditions, the contrast with the sea temperatures creates a land breeze blowing offshore - the reverse of a summer sea breeze - which negates the proximity to the water. This was the case for Crosby in 2010 which at sea-level also became a low point for air drainage, from a mainly flat rural area to the east, with a deep cover of snow. Sandy or dry soil and their temperatures has an effect but this is reduced the greater the depth of snow.

 

 

Thanks for the response, very interesting to note how the coastal location helped create a 'low point' for the dense cold air to collect... Yes a deep cold pool, dense wide area of cold is needed as well. In Dec 2010 we had just that, and over a very wide area thanks to high pressure overhead.

[hillbilly]

Thanks for the response, very interesting to note how the coastal location helped create a 'low point' for the dense cold air to collect... Yes a deep cold pool, dense wide area of cold is needed as well. In Dec 2010 we had just that, and over a very wide area thanks to high pressure overhead.

Both early 2010 and the very cold December gave a mostly nationwide spread of bitter cold but was very cold stagnant air derived mainly from the north.Although severe cold in the north southern Britain was way off what it can produce in Easterly flows that have reached -22 to -23 in the south.

[JacobWilliams]

Would I be right to think then that most locations in the UK just don't have the right topography to record temperatures below -15 to -20 under any reasonably achievable synoptics? 

[Interitus]

Would I be right to think then that most locations in the UK just don't have the right topography to record temperatures below -15 to -20 under any reasonably achievable synoptics? 

 

No, I think most places could achieve -15 to -20 in the right circumstances with some exceptions, firstly those with the most maritime exposure. The Cornish peninsula would always struggle to build a large cold pool with proximity to the sea from most directions.

In any case, coastal sites need the conditions to be just right. With the example of Crosby, 2010 provided three cold spells with that of late December producing an impressive series of minima to -17.6 including also -16.4, -14.2 and -13.5. These compare well with inland Woodford which managed -14.8°C.

However, this must be considered an exception. In the earlier November/December spell Woodford reached -14.7 while Crosby was -7.8, and in January with Woodford down to -17.6 and -16.4 Crosby managed -10.5 which is more the normal situation.

As soon as an offshore flow is reversed the temperature can very quickly recover. For example at Shoreham in Sussex on 3/12/10 the temperature at 7pm was -9.7° and -7.4 at 8pm but an hour later at 9pm with the wind switching from the north to the south off the Channel, the temperature had jumped 11 degrees to 3.6°C.

 

The other exceptions are hilly and upland areas where the cold air can drain away or which lie above the lowest temperatures in an inversion. The Leek Thorncliffe station lies at around 300 metres in an upland valley on the edge of the Peak District which might be expected to achieve low temperatures. Whilst average temperatures in winter and year round are colder than lower elevations, the site struggles for extremely low temperatures with January and the two December 2010 spells reaching -8.6, -7.9 and -7.3°C respectively.

 

Finally, the urban heat islands of large towns and cities will have a reduced likelihood to a varying extent, with eg Manchester city centre often six degrees or more warmer than Manchester airport or Woodford.

Edited April 19, 2015 by Interitus

[SP1986]

Indeed, if you're coastal and on an incline you may never record low temperatures like those inland. Crosby was favoured because it is on land which is flat and sandy - also worth noting its station is actually more or less on the beach, so its not like it is set back. The temperatures back in 2010 were pretty uniform across the area, with Southport recording -15C (from a similar situation), and Prestatyn in Wales, also near sand dunes, recording -11C.. but then you'd get to Talacre (on the corner just to the east of Prestatyn, and it only recorded around -4C. Here on the hill, that night it was very interesting, on Christmas Eve 2010, we broke our temperature record, but being in relatively high elevation, compared to those around us, we, at one point became the warmest in the country. 

 

The temperature around midnight was -7C here and in most places on the Wirral Peninsula.. very quick here it rose to -3.9C by 3am in the morning (probably due to the cold draining off the hill), then when I woke up at 7am, it was down to -10.8C which is the record low for here.. astonishing drop, and never seen before.. the lowest temperature before that was -10.2C in December 1981. The sharp drop may have been cause by the ebbing and flowing of the inversion. Yet places lower down in our town only recorded -6C - maybe it was a thin inversion layer at a specific level.

[Wildswimmer Pete]

Going from memory, I'm sure the lowest lowland temps occured at Shawbury on the Shropshire plain.  Again if I remember correctly the lowest overnight mins were around -26C shared with a Highland location (Braemar?)

[hillbilly]

I did record a few days at -12deg and one -13deg at my old address at 340m in December 2010 and also remember recording very many days at a similar -12 deg in December 1981.I do not know what readings to expect at higher altitudes but Tm in Buxton got nothing like these readings at his 330m so topography has a lot to play!

[Summer of 95]

This is a reasonable assessment, just to add that very obviously a cold, dry airmass and low sun inclination will assist.

Topography such as valleys and glens help cold air drainage and pooling, provides shelter to the surface air and reduced insolation.

The lowest temperatures though tend to be achieved when the deeper a temperature inversion can be created, and over as large an area as possible so the effects are less localised. The increased stability of the air reduces turbulent mixing and erosion of the cold pool at the surface. Topography will still tend to focus the lowest minima such as in the Severn valley in Shropshire at Newport and Shawbury, but the extreme lows are more widespread.

When this cold pool is sufficiently large and in sufficiently calm conditions, the contrast with the sea temperatures creates a land breeze blowing offshore - the reverse of a summer sea breeze - which negates the proximity to the water. This was the case for Crosby in 2010 which at sea-level also became a low point for air drainage, from a mainly flat rural area to the east, with a deep cover of snow. Sandy or dry soil and their temperatures has an effect but this is reduced the greater the depth of snow.

Newport and Shawbury are nowhere near the Severn- that flows SE from Shrewsbury towards Bridgnorth. Neither is particularly hilly; Shawbury is in a bit of a "bowl" but the hills are low (150m). I wonder what the minima were at Preston Montford (right by the Severn, but averages slightly higher minima than Shawbury) or Tern Hill (about 7 miles NE of Shawbury) or Cosford (10 miles S of Newport, very flat) those nights.

Shawbury seems to be a cold spot on summer days too, often 2C or so colder than Shrewsbury and Cosford, and 3-4C below Hereford.

[Polar Maritime]

Dropped to -18c here in Winter 2010. Topo plays a big part round here with warm/cold air getting trapped between the Valleys then pushed and sucked about by funnelling winds sweeping through and over. TM gave a very good explanation somewhere on the Forum.  

Edited April 20, 2015 by Polar Maritime

[JacobWilliams]

Interesting. There must be a much higher standard deviation for very low temperatures at almost all locations - I recall reading somewhere that even 300 years of records probably isn't enough to expect the far left hand tail of the distribution to have been reached in most locations. Does anyone know what the lowest daily minimum CET reading was, and whether the CET stations are topographically appropriate for very low readings?

[Interitus]

Newport and Shawbury are nowhere near the Severn- that flows SE from Shrewsbury towards Bridgnorth.

 

That is true, they are not directly on the river but they are in part of the Severn drainage basin on the north Shropshire plain.

[Summer of 95]

That is true, they are not directly on the river but they are in part of the Severn drainage basin on the north Shropshire plain.

Shawbury is (the Tern and Roden is a tributary of the Severn) but I'm not sure about Newport. It must be close to the Severn/Trent watershed.

[Interitus]

Shawbury is (the Tern and Roden is a tributary of the Severn) but I'm not sure about Newport. It must be close to the Severn/Trent watershed.

 

Newport is also in the Shropshire Middle Severn catchment area, drained by the Meese and Strine, as can be seen on the Environment Agency website - http://environment.data.gov.uk/catchment-planning/ManagementCatchment/61. The Severn and its numerous tributaries in this area form an almost flat or only gently undulating basin.

The Met Office data actually comes from a couple of miles to the northwest at Harper Adams University which at 64 metres asl is officially 8 metres lower than the 72 metres of Shawbury airfield.

[damianslaw]

Braemar and Altnaharra 2 spots which have recorded lowest ever min in the UK at a bone biting -27.2 degrees. Braemar however, is a fair bit higher than Altnaharra but both lie within hilly terrain some distance from the moderating effects of the Sea.

 

Cold dense air can sit within these places producing freezing fog and very very supressed maxima, I think Braemar has reported the lowest ever max at about -15 degrees.

 

The Lake District is a very good place to experience the effects of local topography and proximity to the sea have on the weather and temperatures. In winter there can easily be a 10 degree drop in temperatures from coastal spots such as Millom, Seascale and Barrow in Furness to spots such as Ambleside, Windermere and Grasmere. Another phenomenon here is the dreaded cloud bank that tends to unveil the central lakes in mid summer whilst the coast basks in sunshine - often the day can start sunny here but by late morning the clouds start to develop over the fells, creating a blanket by afternoon but then dispersing in the evening.