I have a second-hand copy of HH Lamb's book "The English Climate", published in 1964. This was the second, heavily revised, edition of CEP Brook's book of the same title published in 1954. Appendix 2 is a "Calendar of Historic Weather Events". I realised that with its help I could probably compile a short list of the coldest days in the London area since reliable thermometers became available in the early 18th century. By "coldest day", I mean having the lowest maximum temperature. So here they are in chronological order:

11 January (New Style), 1740. "Afternoon temperatures in Holland -20C, in London about -9C. E gale." If I correctly remember what Gordon Manley wrote about this day in his book "Climate and the British Scene", there were no outside thermometers in the London area at the time. They were still too expensive and fragile to be risked outside and were usually hung indoors in an unheated room, so of minimal use. The London temperature was estimated based on readings in Holland, where there were some outside thermometers. Given the strength of the wind that -9C looks as if it's a fairly conservative estimate.

20 January, 1838. "Lowest [minimum] temperatures of 19th century in London: -16C reported at Greenwich about sunrise, -20C at Blackheath, -26C at Beckenham. Temperature in Greenwich -11C at midday." It seems likely that the day's maximum would have been several degrees higher than the midday value. With winds probably light and skies clear, judging by the low minima, the fact that proper screens for thermometers were yet to be in use may have depressed the values obtained.

14 December, 1890. "Coldest known December day in London, temperature never rose above -6C."

19 January, 1963. "Coldest day [of that winter]: afternoon temperatures of -7C in Surrey with E gale." [Note: Further research suggests that either the value or the date is likely to be in error. The coldest days, when a -7C value somewhere in Surrey would be plausible, were 23 and 24 Jan.]

12 January, 1987. Maxima between -6C and -8C over much of England, with -9.1C at Warlingham.

From this list, it looks as though the very coldest days mostly occur around the middle of January, and that the lowest possible maximum in SE England is close to -10C. That 1987 value at Warlingham may be the lowest maximum for the London area that we can be certain is accurate, since the lower value in 1838 was a midday rather than a maximum temperature.

A useful compendium. Of note are that the sea froze significantly down here in both 1890 and 1963, out to a considerable distance. IIRC you need about -6 or -7 for the sea to start to freeze over, which ties in with those figures.

Though there was some freezing in 1987, it wasn't anywhere near as great as those other two events.

It'd be interesting to see what the maxima were in 1947, as that was a late winter. A bit like Feb 1986 in that regard - a month here of plentiful snow, temperatures generally below freezing by day, icicles, powder and all the rest. The echo of it in the latter half of Feb 2005 was less impressive, temperatures rose above freezing by a degree or two by day (drip, drip), but there was plentiful snow and of course it froze solid overnight, each night. There was snow on the ground for over 2 weeks from that one.

Interesting remember the 80s cold spells.  Jan 85 had a pretty severe spell too with sub zero max her on sussex coast and Feb 91 on one day it was only -5 max one day here . Overall the 12 Jan 87 was the coldest day I've ever known probably the harshest winter weather to hit the SE in living memory 

I'm still alive and was living in the SE in 1947 (Orpington), 1963 (CambrIdge) and 1987 (Maidstone). On the last occasion, I had to go and escort my wife home from the boarding school where she was teaching in deeply rural country near Maidstone in a blizzard. We walked for a mile-and-a-half in vile conditions, but I still think that it didn't compare to some days in 1963 (and in 1947 I was only five years old so was probably being kept inside in the worst of the weather). But 1987 in Kent at least didn't last more than a couple of weeks unlike the other winters.

Trevor Harley gives lowest lowest daytime maxes winter by winter since 1900; 

https://www.trevorharley.com/coldest-days-of-each-winter-from-1900.html 

and the lowest overnight minima since at least 1880 from Torro may be of interest;

https://www.torro.org.uk/extremes/date-records/min-temp 

Date records for lowest daytime max, annually by winter, if these occurred in the SE

-18.8 Swarraton (Hants.) 9 January 1901

-16.8 Woodbridge (Suffolk) 30 December 1906

-18.3 Liphook (Hants.) 30 December 1908

-20.0 Benson (Oxon) 6 February 1917

-17.2 Woburn (Beds.) 9 February 1919

-21.1 Elmstone (Kent) 30 January 1947

-13.3 Grendon Underwood (Bucks.) 12 December 1991

-14.5 Santon Downham (Suffolk) 3 January 1997

Date records for lowest minima if these occurred in the SE

19/1/1966 -18.9 Elmstone

23/1/1963 -20.6 Stanstead Abbotts, Herts (that day none of the rural buses into Cambridge ran if they'd been parked outside - the diesel had frozen!)

29/1/1947 -20.6 Writtle, Essex

30/1/1947 -21.3 Elmstone

6/2/1917   -20.0 Benson, Oxon

25/2/1947 -20.6 Woburn, Beds

10/3/1931 -15.0 Rickmansworth, Herts

https://www.torro.org.uk/extremes/date-records/min-temp 

Strangely your 3rd date of 19/1/1963 does not even figure in the top 100 in my list and is showing as -1.1C. I would like to know where in Surrey managed a max of -7C on that day as it can't have been that far away from my site. Makes me wonder whether one of our datasets is in error.

The top 15 in the list at Kew features 13 entries from between 1881 and 1895 so for "coldies" in London that must have been a great period. The only other date to appear in the top 15 apart from this range and 1987 was 02/02/1956 with a max of -4.3C

I do believe in January 1740 there were daytime maxes in East Anglia well below -10C. We had some impressive ones in January 1987 -8.3C at Stansted and -9C in Surrey.   🙂🙂

I imagine that the lack of substantial freezing of the sea in 1987 was because the cold spell was shorter than in 1890 or 1963. 

Interestingly, the coldest days at Oxford, which doesn't quite count as SE England, have been rather different:

-9.6C 8 Jan 1841

-7.9C 2 Feb 1830

-7.8C 19 Jan 1823

-7.4C 20 Jan 1838 (the one day in common with my original list)

-7.1C 15 and 19 Jan 1838, 13 Jan 1982

Again the values from the early 19th century have to be caveated with the thermometer screening possibly being inadequate or altogether lacking. The severity of the cold spell in January 1838 was clearly exceptional, though. The data comes from "Oxford Weather and Climate Since 1767" by Stephen Burt and Tim Burt (who remarkable are not related) published in 2019. It supports the ideas that the lowest possible maximum for lowland southern and central England is close to -10C and that the most likely time for the severest cold is around mid January.

Unfortunately, though Trevor heads his table "Coldest day of each winter from 1900" the data is actually for the lowest night-time minima rather than the lowest day-time maxima. But I did find: "The lowest lowland maximum since 1900 is -9.2 at Warlingham (Surrey) on 12 January 1987; Pershore saw a maximum of -8.2 on 19 December 2010, and Altnaharra -15.8 on 22 December 2010."

Thanks. That leaves me wondering if Lamb could have made a mistake with 19/1/1963, either regarding the precise date or - assuming originally hand-written log-books - maybe -1.1 could have been misread as -7.1 or vice versa.

Thanks. I can certainly believe that a sub-zero maximum below -10C would have been possibly in January 1740 in East Anglia - though as always there has to be a caveat about the quality - or even total absence - of screening of the thermometers at that time.

I'm sure there were some exceptionally low temperatures in the 1683-4 winter. The problem is the lack of accurate thermometers that early. The Wikipedia article on the CET series explains it quite well: "Although best efforts have been made by Manley and subsequent researchers to quality control the series, there are data problems in the early years, with some non-instrumental data used. These problems account for the lower precision to which the early monthly means were quoted by Manley. Parker et al. (1992) addressed this by not using data prior to 1772, since their daily series required more accurate data than did the original series of monthly means. Before 1722, instrumental records do not overlap and Manley used a non-instrumental series from Utrecht compiled by Labrijn (1945), to make the monthly central England temperature (CET) series complete."

To confirm that it looks like Lamb made a mistake regarding 19/1/1963, as it seems that the maximum at Gatwick on that day was -2C (to the nearest whole degree).

-6.8C 12 Jan 1987

-5.6C 24 Jan 1963

-5.0C 29 Dec 1908

-4.5C 23 Jan 1963

-4.5C 16 Jan 1985

The other dates in the top 10 were 13 Jan 1987, 1 Feb 1956, 7 Feb 1991, 28 Dec 1908 and 24 Jan 1945.

Those can't be maxima - must be minima.

I see that has already been mentioned. But regarding January 1963, it may not be an error. Trevor Harley says:

There was another notable blizzard on the 19-20th, particularly affecting the southeast, with widespread maxima of -5C in the south.

Hi Rob

it was a weekend 19th Saturday not a lot of snow about in the morning but a disturbance came up from the s/e, bitter se gale wind blowing snow i remember it was difficult to see what was falling or what was actually being drifted

severe cold had dug in europe one hell of a battle 

very unusual fax charts one low approached went into med another approached and made a complex area area to the s west 

very severe penetrating frost at night, on the Sunday we had a glaze of freezing rain on top.

fantastic times 

http://www.mikett.plus.com/winter-62-63-maps/january/19jan.jpg 

Look at that severe cold across Eastern Europe on Brian’s chart

https://www.theweatheroutlook.com/twodata/reanalysis.aspx 

You can see from these why those of us in East Anglia and Kent frequently go on about "the good old days" last century and you have to bear in mind there were plenty of other low temperatures and snowfalls that would be considered severe here now.

It's a shame Ron Button isn't around any more as he kept a daily diary from around the end of WW2. Whilst I have many of his old weather books his family retained (quite rightly) his hand written diaries.

Now that's what I call a Greenland High! Can you imagine the traffic on here if we ever saw that again! I doubt we've even seen it in F1 this century!

That was the occasion of the famed British Rail "the wrong type of snow" comment due to the salt being unable to thaw it.

I do remember that Thursday morning  like it was yesterday.  We did manage to get to school on the bus, only to find out the school had closed. Then we had to find our way back home. Fortunately for me, a dad of another boy who lived near me took me home in his car after said boy called him from school. It took hours but we made it.  And of course no mobile phones so my parents had no idea where I was until I knocked on the door.  They didnt seem at all concerned.   How different would things be nowadays!

Just up the road from me. Only later discovered that my location was the sweet spot in terms of snow depth for the entire country.  Also repeated in Feb 2009 and Dec 1 2010 from similar Thames streamers.   But those more recent snowfalls were not accompanied by the extreme low temps we got in 87 and 91

The point really is that uppers were never  that cold but just enough with the cooler seas at that time.

I guess if that set up existed today with warmer seas it would be mainly rain.

11th Dec 1981 is a good example of this

https://www.theweatheroutlook.com/twodata/reanalysis.aspx