Two remarkable British summers – ‘perfect’ 1911 and ‘calamitous’ 1912



Recent work by the Met Office National Climate Information Centre (NCIC) to extend the monthly temperature and rainfall series for the UK from 1914 back to 1910 drew two remarkable British summers to our attention.

The ‘perfect’ summer1 of 1911 was one of the warmest on record. It included the driest July over England and Wales in the last 100 years, an exceptionally sunny July and one of the warmest Augusts. In marked contrast, the ‘calamitous season’2 of summer 1912 was the wettest in over 200 years and included the coolest and wettest August on record.

The promising early start to summer 1911 – May was warm and sunny – and forthcoming coronation of George V created, briefly, a national air of optimism. The new king was ruler of over 400 million subjects across the Empire, and Britain held a leading place in science, engineering, exploration and literature.

A rigid, class-based society was still in place, with 1% of the population owning 60% of the wealth, and there was real poverty in places such as London's East End. However, the long-established status quo was being challenged, sometimes in violent ways – for example by suffragettes, Irish nationalists and sections of the work force. The working classes were mostly involved in manual labour – manufacturing, farming, mining, transport and domestic service – often with poor pay and lack of job security. During summer 1911, resentment spilled onto the streets with rioting, and strikes by trade unions representing miners, dockers and factory workers. The heat and drought of summer 1911 and the floods and cold of summer 1912 are likely to have exacerbated the social and political tensions of the time.

Data sources

The NCIC maintains monthly gridded datasets covering the UK at 5km × 5km resolution for 36 climate variables, mostly from 1961. The method used to generate these is described in Perry and Hollis (2005). The datasets are used to create (i) areal values for the UK, England, Wales, Scotland, Northern Ireland and regions and (ii) colour-shaded maps, and they form a key part of the UK climate record, available at

An important aspect of these datasets is the series' length. Those for rainfall and air temperature have been extended back, first to 1914 and recently to 1910, following digitisation of published station data, thus drawing our attention to the interesting summers of 1911 and 1912.

The gridded datasets provide a useful resource for the last 100 years, but their quality is highly dependent on the density of the station observations. In 1911, the Meteorological Office and the British Rainfall Organization administered UK networks of approximately 200 climate stations and 4800 rainfall stations respectively.3 The 1911 editions of the Monthly Weather Report (Meteorological Office, 1911) and British Rainfall (British Rainfall Organization, 1912) provide monthly temperature data for almost 200 sites and monthly rainfall for about 550 sites, and these were used for gridding. Going further back from 1910 the numbers of published stations diminish, so maintaining the homogeneity of the gridded series would become increasingly difficult.

For a longer-term perspective, we have therefore also made comparisons with the monthly Central England Temperature (CET) series from 1659 (Manley, 1974) and the England and Wales Precipitation (EWP) series from 1766 (Wigley et al., 1984). These are the longest instrumental series of their type in the world.

Summer of 1911

Summer 1911 was warm, dry and sunny for many weeks, particularly in the south, with high pressure near or over the UK. The warmest spells were in early June, late July to mid-August and early September. This was a prolonged dry period, with much of England having less than 70% of average rainfall from April to September. The drought through July and August was particularly severe because it coincided with the greatest heat. Despite the heat, there were relatively few thunderstorms. Summer 1911 rainfall is shown in Figure 1.

Figure 1.

Summer 1911 rainfall (% of 1961–1990 average). 1961–1990 is the earliest 30-year standard period for which gridded averages are available.

May 1911 was warm, sunny and very dry – although the month ended with violent thunderstorms across southeast England, including the Epsom area of Surrey immediately after the Derby Day races (Jebson, 2011). The fine weather resumed for the first half of June, but the second half was wet and very unsettled – disrupting the tennis championships at Wimbledon – and June overall was an unexceptional month. The most notable feature was a very heavy rainfall event on the 23rd and 24th, mainly near the east coast of England, described in British Rainfall as the most remarkable fall of the year. The rain began in London at 1900 gmt on 23rd just after the royal procession of King George V and Queen Mary – the day after their coronation. Over 75mm fell in these two days across parts of northeast England.

July was an exceptionally warm, dry and sunny month. High pressure dominated the weather during the first half and, although Atlantic depressions affected the north and west in the second half, the south and east remained under the influence of high pressure from the near continent. Many places here had no rainfall for three weeks or more, although there were thunderstorms towards the end of July in the London area – one giving 28mm of rain in 15 minutes at South Kensington on the 28th (Meteorological Office, 1911). Much of central and southern England had less than 20% of average rainfall and an area around Bath, Somerset, was completely dry for the whole month. The mean England and Wales total of 14.3mm makes this the driest July in the series from 1910. In the EWP series from 1766 it is the third driest (15.8mm): only the Julys of 1800 and 1825 were drier, each with less than 10mm.

The warmth of July 1911 was particularly pronounced across the Midlands, southern and southeast England, with the temperature frequently exceeding 27°C and reaching 33–34°C in the London area on 22nd. In the gridded mean temperature series for England and Wales, July 1911 ranks equal fourth (17.6°C), that in 2006 being the warmest July (19.1°C). At Eastbourne, Sussex, 383.9 hours of sunshine were recorded – an average of 12.4 hours per day. This remains the highest monthly total ever recorded in the UK.4 A further 18 stations along the south coast recorded more than 360 hours (Meteorological Office, 1911). The grid-based sunshine series extends back to 1929. In this, the sunniest July in England and Wales was that of 2006 (Prior and Beswick, 2007), when the highest station total was 340.4 hours at Shanklin, Isle of Wight.

The main heat-wave occurred in the second week of August. On the 8th, high pressure over northern Germany fed a southeasterly flow from the very hot near continent. On the 9th, the temperature exceeded 35°C over a wide area of England, with 98°F (36.7°C) recorded at Raunds, Northamptonshire, and also at Canterbury, Kent and Epsom, Surrey5. This record stood for nearly 80 years, a remarkably long time – even more so given the increasing number of observations during the twentieth century. It was finally broken on 3 August 1990 when 37.1°C was recorded at Cheltenham, Gloucestershire (Brugge, 1991; Burt, 1992)6. A temperature of 100°F (37.8°C) was recorded at Greenwich, London, on 9 August 1911, the highest recorded there since observations began in 1841. However, this was from a Glaisher stand, not a Stevenson screen, and it is not accepted as an official record (Laing, 1977).7

In the gridded mean temperature series for England and Wales, August 1911 ranks equal third (17.9°C), that in 1995 being the warmest August (18.4°C). Many places in southern and eastern England had no rain for 18 consecutive days during August and the hot, dry weather continued into early September, with Raunds recording 94°F (34.4°C) on the 8th. After the 12th the weather became cool and changeable, and the long, hot summer was at an end.

The meteorological conditions required for very high temperatures in the UK are set out by Brugge (1991). All the key ingredients were present in 1911:

  • Timing – early August, when land and sea surface temperatures were near their annual maxima.

  • Anticyclonic conditions, with clear skies and a southeasterly flow drawing air off the hot continent over a short sea-track.

  • Very dry ground – little surface moisture to evaporate, so solar radiation was readily converted to sensible heat.

In the CET series from 1659, summer 1911 is the equal eighth warmest summer, and at the time was the warmest summer since 1846 (Table 1). Summer 1976 is the warmest in this series but notably those of 1995, 2003 and 2006 are also within the top five, consistent with the warming trend of recent decades.

Table 1. The 11 hottest summers (June, July and August) from the monthly Central England Temperature (CET) series from 1659.
SummerMean CET (°C)


Harding (1912) and Nicolson (2006) noted many impacts from the hot, dry summer.

  • Death rate

    There was a steady increase in death rate through the summer as the heat intensified and, by 17 July, The Times began a regular column on Deaths from Heat. In London, there were 17 400 deaths during July, August and September 1911 compared with 13 300 for the same three months in 1910. Infant deaths accounted for most of this increase; in August alone there were over 2000 from diarrhoea associated with rotted food and bad milk. The technology that we now rely upon during heat-waves, such as refrigeration and air-conditioning, was not available to most households a century ago.

  • Drought

    Water supplies were cut off for several hours each day in cities such as Manchester and Bradford. Reservoir levels in the Peak District were very low, with concerns about water supplies to canals (still used then for transporting goods). The dry conditions also led to problems of settlement of many houses in the London area built on clay.

  • Agriculture

    In the parched English landscape, good grazing for sheep and cattle became scarce and farmers struggled with water and feed for livestock, causing the price of milk to rise. Farm workers bringing in the harvest found the heat overpowering but, generally, conditions favoured an early harvest.

  • Civil unrest

    Widespread unrest amongst the work force, owing to dissatisfaction with wages and working conditions, spilled into riots and strikes involving over a million trade union members. The hot weather exacerbated growing anger as everyday life became ever more difficult. Factory and mill workers were laid off due to the water shortages and working hours were altered to avoid the afternoon heat. A dockworkers strike left food to rot at ports, increasing the threat of shortages.

  • Positive effects

    The fine weather resulted in a very large increase in the number of visits to seaside resorts such as Blackpool and Brighton, particularly during the bank-holiday weekend of Saturday 5th to Monday 7th August: ‘bathing machines’ were very popular. Lions, cheetahs, rhinoceros and giraffes at London Zoo became unusually active. Dry, flat pitches made it an excellent year for cricket batsmen.

Summer of 1912

The following summer was almost the exact opposite of summer 1911 – notably wet, cool and gloomy. Over twice the average amount of rain fell in a broad swathe from Cornwall to Norfolk, with more than 250% in places (Figures 2 and 3). This was the wettest summer for at least the previous 50 years across England and Wales (Salter, 1913), although it was only marginally wetter than that of 1879. It stands as the wettest summer in the EWP series from 1766 (Table 2). England and Wales received 197% of the 1961–1990 average rainfall, significantly more than in the wet summer of 2007 (169% of average) (Figure 4), although the wettest areas were very similar (Mayes, 2008).

Figure 2.

Summer 1912 rainfall (% of 1875–1909 average) reproduced from British Rainfall, 1912.

Figure 3.

Summer 1912 rainfall (% of 1961–1990 average) from NCIC gridded data (the legend scale is identical to Figure2but differs from Figure1). The wettest areas in 1912 are in close agreement (note the different averaging period).

Figure 4.

Summer 2007 rainfall (% of 1961–1990 average).

Table 2. The 10 wettest summers (June, July and August) from the monthly England and Wales Precipitation (EWP) series from 1766.
SummerTotal rainfall (mm)
  1. For comparison, the values from NCIC gridded data are 400.3mm for summer 1912 and 342.5mm for summer 2007.


June was very unsettled with a series of shallow, slow-moving Atlantic depressions moving eastwards or northwards and a complete absence of anticyclones. The sole exception was a brief spell of warm weather in the third week east of a line from The Wash to the Isle of Wight. However, winds were mostly light. Rain was abnormally frequent with an unusually large number of thunderstorms. Snow fell on the upper levels of Snowdon on the 4th. The 10th saw torrential rain and flooding in Northern Ireland, whilst an observer at Roden, Shropshire, reported a very heavy snow (sic) and hail storm doing immense damage to crops. Most stations recorded well-below average sunshine for the month (only 43% at Stonyhurst, Lancashire, for example).

July was the ‘least bad’ summer month. There were two spells of fine, warm weather in the first half. During the second spell (from 12th to 17th) temperatures rose above 30°C widely. However, the month was otherwise unsettled, dull and cool. From the 23rd to 29th the weather became very wet and thundery with low pressure dominant and this type of weather persisted into August.

August was characterised by a relentless succession of Atlantic depressions, which would be considered unusual, even for a winter month (Meteorological Office, 1912). There were frequent southwest to westerly gales along the south coast and a remarkable absence of summer warmth, with the highest temperature of the month only 73°F (22.8°C) on the 4th at Greenwich and Camden Square, London, and Welshpool, Powys. On some days temperatures widely failed to rise above 60°F (15.6°C).

The most significant rainfall event of the summer was the Norfolk rainstorm of 25–26 August. A depression moving northeast intensified as it crossed the Thames Estuary, with 180 square miles of East Anglia from the Broads to Wymondham (Norfolk) receiving more than 175mm of rain in 48 hours. The 160.3mm that fell at Norwich in the 24 hours to 0900 gmt on 26 August was reported at the time as the largest amount on record for a rainfall day in eastern England (Meteorological Office, 1912).

The monthly mean CET was 12.9°C, making this the coldest August in the series from 1659. The August 1912 mean temperature in England and Wales was more than 5 degC lower than that of August 1911. This was also the wettest August in the EWP series from 1766. The extreme contrast between July 1911 and August 1912 rainfall is shown in Figure 5(a) and (b) – with the latter recording over ten times more rainfall across England and Wales than the former (Table 3). A large number of stations received less than half the average sunshine.

Figure 5.

Two remarkably contrasting months: the driest July (1911) and the wettest August (1912) in England and Wales in the last 100 years.

Table 3. Rainfall of July 1911 and August 1912 compared (from NCIC gridded data).
MonthEngland and Wales rainfall (mm)1961–1990 average (mm)% of average
July 191114.362.323
August 1912182.176.7237

On 25 September 1912, W. N. Shaw of the Meteorological Office noted several correspondents describing a peculiar persistent whitish haze which has covered the sky even on apparently cloudless days during the past summer. This phenomenon was reported across Europe, including the UK, Germany, Switzerland and Sweden. For example, W. D. Addison in Riga (on the Baltic coast) wrote: [during the summer] never once have we beheld a sky which would be reasonable to term blue. A ‘veil’ seems to have been imposed between us and our sky. Shaw commented: All observers are unanimous in regarding the effects as taking place at very high levels, well above the lower cloud level (Shaw, 1912).

Unknown to these observers, this ‘veiled’ appearance was due to fine volcanic ash which had been projected high into the stratosphere by the major eruption of the volcano Mount Katmai in Alaska between 6 and 9 June 1912. This ash would have remained in the stratosphere for several months, too high for any rain to remove it.8 Volcanic dust from this eruption has subsequently been found in ice-cores drilled in the Greenland ice-cap. (More information from the Alaska Volcano Observatory There may have been a connection between this eruption and the poor British summer.


The wet summer of 1912 was particularly difficult for agriculture, following on from the heat and drought of summer 1911. The two rainiest periods in June and August coincided with the hay and grain harvests in England, Wales and Ireland and in many areas the effect was disastrous: crops were rotting in the fields and those that had been cut had not been gathered to dry. For example, The Times of 27 August 1912 reported wheat and barley crops in Suffolk to be irretrievably ruined.

The rainstorm of 25–26 August caused widespread damage, not just to crops. Bridges were destroyed and railways across East Anglia and the east Midlands were flooded. There was some loss of life (Meteorological Office, 1912). The low-lying part of Norwich was worst affected; a flood-water mark from August 1912 was 38cm higher than the next mark, bearing a date of 1614. In the Fens between Spalding and Peterborough, floodwaters formed a lake 20 miles long and a mile wide in places; a large area of land remained underwater throughout the following winter (British Rainfall, 1913). The water level of the Norfolk Broads rose several feet, but these fortunately stored the floodwater and prevented much worse inundation elsewhere (for example, in Great Yarmouth).


The contrasting summers of 1911 and 1912 were remarkable not only for their individual characteristics but the fact that they occurred in successive years.

The hot, dry summer of 1911 saw the third driest July in England and Wales in over 200 years and the UK's highest monthly sunshine total on record. During the August heat-wave, a UK temperature record was set which stood for almost 80 years.

The cool, wet summer of 1912 was significantly wetter even than summer 2007. This was the wettest summer in over 200 years for England and Wales, with only summer 1879 coming close. August 1912 was the coldest August in the CET series from 1659 and the wettest in the EWP series from 1766.

A century ago, life in the UK was very different from that of today, so the weather of these summers would have had different impacts. Technology was, of course, far less advanced and this was reflected in domestic life, health-care, agriculture, transport and mass communication – including the production and dissemination of weather forecasts. Working conditions were also very different, although a National Insurance Act had been proposed to provide some medical and unemployment protection. So in many ways people are likely to have been more susceptible to the impacts of extreme weather. The agricultural sector probably suffered most, being affected by both abnormal temperature and rainfall – and without the benefit of modern mechanisation (Figure 6). However, society was less complex in 1911: expectations were lower, people were more self-sufficient and journeys were fewer and shorter. Perhaps most importantly, the UK population in 1911 was only around two-thirds of today's. These factors imply that extreme weather had less impact 100 years ago. It would be a difficult, but thought-provoking, exercise to assess the net overall difference in impacts between then and now.

Figure 6.

The harvest in 1911 (threshing oats) – a mix of man-, horse-, and steam-power. (Courtesy the Museum of English Rural Life, University of Reading,

It is possible that the very poor weather of summer 1912 was influenced by a volcanic eruption in Alaska. Nevertheless, these summers emphasise the large variability of the UK's climate and the fact that the weather of one year can bear little or no relation to that of the next.

The summers of 1911 and 1912 serve as a reminder that extreme weather events occurred in the past just as they do today – there are parallels with the record-breaking summers of 2006 and 2007, for example. They also aroused the same scientific interest: in discussion following Harding's summer 1911 paper (Harding, 1912), Mr Gold said Mr Harding had given an excellent account of the summerbut that the ordinary person was even more anxious to knowwhyit had been hot thanhowhot it had been. The exceptional summer must be due either to an exceptional but accidental course of the circulation of the atmosphere arising from terrestrial causes, or to a change in the intensity of solar radiation. 100 years on, such questions remain central to climate science.


Thanks are due to Margaret Elms, formerly of the Met Office, who carried out the 1910–1913 digitisation work, and to David Parker of the Met Office for identifying the link to the Mount Katmai volcanic eruption.

The authors are also grateful to the staff of the Museum of English Rural Life, University of Reading for help with researching agriculture in the early twentieth century. See

Information on the climate of the UK, including monthly, seasonal and annual summaries and statistics, is compiled by the Met Office NCIC and available at

  • 1

    Only ‘perfect’ for some. Nicolson (2006) acknowledges that for many people the summer heat made life more difficult.

  • 2

    The ‘calamitous season’ appeared as a headline of an article in The Times of 27 August 1912, recalling the many impacts of the poor weather.

  • 3

    The figures refer to the current United Kingdom of Great Britain and Northern Ireland. In 1911 and 1912 the UK comprised Great Britain and the whole of Ireland.

  • 4

    Discussion following a paper describing summer 1911 by Harding (1912) states: Mr Brodie had mentioned there had been an increase in the sunshine recorded during the last few years…this might be accounted for by the fact that the cards now used [in Campbell-Stokes sunshine recorders] are more porous and ready to burn than formerly, and so a larger amount of sunshine is recorded. It was, therefore, doubtful whether the records of the earlier years and those of the present would be strictly comparable. Today's scientists working with observational data will identify with this type of issue!

  • 5

    Temperatures were published to the nearest whole degree Fahrenheit. Burt (1992) expressed reservations about the Epsom and Canterbury readings.

  • 6

    The current official record is 38.5°C at Faversham, Kent, on 10 August 2003.

  • 7

    The Glaisher stand was brought into use at Greenwich Observatory in 1841. It comprised a vertical board on which thermometers could be mounted, which could be rotated around a central pivot to face away from the sun. It was reliant on the observer turning the stand regularly. Laing (1977) compared temperatures from Glaisher stands and Stevenson screens and found differences of up to several degrees Celsius on warm sunny days due to the Glaisher stand's design. The Stevenson screen was invented in 1863.

  • 8

    In summer 1912, aviation was still at a very early stage (Bleriot flew an aircraft over the English Channel for the first time only three years earlier). The volcanic ash from Katmai thus caused no travel disruption, in contrast to that from Ejafjallajökull in Iceland in April 2010, which grounded aircraft across Europe.