A blog post by Mark McCarthy (NCIC, Met Office) - 21st February 2017

Extending our historical weather and climate records

In the late 17th and early 18th centuries the philosopher John Locke provided his meticulous weather observations and thermoscope (a predecessor to the modern thermometer) measurements to the Royal Society. Accompanying his return for 1692 he said:

“I have often thought that if such a register as this, or one that were better contrived, ..., were kept in every County in England, and so constantly published, many things relating to the Air, Winds, Health, Fruitfulness etc. might by a sagacious man be collected from them, and several rules and observations concerning the extent of winds and rains etc. be in time established, to the great advantage of mankind.”

It took until the late 19th Century before John Locke’s vision was fully realised, but his measurements are still very much used today to help us do exactly as he imagined, and they form part of the long-running Central England Temperature series. The National Climate Information Centre (NCIC) at the Met Office retains UK climate datasets that span, in some cases, over 350 years of instrumental measurements. John Locke’s records demonstrate that the value of these historical measurements does not diminish with age, in fact quite the reverse. Climatologists in the UK are very fortunate because we have such a long and established history of both talking about and observing our weather and climate and consequently we have an archive of meteorological observations spanning centuries, much of which is stored in the National Meteorological Archive in Exeter. The digital age is really relatively new in comparison and turning centuries’ worth of careful handwritten observations into something that can be read and interpreted by a computer is not a task to undertake lightly. This means that even though we do have millions of historical weather records that have been digitised for the UK, it is still only a fraction of our full paper archive.

As part of the Historic Drought project the Met Office has been working to digitise more UK rainfall and temperature data. These data will then form part of our digital memory of weather and climate, and importantly can help add to our knowledge and understanding of drought and other climate phenomena. With such a mountain of potential data it has been important to be incredibly focussed. We have the best data coverage back to the late 1950s, and quite a few stations covering as back as 1910 (although we still have plenty of gaps to fill), so this project has looked in particular at the period 1862 to 1909. The late 19th Century and early 20th Century is a period containing a number of very notable hydro-meteorological extremes, such as a severe drought during 1887 and early 1888, and the long-drought of 1890-1910, which was itself interspersed with some exceptional wet episodes. The long-running England and Wales Precipitation series puts October 1903 as the wettest, and February 1891 as the driest calendar months in a series from 1766. In 250 years of observations the wettest and driest months are separated by just 12½ years. The period also contains the wettest year in the series (1872) and the third driest (1887). It was also an era of rapid developments in meteorological observations, and most notably the pioneering work of George Symons who established the British Rainfall Organisation. Symons succeeded in not only collecting and publishing rainfall measurements from amateur observers across the UK and Ireland, but also recognising the vital importance of standardising how these measurements were made. The British Rainfall publication which ran from 1862 to 1991 collated these rainfall measurements, providing a wealth of historical observations some more of which have now been recovered (see below). We can use the digitised data to gain more detailed insights into the characteristics of some of these historical climate extremes to allow us not only to achieve the aims of the Historic Drought project, but also to provide the nation with longer, more detailed, and more accessible historical climate datasets tailored for our digital age. 

British Rainfall

Excerpt from British Rainfall from 1891

Rainfall October 1903So what do the UK’s wettest and driest months look like? October 1903 saw a rapid succession of Atlantic depressions affect the British Isles. Many parts of the country saw rain on nearly all days of the month, with some intense downpours and thunderstorms causing extensive damage. Flooding was widespread and many parts of the country experienced disruption. Using established NCIC techniques to map rainfall from station observations, the new data allow us to paint a much more detailed picture about this extreme month to show where the rain actually fell, and we can compare this with extremes of more recent times. For example Cumbria saw the highest rainfall totals during October 1903 with 600 mm at Seathwaite, which in itself is not nearly as extreme as December 2015 for which Seathwaite recorded nearly a metre of rain (967 mm). However, heavy and persistent rain in October 1903 was much more widespread across southern and central England than it was during December 2015. On the basis of the UK average rainfall October 1903 is considered a wetter month than December 2015, however in terms of regional extremes such as those experienced across north-western areas including Cumbria, then December 2015 is still an exceptional record-breaking month. The historical records also include much more information from the time, including the observer’s notes. These provide some real insights into the impacts from this extreme weather over a hundred years ago. For some the consequences were devastating with widespread flooding and ruined harvests. For example the observer at Altarnon, Cornwall in 1903 said:

“All crops were in good bulk, but when the dry period came in November, favouring the gathering of roots and apples, much of the grain had rotted in the ground.”

Rainfall February 1891In contrast February 1891 was extremely dry and the driest calendar month in our records. February has a slight advantage here being our shortest month, but even taking this into consideration it really was a notable month. Many rain gauges recorded very little rain or even no rain at all. The settled conditions favoured frequent fog, and the totals reported from some rain gauges were in fact small amounts of water from the fog and heavy dew rather than rainfall. Outside of the fog however there was plenty of fine sunny weather. In London the observer in West Kensington celebrated this unusual occurrence:

“A rainless month has not occurred in London since July 1800! So we may congratulate ourselves on having seen what hardly anyone living has seen before.”

While it is of interest to contrast these two extreme months, the additional geographical detail they provide will now feed into hydrological reconstructions for the period as part of the wider Historic Droughts project, to provide more detailed insights into the character of these notable hydro-meteorological extremes of the past. Following final checks and research into the data quality, they will be adopted within our formal national climate archive where they will reside in perpetuity for the nation.