SEARCH

SEARCH BY CITATION

Charles Higman Griffith (Figure 1) was a Hampshire clergyman who, by conducting a series of thorough instrument comparisons in the late 1860s and early 1870s, played a leading role in the adoption of the Stevenson screen and the five-inch ‘Snowdon’ copper raingauge as UK and Ireland standard climatological instrumentation. They remain standards to this day. With the British Empire at its height, British meteorological instruments and standards spread quickly around the globe. Neither instrument is perfect, but because their design and construction have changed little in more than a century, today's climate researchers have a wealth of relatively consistent long-period temperature and rainfall measurements from around the world. Without the comparative trials undertaken by Griffith in the garden of a north Hampshire rectory over 140 years ago, and the rationalisation and standardisation of instrument types that followed, the world's climatological records would be much more fragmented and inconsistent.

image

Figure 1. Reverend Charles Higman Griffith, photographed about 1879. This photograph was included in the ‘Symons Book’, and appears courtesy of the Royal Meteorological Society.

Download figure to PowerPoint

Charles Griffith was born in Tottenham (north London) on 10 August 1830. Not much is known about his early life, but he married Hannah Mockridge from Taunton (Somerset) in 1853 and lived for a time in Kent, where his eldest child, William, was born in 1855. By 1860 he was a member of the clergy and had moved as a curate to the rectory at Strathfield Turgiss (now Stratfield Turgis), part of the Duke of Wellington's estate which lies in north Hampshire midway between Reading and Basingstoke. We know he was there in 1860, because some of his entomological observations from that year survive, while he and his wife and their three young children are listed as living in the rectory in the April 1861 Census. In 1862, on the death of the previous incumbent, he became rector of All Saints Church in Stratfield Turgis, with the Duke of Wellington as his patron. He was already a keen naturalist and microscopist, and his diaries abound with detailed observations of plants, insects and natural phenomena, including weather events. He began rainfall records in 1862: these were first published in British Rainfall in 1865, and this was presumably the first time he corresponded with George Symons, the founder of the British Rainfall Organization or BRO (see the special issue of Weather, May 2010, for more on Symons and the BRO).

The raingauge trials at Stratfield Turgis

  1. Top of page
  2. The raingauge trials at Stratfield Turgis
  3. The thermometer screen trials
  4. Global impact
  5. Epilogue
  6. Acknowledgements
  7. References

The founding of the British Rainfall Organization in 1860 was a landmark event in British climatology. In its early years, there were almost as many different types of raingauges, exposures and observational methods as there were observers. As the BRO expanded, the need for guidance to observers on a ‘standard raingauge’ and observing protocol became pressing. In 1863 Symons encouraged and part-funded Colonel Michael Foster Ward, at Castle House in Calne (Wiltshire), to set up a series of gauges of different size, construction and materials, exposed at various heights above the ground, to examine their relative performance. A parallel trial was run at Framfield (Sussex) in 1866–1867 by Rev. T. E. Crallan, and the results from these early comparisons were published in British Rainfall in 1866 and 1867. When Col Ward moved to Switzerland in 1868, Griffith agreed to continue the trials at Stratfield Turgis. The gauges from Calne and Framfield were duly re-installed there, together with other models, such that by the end of 1868 the rectory grounds contained no less than 42 different daily-read raingauges, as shown in the famous Frontispiece to British Rainfall 1868 (Figure 2).

image

Figure 2. The grounds of Stratfield Turgis rectory in north Hampshire, showing the 42 experimental raingauges run by Charles Griffith. This is the Frontispiece to British Rainfall 1868.

Download figure to PowerPoint

All gauges were read daily at 9 a.m. during 1868, and monthly in 1869, with the results promptly written up in British Rainfall in 1868 and 1869. In 1870 the gauges from Stratfield Turgis were dismantled and relocated to a windier and more exposed site at Hawsker, near Whitby (Yorkshire), run by yet another clergyman, Rev. F. W. Stow. A parallel trial was also set up at Boston Reservoir, near Rotherham (Yorkshire), while various trials on different gauge types and heights were also carried out by Glaisher and Airy at the Greenwich Observatory (London).

The results from over a decade of field trials in differing climatic regimes led to the formal adoption in 1875 of the copper Snowdon ‘five-inch’ (127mm) diameter gauge as the BRO standard, mounted 1 foot (300mm) above ground, and read at 9 a.m. daily. The Snowdon gauge with its deep funnel and an accurately-turned brass rim was found to offer a fair compromise between the various factors influencing rainfall catch, although for many years the Meteorological Office and a few private observers continued to use eight-inch (203 mm) gauges. Some of the latter were still in daily use up to the 1980s, perhaps one or two still are, although the Meteorological Office finally adopted the BRO standard five-inch gauge during the 1920s – the change no doubt helped along by the assimilation of what remained of the BRO into the Meteorological Office in 1919.

As an aside, it has always puzzled me how Griffith managed to get round 42 raingauges at 9 a.m. Clearly if it was raining at the time, the daily totals between the first-read and the last-read gauge would differ, perhaps substantially. Following a presentation to my local history society in 2001, referring to the role of Rev. Charles Griffith and Stratfield Turgis (Stratfield Turgis Rectory lies less than 5km to the southeast of my home at Stratfield Mortimer, on the Hampshire/Berkshire border), one member suggested that Griffith may have run a boarding school in the rectory. A check of the 1871 Census returns shows that the only people registered as living at the property were Griffith and his wife, their five children (then aged from 9 to 16) and two domestic servants – a cook and a maid. Local parish records do confirm a ‘boarding and tutorial establishment for boys’ at the rectory run by Griffith, so perhaps, with five of his own children already living in the rectory at that time, the pupils boarded locally in 1871. The 1881 Census does show five teenage ‘scholars’ living in the rectory with Charles and Hannah Griffith and their two remaining unmarried daughters. The mental image of a horde of small boys running around to the 42 gauges, wrapping up the rainfall observations within minutes, has great appeal, even if there is no unambiguous evidence for it!

The thermometer screen trials

  1. Top of page
  2. The raingauge trials at Stratfield Turgis
  3. The thermometer screen trials
  4. Global impact
  5. Epilogue
  6. Acknowledgements
  7. References

Griffith commenced measurements of air temperature around 1864 or 1865 but, unlike the rainfall measurements which were preserved in the archives of the British Rainfall Organization, only a few fragments of these early records remain. At that time, types and patterns of thermometer housing were almost as numerous as raingauge designs, and in fact Griffith used his own design of screen for these early records. The Glaisher screen, an open stand first brought into use at the Greenwich Observatory in 1841, was probably the most widely-copied model, but it was known to suffer from warming owing to the effects of unshielded infrared radiation from the ground and sky on the thermometers. Thomas Stevenson's short note on a proposed design for a thermometer shelter which used wooden louvres to screen the interior instruments from solar and infrared radiation and precipitation appeared in the Journal of the Scottish Meteorological Society only in 1866 (Vol. 1, p 122) and at the time was only one of a multitude of competing designs. Clearly not one to pass up a challenge, Griffith also offered to run comparative trials of ten types of thermometer exposure in the grounds of the rectory, including one constructed to Stevenson's design, using identical calibrated thermometers (Figure 3). Equipment costs (£45) were met by a grant from the Royal Society procured by George Symons, while ‘Mr Casella’ of the instrument firm of the same name kindly lent a large number of excellent thermometers, carefully calibrated at Kew Observatory beforehand. It is interesting that one of the trial sensors was an aspirated thermometer, operated by a water-driven piston. Despite their many advantages, it is only within the last decade or so that aspirated sensors have become practical for widespread field use in climatological field-measurements.

image

Figure 3. The thermometer screen trials in the grounds of Stratfield Turgis rectory. This is from Gaster, Quarterly Weather Report 1879, Appendix II, p [24], published in 1882 but probably drawn at the same time as the engraving used for the Frontispiece of British Rainfall 1868 (Figure 2).

Download figure to PowerPoint

The thermometer screen trials at Stratfield Turgis ran from November 1868 to April 1870. Observations were made three times daily (9 a.m., 3 p.m. and 9 p.m.), together with maximum and minimum temperatures in the screens that were capable of holding such thermometers. It should be noted that part of this period also coincided with the daily 9 a.m. reading of the 42 raingauges!

The work involved in collating, analysing and writing up such a large trial dataset would be significant today, with ready access to powerful computers and spreadsheet software, but in the 1870s when it would have all been tabulated and summarised by hand it would have been a truly daunting task. It is probably for this reason that the results did not finally appear in print until 1882, when Frederic Gaster's 27-page paper, dated May 1880 and snappily titled Report on experiments made at Strathfield Turgiss [sic] in 1869 with stands or screens of various patterns, devised and employed for the exposing of thermometers, in order to determine the temperature of the air, finally appeared in the Meteorological Office's Quarterly Weather Report for 1879, from which Figure 3 is taken. Gaster's conclusion and recommendation (Appendix II, page 31) was that Stevenson's screen is better adapted than any of the other stands tested to register the temperature of the air. Even before publication of Gaster's analysis, however, … the Meteorological Office adopted the Stevenson screen for use at telegraphic stations, and in 1874 the Council of the Meteorological Society recommended its adoption by the Society's climatological stations (Bilham, 1937).

This was not quite the end of the matter. The Meteorological Society – which became the Royal Meteorological Society only in 1884 – established a thermometer screen committee which undertook further comparative tests, eventually recommending slight modifications to Stevenson's original design, culminating in the endorsement and recommendation of a standard design in late 1883 (Report of the thermometer screen committee: Quarterly Journal of the [Royal] Meteorological Society, 10, pp. 92–94).

Changes have been minor since 1883, at least in the UK and Ireland. A double-width ‘large’ Stevenson screen was introduced during the First World War to accommodate clock-driven autographic instruments recording on paper charts, such as the thermograph and hygrograph. More recently, screens have been made from UV-resistant plastic, with consequent reductions in maintenance requirements (thereby removing some of the inconsistency of readings resulting from deteriorating paintwork). Trials undertaken by the Met Office showed that there was little significant difference between wooden and plastic screens (Perry et al., 2006), and plastic Stevenson screens have since been formally adopted by the Met Office, with a rolling replacement programme as wooden screens reach the end of their working lives.

Global impact

  1. Top of page
  2. The raingauge trials at Stratfield Turgis
  3. The thermometer screen trials
  4. Global impact
  5. Epilogue
  6. Acknowledgements
  7. References

It is hard to underplay the impact of those two painstaking instrument trials in the garden of a north Hampshire rectory just over 140 years ago. Both eventually led to the clear endorsement of a particular design of instrument, which with only minor modifications then quickly became ‘standard’. At the height of the British Empire, such standard instruments soon found their way around the globe. Louvred Stevenson-type screens are the closest that we have to a global standard thermometer housing, and at a conservative estimate there are probably 35 000 in daily use around the world. In the United States, the similar louvred-type thermometer screen known as the Cotton Region Shelter remained as the standard until very recently (many are now being replaced by the slightly smaller Stevenson pattern, or smaller plastic automatic weather station (AWS) screens).

The dominance of the ‘five-inch’ raingauge is less pronounced, owing perhaps more to climatic factors than to any limitations in design: in regions with significant annual snowfall, a raingauge mounted 300mm above the surface is quickly buried. Most funnel-based raingauges cope poorly with snowfall, particularly in windy conditions, and designs combining wind shields and open-top gauges perform better in snowy climates. Even so, a World Meteorological Organization report in 1989 (WMO, 1989) established that the basic Snowdon pattern, or minor variations thereof, was the standard precipitation sensor in 29 countries around the world, with approximately 18 000 gauges in use by national or regional weather services alone. It is likely that there are at least as many again in the care of hydrology or water supply bodies, farms, universities, research organisations and private individuals.

Of course, the numbers of both Stevenson screens and Snowdon-type raingauges are likely to decline in the future as automation continues to erode the dominance of, and long-term standards established by, ‘traditional’ instruments. Such changes of instrument type often result in catastrophic loss of homogeneity in long-term records where little or no period of overlapping observations has been allowed for prior to withdrawal of the ‘traditional’ instruments. Although the limitations of both the Stevenson screen and the simple Snowdon raingauge have been known for more than a century, the existence of long periods of conservative record made using such standardised instruments today provides the essential baseline for assessing the extent and magnitude of climate change.

Epilogue

  1. Top of page
  2. The raingauge trials at Stratfield Turgis
  3. The thermometer screen trials
  4. Global impact
  5. Epilogue
  6. Acknowledgements
  7. References

Charles Griffith remained rector of the parish of Stratfield Turgis for the rest of his life, and in 1892 he published a book on the history of Stratfield Saye parish. After 1870, he continued to make twice-daily readings of temperature, rainfall, pressure, wind and weather, with only a few short gaps in his record. The daily registers are peppered with phenological and entomological notes as well as meteorological observations. His original handwritten records have been preserved in the National Meteorological Archive in Exeter, and I have recently copied his entire observational record from their files. Because Stratfield Turgis rectory is so close to my current observing location, and the records (and instruments) are directly comparable, his notes provide a fascinating local window on the weather of the second half of the nineteenth century compared with today.

His record ends abruptly. The 9 a.m. observation on 25 March 1896 (‘cloudy, fine’) is the last to appear on the monthly return – the rest of the page is blank (Figure 4). Charles Griffith must have been suddenly struck down with a terminal affliction, for he died exactly a week later, on 1 April 1896 (his death notice appeared in The Times, 4 April 1896: issue 34 856, page 1, column A). Both the observation register and the incomplete March 1896 monthly return are marked as being received by the Royal Meteorological Society in June 1896. There are only brief references to his demise in British Rainfall 18951 and in Symons Meteorological Magazine (SMM, April 1896, opening page), but considering his many contributions to meteorology over four decades, it is surprising that no formal obituary ever appeared in either SMM or the Quarterly Journal of the Royal Meteorological Society2. George Symons’ brief note in SMM, however, concluded with the following handsome eulogy: Mr Griffith was a worker rather than a writer; had he written more, the parallelism between the two Hampshire rectors, Gilbert White of Selborne in the 18th century, and Griffith of Strathfield Turgiss in the 19th, would have been noteworthy.

image

Figure 4. Charles Griffith's monthly observations return for March 1896, showing the abrupt end to the observational record at Stratfield Turgis. The page is already signed, ready to post at the end of the month. The original is in the National Meteorological Archive at Exeter.

Download figure to PowerPoint

The rainfall record was carried on by Hannah until 1904, until, at 75, she became too old to continue the record which her husband had begun at the rectory 42 years previously; she died early in 1905. I have used those records extensively in assembling a homegeneous series of monthly rainfall totals for this area stretching back 150 years.

Stratfield Turgis rectory is still there today, recognisably the same as the buildings and gardens depicted on the frontispiece of British Rainfall 1868. A few years ago I contacted the current owners, who (unsurprisingly) had little inkling of the important climatological research that had taken place in the grounds almost 150 years ago, and they kindly allowed me to take photographs of the site from a vantage point close to where the 1868 sketch was made (Figure 5). It is surprising how little had changed, although of course no traces remained of the screens or the raingauges. It was very easy to half-close my eyes and imagine the Victorian clergyman and his small army of pupils racing into the grounds to read the instruments at 9 a.m.

image

Figure 5. Stratfield Turgis rectory today. Compare with Figure 2.

Download figure to PowerPoint

If not for the Victorian clergyman and his pupils making those meticulous records from so many different types of instrument in a north Hampshire rectory garden, day in and day out, standards might still be as ill-defined as they were in the 1860s. For this, Rev. Charles Higman Griffith deserves posthumous recognition. He could not have imagined what an impact his painstaking investigations would have on global climatology over the course of the following 150 years.

Acknowledgements

  1. Top of page
  2. The raingauge trials at Stratfield Turgis
  3. The thermometer screen trials
  4. Global impact
  5. Epilogue
  6. Acknowledgements
  7. References

Thanks to Mark Beswick in the Met Office Archives and Steve Jebson in the National Meteorological Library, and to Philip Eden, who searched for CHG's obituary, without success, in both SMM and QJ. An abbreviated version of this note appeared in the Society's History Group Newsletter, October 2012.

  1. 1

    British Rainfall 1895 was published in July 1896, and this note was presumably inserted close to the volume going to press.

  2. 2

    It seems probable that the reason there was no obituary of Griffith in the Quarterly Journal is that he was not a member of the Society at the time of his death. He was elected a Fellow of the (then) Meteorological Society on 20 November 1867 and resigned his membership on 31 December 1886. (Malcolm Walker, pers. comm.).

References

  1. Top of page
  2. The raingauge trials at Stratfield Turgis
  3. The thermometer screen trials
  4. Global impact
  5. Epilogue
  6. Acknowledgements
  7. References
  • Bilham EG. 1937. A screen for sheathed thermometers. Q. J. R. Meteorol. Soc. 63: 309319.
  • Perry MC, Prior MJ, Parker DE. 2006. An assessment of the suitability of a plastic thermometer screen for climatic data collection. Int. J. Climatology 27: 267276.
  • World Meteorological Organization. 1989. Catalogue of national standard precipitation gauges. Instruments and Observing Methods Report No 39, WMO/TD-No. 313. WMO: Geneva, Switzerland.