1
Introduction

The appearance of the sky and its relationship to the atmosphere's properties have, no doubt, always provoked curiosity, with early ideas on explaining its variations available from Aristotle. A defining change in the philosophy of atmospheric studies occurred in the seventeenth century, however, with the beginning of quantitative measurements, and the dawn of the instrumental age. Since then, elaborate devices to monitor and record changes in the elements have continued to develop, providing, along the way, measurements underpinning the instrumental record of past environmental changes, most notably in air temperature. This means that characterising and understanding early meteorological instruments are of much more than solely historical interest, as recovering past measurements, whilst recognising their limitations, can also have immediate geophysical relevance.

An important meteorological example is the reconstruction of past temperature variations from the miscellaneous thermometer records originally undertaken to satisfy personal curiosity. Ships' logbooks provide another example, in terms of geomagnetic field changes. Beyond the actual data produced in either case, this also provides a reminder that all measurements can have unforeseen applications well beyond their original motivation [1], either through a change of context in which the measurements are evaluated, or because other subsequently important information has unwittingly been included.1 Such future scope is probably impossible to predict completely, but it can to some extent be allowed for by ensuring a full appreciation of the related measurement science through careful description of the construction, calibration and recording procedures for the instrumentation employed. The possible future legacy implied by taking this historical perspective adds further motivation for rigour in the modern science of atmospheric measurement.

This chapter briefly highlights some of the major historical landmarks in development of instrumentation science for meteorology, and concludes with an overview of the book's material.

1.1 The instrumental age

Many of the early atmospheric measuring instruments were developed in Florence, due perhaps in part to the experimental physical science tradition inspired by Galileo, and availability of the necessary craftsmanship. This included early thermometers, such as the thermoscope produced during the late 1500s to determine changes in temperature. Following key instrument advances such as the invention of the barometer by Evangelista Torricelli in 1643 and an awareness of the need for standardisation of thermometers, modern quantitative study of the atmosphere can be considered to date from the mid-seventeenth century.

Early measurement networks followed from the availability of measuring technologies combined with the formation of learned scientific societies, which together provided the means to record and exchange information in a published form. Comparison of measurements required a system of standardisation, such as that achieved through common instrumentation, and in many cases, common exposure. For thermometers, an agreed temperature scale was necessary and the Celsius,2 Fahrenheit3 or Réaumur4 scales all originated in the eighteenth century [2]. The meteorological values were published as tables of readings, in many cases without any further processing, but which were sufficiently complete for analysis to be made later.

1.2 Measurements and the climate record

Early weather records can be found in 'weather diaries', which were usually kept by well-educated and well-resourced individuals able to purchase or construct scientific instruments such as barometers and thermometers. In some cases, these diaries contain considerable descriptive and quantitative geophysical data, such as those of temperature and rainfall measurements (Figure 1.1).

Figure 1.1 Example page from a weather diary (kept by an apothecary and surgeon, Thomas Hughes at Stroud, Gloucestershire, between 1771 and 1813), in which daily measurements of air pressure, temperature, humidity, rainfall and weather were recorded. As well as quantitative weather information, this particular diary includes other geophysical information, such as timings of earthquakes and even occurrence of the aurora borealis, an indirect measure of solar activity [3]. (Reproduced from Reference 3 with permission of The Met Office.)

Such early data sources are important because of the reference information they provide for the study of climate change, and they therefore remain of scientific value many centuries later. This is particularly true of the disparate thermometer measurements made in southern England from the 1600s, which, although made originally by individuals in an uncoordinated way, now provide an important climate data resource. The temperature readings were cross-checked and compiled5 in the 1950s, drawing on knowledge of the different instruments used and understanding of their exposures [4]. This important synthesis generated a long series of temperature data for an area conveniently described as 'Central England', amounting to an approximately triangular region bounded by Bristol, Manchester and London.

The Central England measurements form the longest continuous set of monthly instrumental atmospheric temperatures available anywhere in the world, beginning in January 1659. (Daily values are also available, beginning in 1772; see Reference 5.) Figure 1.2 shows minimum, maximum and mean annual temperatures of the monthly Central England Temperature (CET) series.

Figure 1.2 Monthly temperatures of 'Central England', originally constructed from historical thermometer records by Manley, and continued using updated modern measurements by the Hadley Centre of the UK Met Office [5]. The thick central line shows the annual mean temperature, with the upper and lower lines the mean values for summer (June-July-August) and winter (December-January-February) respectively (the degraded resolution of the early thermometers is also evident). (Reproduced from Reference 5 with permission of The Met Office.)

1.3 Clouds and rainfall

In the nineteenth century, classification, quantification and taxonomy became an important aspect of many sciences, particularly in the life sciences and geology, so it was natural for similar approaches to be extended to meteorology. The classification of clouds6 was one early aspect, and the compilation of rainfall data also helped further develop the quantitative basis for environmental description. Major developments in meteorology continued in the mid-nineteenth century, following the foundation of the Meteorological Society in 1850, and the establishment of the early Met Office in 1854 under Admiral Fitzroy.7 The British Association for the Advancement of Science convened a Rainfall Committee, with G.J. Symons as secretary. Compilation of historical rainfall data for the United Kingdom was a herculean undertaking, but, following adverts in many local newspapers leading to thousands of replies, Symons [6] did conclude in 1866 that 'there are not now very many records in private hands of which copies are not already obtained and classified.' The legacy of this work is the series of annual volumes of Symons British Rainfall. Further, a continuous series of monthly data [7] for England and Wales Precipitation (EWP) exists from 1766 (see Figure 1.3).

Figure 1.3 Annual rainfall for England and Wales. (Reproduced with permission of The Met Office.)

1.4 Standardisation of air temperature measurements

Standardised exposure for air temperature measurements began in the nineteenth century [8], when meteorological instruments were becoming increasingly available commercially.8 Early (1841) exposure of thermometers for air temperature measurement was through use of a Glaisher stand,9 a simple shading board which was rotated manually to prevent direct solar radiation reaching the thermometer [9]. The Glaisher stand's effectiveness depended on the diligence of the observer required to turn the stand after each reading. If the interval between readings became too long, direct sunlight could still reach the thermometer. The practical difficulty in manually turning the shade board yet retaining good ventilation was solved by Thomas Stevenson10 in 1863, in the form of a double-louvered wooden box painted gloss white. This gave protection to thermometers from solar radiation in all directions, and ensured long wave radiation exchange was with the interior of the screen. The use of a double-louver increased the length of the air path through the screen, which brought the interior of the screen material closer to air temperature than alternatives of simple slits or mesh. In its original form, the Stevenson screen was a wooden box 15 inches high, 14.5 inches long and 7.5 inches wide. It had a solid roof with integral ventilator, and the thermometers were mounted horizontally 4 feet above the ground.

Many minor variants on the Stevenson screen were made. The Scottish physicist John Aitken investigated screen properties [10], noting much later [11] that nothing had been done to mitigate the effects of thermal inertia of...