






































Microsoft Word - 10wheeler.doc


 History of Meteorology 2 (2005) 133 
 

 

 

 

 

 

 

 

 

 

British Naval Logbooks from the Late Seventeenth Century:  

New climatic information from old sources 
 

Dennis Wheeler 

 
University of Sunderland, Sunderland SR1 3PZ, UK 

e-mail: denniswheeler@beeb.net 

 

Introduction 

 

The recently completed CLIWOC
1
 project funded by the EU has demonstrated the value 

of the climatic information to be found so abundantly in old naval logbooks (Garcia-Herrera et 

al, 2005). This project was, however, confined to the period from 1750 to 1850 and, whilst many 

logbooks, especially in the UK, survive from much earlier times, they were not included in the 

analysis. This paper takes the opportunity to examine the oldest collection of logbooks that exist 

in sufficient quantities to provide a daily, regionally-based, series of weather observations and to 

gauge their value as sources of scientific data and information. Logbooks of this antiquity are 

those of Royal Navy officers, lieutenants and captains, preserved in the National Maritime 

Museum at Greenwich, South East London and in the National Archives in Kew, South West 

London. They cover the period from 1685 onwards that, fortuitously, embraces both a large part 

of the Late Maunder Minimum and the coldest phase of the so-called Little Ice Age. The 

character of the data is examined and some preliminary conclusions are drawn regarding the 

climate of the period. 

 Climatic data are abundant for this period, particularly in Europe: ice core records, 

dendrochronological information, terrestrial and lacustrine sediments  combine with a rich legacy 

of documentary sources to yield a comprehensive picture of the contemporary climate (Jones and 

Mann, 2004). Equally importantly, the first instrumental data also extend back into the 

seventeenth century in the form of Gordon Manley’s Central England Temperature series, the 

oldest such data set in the world (Manley, 1974). Luterbacher et al. (2002) have forged these 

sources together to create valuable multi-proxy pressure field reconstructions for the period. It 

might, therefore, be asked what new insight logbooks can lend to this already well-developed 

field of study? The answer is three-fold. In the first instance, no new data source should be 

overlooked in the search for a clear understanding of the changes that climate has undergone in 

the past centuries. Secondly, and more significantly, the logbook data provide information for a 

                                                
1
 CLIWOC: Climatological Database for the World’s Oceans: 1750 to 1850. www.ucm.es/info/cliwoc . 

 



 British Naval Logbooks 134 
 

part of the Earth’s surface that is notoriously under-represented in such studies – the seas and 

oceans – and do so in the form of direct observations that are ‘non-proxy’ in character. Finally, 

logbook data provide an unrivalled scale of temporal resolution, with observations being made at 

least daily, often more frequently. 

 

The nature of logbook climatic information 

 

Almost as soon as mariners sailed out of sight of land and needed to maintain a record of 

their vessel’s speed, direction of sailing and other navigational details, some form of account was 

kept on board ship. Christopher Columbus provides one of the earliest examples of this type of 

record (Fuson, 1987). From haphazard beginnings, matters became more regularised as the years 

went by and the size of navies became larger and voyages more frequent. By the late seventeenth 

century logbooks assumed a form that has endured with only minor changes until the present-

day. A typical example of such an early logbook is given in Figures 1a and 1b. These reveal the 

degree of detail contained within those venerable documents. On the left-hand page were 

recorded the date, latitude, longitude, bearings, wind directions and ship’s course. The right hand 

side contained a daily account that includes notes on wind force and weather. Other information 

recorded here were the formal proceedings and accounts of the management of the vessel.  

 By these means all weather observations are located in space and time; the former by date 

and the latter by latitude and longitude or, when close to shore, by bearings to observed 

landmarks. Both, however, require cautious interpretation. Dates in English documents of the 

period were based on the Julian Calendar, which was then ten days behind the more accurate 

Gregorian Calendar that was used across much of Europe. The latter was not adopted in England 

until 1752. Latitude could be determined with reasonable accuracy at this time, but longitude was 

a problem not fully resolved until the closing decades of the eighteenth century when the marine 

chronometer had been perfected (Hewson, 1983). Longitudinal corrections can, nevertheless, be 

applied by plotting the route of the vessel and checking against points of known longitude on the 

voyage, such as ports of arrival and departure. These issues are discussed more fully in Koek and 

Können (2005). 

 



 History of Meteorology 2 (2005) 135 
 

 

 

 

 

 



 British Naval Logbooks 136 
 

 

 
Figs. 1a and 1b. Facing pages from a logbook typical of the period. These were written by an unknown officer on 

board HMS Experiment and date from June 1697.  Courtesy of the National Maritime Museum. 



 History of Meteorology 2 (2005) 137 
 

 

 Observations of wind direction were made, as they are today and had then been for over a 

century, on the basis of the 32-point compass. As all such records were in the form of magnetic 

bearings, and corrections for magnetic variation need to be applied to express them by reference 

to true north. Variation changed, however, through time and space, although studies have now 

provided detailed reconstructions of this phenomenon with which it is possible to convert any 

wind direction in magnetic form to its ‘true’ equivalent (Bloxham & Gubbins, 1985). In the 

present case, such corrections were minimal. 

 Weather descriptions, included within the general accounts, were written in the 

unadorned language of the day, and would have been readily understandable to any 

contemporary reader. Most remain comprehensible today, if the spelling (‘raine’, ‘rayne’ etc.) is 

idiosyncratic and rules of punctuation applied in an anarchic fashion. The handwriting is often 

surprisingly clear, and presents few difficulties. 

 The greatest challenge is offered by the wind force terms, which were included within the 

general descriptions. These were of a more technical nature, less familiar to those unaccustomed 

to the ways of the sea and formed part of the arcane vocabulary that mariners employ even today. 

The Beaufort Scale was not adopted by the Royal Navy until well into the nineteenth century 

(Wheeler and Wilkinson, 2004), but the logbook entries point unequivocally to the adoption, 

even in these formative times, of a limited number of widely-used descriptors, suggesting an 

unofficial precursor to the scale proposed by Francis Beaufort. 

 The collection of early logbooks held at the National Maritime Museum was used in this 

study, and a sample compiled from ships sailing in the confined but frequently navigated area of 

the English Channel. Although some logbooks survive from the 1670s (Wheeler, 2004) they are 

sufficiently abundant to provide for a daily series of observations only from 1685. The abstracted 

series concludes in 1700. The sample was comprised of 52 logbooks, with a smaller sub-set of 

randomly selected duplicate logbooks to act as checks for consistency of the record. The daily 

series derived from the sample yielded 5500 days of data. These were gathered into a database, 

the elements of which are listed in Table 1. 

The content of the database gives notable testimony to the volume and variety of climatic 

information that can be abstracted, and inferred, from logbook daily entries. Whilst none of the 

information is based on instrumental evidence, and much depends upon the judgement of the 

recording officer, it must be recalled that such officers would have enjoyed many years of 

experience on board Royal Navy vessels and would have been tutored in the manner of 

observation, ship management and navigation. Even today it is commonplace for VOSs 

(Voluntary Observing Ships) to report important meteorological data based on just such 

judgment-based observations. The published advice to present observers could apply with equal 

veracity to those of the late seventeenth century: 

 

Non-instrumental observations are very important and, being estimates, they are 

dependent upon the personal judgement of the observer. This judgement is the product of 

training and experience at sea, together with practice in making the observations. 

(Meteorological Office, 1977, p.37). 



 British Naval Logbooks 138 
 

 
Table 1. Summary of climatic data derived from seventeenth century English logbooks. 

 
variable notes 

 

ship’s name 

 

As recorded in the logbook 

 

NMM catalogue 

reference 

The NMM has a printed catalogue of all logbooks in its collections. Currently 

this is not available on-line. 

 

Julian date As given in the logbook, but note should be made that the nautical day begins at 

noon and 12 hours ahead of the civil day. 

 

Gregorian date At this time the Julian date was ten days behind the Gregorian. 

 

bearings to 

landmarks 

It was common practice when in the English Channel to give location by 

bearings to known landmarks. 

 

wind force The original term is entered here. 

 

wind direction As recorded on a 32-point compass. No local corrections for magnetic variation 

were required for this period. 

 

weather As recorded in the logbook (original spellings retained). 

 

Rain, snow, 

thunder, hail, fog 

These are derived data, and columns (yes/no entry) are provided to indicate 

when the ‘weather’ text includes any of these distinct elements. 

 

frost/cold day When the weather text indicates cold or frosty conditions, that inclusion is 

indicated here. 

 

Beaufort wind 

force 

Each original entry is re-expressed as its present-day Beaufort Scale equivalent 

(see below). 

 

gale day If the Beaufort expression is gale force or stronger, it is flagged here. 

 

wind by cardinal 

quarter 

Each wind direction is re-expressed in its appropriate cardinal quarter (N, S, E 

or W) 

 

 

The seventeenth century wind force scale 

 

The most well-known summary of the wind force scale said to be used by mariners of the 

late seventeenth century is that offered by Daniel Defoe in his collection of letters and accounts 

that describe the ‘Great Storm’ of 1703 (Defoe, 1704). This scheme is reproduced in Table 2. It 

is interesting to note the similarities to the Beaufort Scale, but it is clear that it does not compare 

in every respect with the vocabulary found in the logbooks of the period. 

Defoe claimed the terms in Table 2 to have been in popular usage, and most were indeed 

encountered in the logbook sample, but no record of either ‘top sail gale’ or of a ‘fret of wind’ 

was found, while ‘fine breezes’ were only rarely encountered. On the other hand, there are terms 

in popular use at the time that appear frequently in the database, for example ‘moderate gales’ 

and ‘blows hard’, which Defoe did not recognize. The principal logbook vocabulary of the age is 

summarized in Table 3 in which the eleven most widely used terms are included. Over 70 



 History of Meteorology 2 (2005) 139 
 

 

different wind force terms were found but, as is evident from the cumulative frequencies, the 

eleven most frequently used items account for nearly 88 per cent of all entries, and many of the 

remainder were used only once or twice. Some, but not all, of these terms survived to be adopted 

a century later by Francis Beaufort. The evolution of the vocabulary during that period and the 

means by which the scales have been rendered comparable have been reviewed in Wheeler and 

Wilkinson (2005). Fortunately most of the archaic wind force terms could be expressed in 

present-day Beaufort equivalent descriptors, leaving only some 1 per cent of wind force entries 

that could not be translated. 

 
Table 2. Defoe’s ‘table of winds’ as described in his account of the ‘Great Storm’ of 1703. Defoe did not number 

his scale, the system being adopted here only for purposes of clarity. It is not intended to translate into 

Beaufort’s terms. 

 
Term grade* Term grade 

stark calm 0 a top sail gale 6 

calm weather  1 blows fresh 7 

little wind 2 a hard gale of wind 8 

a fine breeze 3 a fret of wind 9 

a small gale 4 a storm 10 

a fresh gale 5 a tempest 11 

 
Table 3. Absolute and cumulative frequency (over all entries) of usage of the 11 most widely used wind force terms 

from the late seventeenth century. † indicates terms that have appeared in one or more of the Beaufort 

Scales used since 1806. 

 
Rank of usage Term Frequency Cum. Frequency (%) 

1 fresh gales† 1261 30.6 

2 little wind 621 45.7 

3 moderate gales† 457 56.8 

4 hard gale 342 65.1 

5 small gale 215 70.2 

6 blows hard 207 75.2 

7 variable 134 78.5 

8 fine gales 129 81.6 

9 blows fresh 101 83.1 

10 calm† 99 85.5 

11 strong gales† 99 87.9 

 

Preliminary findings from the database 

 

Careful use of the vocabulary allows the database to be completed along the lines noted 

in Table 1. From the resulting daily series it is possible to estimate the monthly means and counts 

that summarise the overall trends and changing climate of the period. The preliminary findings 

can be summarised as follows: 

 

Wind direction  

 

A general picture of prevailing wind directions is recreated in Table 4 in which the 

directional frequencies are resolved to monthly level but are based on data that have been 



 British Naval Logbooks 140 
 

converted from 32-point to 4-point form, each original entry being classified into one of the 

cardinal points.  

 
Table 4. Mean monthly wind directions (in proportions of each month) for 1685-1700. 

 
 Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec year 

N 0.24 0.16 0.22 0.18 0.25 0.17 0.22 0.17 0.18 0.19 0.19 0.20 0.20 

E 0.25 0.26 0.25 0.23 0.28 0.20 0.20 0.14 0.13 0.23 0.21 0.19 0.21 

S 0.13 0.21 0.20 0.22 0.20 0.23 0.12 0.19 0.15 0.22 0.20 0.24 0.19 

W 0.37 0.37 0.34 0.37 0.27 0.40 0.47 0.51 0.54 0.36 0.40 0.37 0.40 

 

This is the first time that it has been possible to view the behaviour of the wind over the 

sea for such distant times at anything other than the coarsest of resolutions. Moreover, whilst 

contemporary land-based records are by no means absent, they are dogged by the fact that wind 

direction is profoundly altered by the boundary layer conditions and may make it 

unrepresentative of the synoptic states that prevailed at the time of their measurement. This 

problem is not wholly resolved by the use of these marine data, but they represent a marked 

improvement on land-based observations. Westerlies, not unexpectedly, dominate the picture, 

but attention should be drawn to the higher frequency of easterlies in late winter and early spring. 

More specifically, Figure 2 shows the changing degree of westerliness from 1685 to 1700 using 

data aggregated to the monthly statistics.  

 

 
Fig. 2. Monthly index westerliness based on logbook observations. A Gaussian filter running mean is superimposed. 

 

This important parameter is closely related to, but not identical to, the North Atlantic 

Oscillation index (the correlation between the two over the study period is r = 0.363 for the year, 

but r = 0.533 for the period December to February: both are significant at the 0.01 level). 

Slonosky et al. (2001) have suggested that the region in question – that of the English Channel – 

is a key location for the study of synoptic scale climatic behaviour over a much wider area of the 

north-east Atlantic region. It is important, therefore, to recall that this ‘index of westerliness’ is 

not a proxy dataset, but is based on direct observations of the conditions of the day and that is 

may represent conditions that prevailed beyond the immediate limits of the study region. There is 



 History of Meteorology 2 (2005) 141 
 

 

a high degree of temporal detail, and particular periods can be readily identified for future study. 

Among them, those around 1688, 1695 and 1697 are noteworthy, as are those particular months 

when westerlies appear to have been absent (February 1688, March 1690, February 1692, 

December 1695 and April 1697). This finding is important because it emphasises the inherent 

variability of climate at this time. The episodic character of the periods in which westerlies are 

poorly represented shows also that the coolness of the British Isles at this time – as witnessed by 

the Central England temperature series (Manley, 1974) – cannot be accounted for by persistent 

winter easterlies drawing cold polar continental air towards the region and that any tendency to 

such meridionality manifests itself in only a discontinuous fashion. Nevertheless, the record 

suggests that winter easterlies were generally more frequent at this time than is the case today by 

a margin of some 10 per cent. 

 

Gale frequencies 

 

The database’s gale series provide a particularly valuable data set that suggests that gales 

were far more frequent in the closing decades of the seventeenth century than they were three 

hundred later at the close of the twentieth (Figure 3). More importantly, as Wheeler and 

Dominguez-Suarez (2006) have shown, the greatest increase in gale frequencies occurred during 

the period March to October and, remarkably, they were less frequent in the winter season.  This 

is probably accounted for by the steeper summer temperature gradients between the sub-tropical 

and high latitudes during summer, drawing the polar jet southwards and providing the system 

with a greater potential for active cyclonic development than is the case in the warmer summers 

that currently prevail. The appearance of well-developed continental anticyclones in winter 

might provide for ‘blocking’ situations in which the cyclones would adopt paths that avoided the 

British Isles, thereby decreasing the incidence of gales at that season.  

 
Fig. 3. Monthly mean frequencies of days with winds of Beaufort force 8 or more for the period 1685-1700 (shaded) 

and for the mid-twentieth century. 

 

 

0

2

4

6

8

10

12

14

J F M A M J J A S O N D

a
ve

ra
g
e
 f

re
q
u
e
n
cy

1685-1700 C20 data



 British Naval Logbooks 142 
 

Precipitation 

 

Even today, measurement of precipitation at sea is fraught with problems. In the late 

seventeenth century no attempt was made to measure the depth of rainfall but officers made 

careful note of precipitation when it was seen, usually taking care to distinguish between snow, 

hail and rain. The latter might be further qualified by the use of terms such as ‘heavy’, ‘light’, 

‘showers’ or, even, ‘small rain’ and ‘mizzling’. The latter probably referring to drizzle (it is a 

term encountered in some parts of Britain today). Typical logbook entries would read “squalls of 

raine [sic] and hail”, or “the weather frosty with snow”. The most common entry would be in the 

form of a reference to wind force, followed “with rain” or “with snow”. 

With these observations it is possible to reconstruct a series of rain days. Figure 4 shows 

this series, in which the daily data have been aggregated into monthly counts. The average 

number of rain days per year was 87, reaching a maximum of 140 in 1697 and a minimum of 50 

in 1695. Unless the climate had been notably drier at this time (and there is no evidence that it 

was) such data suggest under-recording of precipitation, emphasising the need to concentrate on 

the patterns of relative and not absolute frequencies. Greater care yet is needed with reference to 

fog, and it is evident that the term was used to embrace conditions that would not today qualify 

for such an entry and was used as a general description for poor visibility, be it fog or mist sensu 

stricto or those times when sea spray might have obscured visibility; fog [sic] was often recorded 

with strong gales. 

 

 
 
Fig. 4. Series of rain days (by months) for the English Channel 1685 to 1700.Units are standard deviations about the 

overall mean. 

 

Allowing for the fact that there may be unavoidable under-recording of precipitation, it is 

still possible to recreate the patterns of wet and dry spells, and Figure 4 reveals the 

predominantly wet conditions that characterised the close of the study period and the drier 

conditions of the immediately preceding months. 

 



 History of Meteorology 2 (2005) 143 
 

 

Cold weather 

 

No ships of this period carried thermometers but, perhaps because of the difficulties that 

frost and very cold conditions made for ship management, cold weather tended to be noted. Hot 

weather was not accorded the same degree of attention. The CET for this period is expressed in 

monthly form only, and daily temperature data are rare for this period, nevertheless, individual 

cold days are chronicled in the logbooks sometimes identifying protracted periods of such 

conditions and defining with precision the start and conclusions of such spells of weather. The 

most marked of these are summarised in Table 5, in which the corresponding wind directions and 

CET anomalies are included.  

It is informative to note that these spells coincide with notably cold months in the CET 

and that they are commonly the result of outbreaks of easterly winds. They give, however, a finer 

degree of resolution and detail to the temperature regime than has hitherto been possible. 

 



 British Naval Logbooks 144 
 

Table 5. Summary of cold spells noted in the logbooks for the period 1685 to 1700. 

 
dates (NS) 

of cold 

weather 

spells 

description winds CET 

anomaly 

for the 

month 

(deg. C) 

 

details 

10-13 & 

17
th

 Jan. 

1685 

frosty between NE 

and S 

-2.1 Recorded at the Downs on HMS Deptford by an 

unknown officer 

 

29
th

 Dec. - 

7, 9-11, 13, 

15 & 16
th

 

Jan. 1689 

 

frosty and 

frosted  

hard 

exclusively 

from NE 

-2.1 Recorded at Portsmouth on HMS Elizabeth by an 

unknown officer 

31
st
 Jan. & 

4, 5, 11, 14, 

16, 19
th

 Feb. 

1692 

cold and 

frosty 

Mostly NE until 

11
th

 Feb, then 

between N and 

E 

-2.6  

(Feb.) 

Recorded by Lt. W Bridgman  on HMS Hope at the 

Downs. From 10
th

 February the record is from HMS 

Centurion (officer unknown) off the southern Irish 

coast. 

 

29
th

 Nov. – 

3
rd

 Dec. 

1693 

frosty and 

cold raw 

Winds mostly 

NW 

-0.3  

(Dec.) 

Recorded at Plymouth on board HMS Devonshire 

(officer unknown) 

 

14, 17-19 & 

24
th

 Jan. 

1694 

frosty Winds between 

NW and NE 

-2.6 Recorded at St Helens on board HMS Dragon 

(officer unknown) 

 

15-17
th

 

March 1699 

frosty Winds NE to 

ESE becoming 

S on 17
th

 

-0.6 Recorded off North Foreland on board HMS 

Bedford (officer unknown). 

 

 

 

Conclusions 

 

 The use of logbooks for purposes of climatic reconstructions is still very much in its 

infancy. Nevertheless, as the above example shows, the abstracted and carefully treated data can 

provide a unique insight into the climate of past centuries. For the first time, scientists can see 

the how the oceanic weather was behaving at a daily scale of resolution, and can aggregate these 

data to recreate a longer-term, climatic impression at monthly, seasonal or annual scales should it 

be necessary. Equally importantly, the data source has a notable degree of homogeneity; the 

vocabulary is tolerably consistent and the methods of observations in all probability identical 

between officers. While logbooks exist in such huge quantities as they do (over 120,000 exist in 

UK archives for the period 1680 to 1850) there is abundant scope for this modest exercise to be 

repeated at a much larger scale, as has been done by the EU-funded CLIWOC (Climatological 

Database for the World’s Oceans:1750 to 1850) cited in the opening paragraph.  

 



 History of Meteorology 2 (2005) 145 
 

 

References 

 

Bloxham, J. and D. Gubbins, 1985. “The secular variation of Earth’s magnetic field,” Nature 

317, 777-781. 

Defoe, D., 1704. The Storm: or, A collection of the most Remarkable casualties and Disasters 

which happen’d in the late Dreadful Tempest by both Sea and Land.  London. 

Fuson, R.H., 1987. The Log of Christopher Columbus. Southampton, UK: Ashford Press. 

García-Herrera, R., G.P. Können, D. Wheeler, M.R. Prieto, P.D. Jones, and F.B. Koek, 2005. 

“CLIWOC: A climatological database for the world’s oceans 1750-1854,” Climatic 

Change 73, 1-12. 

Hewson, J.B., 1983. A history of the practice of navigation. Glasgow: Brown, Son and Ferguson. 

Koek, F.B., and G.P. Können, 2005. “Determination of wind force and present weather terms: 

The Dutch case,” Climatic Change 73, 79-95. 

Jones, P.D. and M. Mann, M., 2004. “Climate over past millennia,” Rev. Geophys., 42, RG2002 

doi:10.1029/2003RG000143. 

Manley, G., 1974. “Central England temperatures: Monthly means 1659 to 1973,” Quart. J. Roy. 

Meteorol. Soc., 100, 385-405. 

Meteorological Office, 1977. Marine Observer’s Handbook. London: HMSO. 

Luterbacher, J., C. Schmutz, D. Gyalistras, E. Xoplaxi, and H. Wanner, 2002. “Extending north 

Atlantic oscillation reconstructions back to 1500,” Atmos. Sci. Letters, 2, 114-124. 

Slonosky, V., P.D. Jones, and T.D. Davies, 2001. Instrumental pressure observations and 

atmospheric circulation from the 17
th
 and 18

th
 centuries: London and Paris,” Intl. J. 

Climatology, 21, 285-298. 

Wheeler, D., 2004. Understanding seventeenth century ships’ logbooks: An exercise in historical 

climatology, J. Maritime Res., March 2004, published electronically at www.jmr.ac.uk. 

Wheeler, D. and J. Dominguez Suarez, 2006. “Climatic reconstructions for the northeast Atlantic 

region AD 1685-1700: A new source of evidence from naval logbooks,” The Holocene, 

11, in press. 

Wheeler. D. and C. Wilkinson, 2004. From calm to storm: The origins of the Beaufort Wind 

Scale.  Mariner’s Mirror, 90, 187-201. 

Wheeler. D. and C. Wilkinson, C., 2005. “Determination of terms for wind force/present 

weather. The English case,” Climatic Change 73, 57-77.  

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 British Naval Logbooks 146 
 

 

 

 

 


