Identifying New Zealand, Southeast Australia, and Southwest Pacific historical weather data sources using Ian Nicholson's Log of Logs


  • This work was supported in part by the Government of New Caledonia and the French Pacific Fund [Recovery, safeguarding and dissemination of historic French meteorological and climatological data: a Pacific component of the international ACRE programme of Atmospheric Circulation Reconstructions over the Earth (Récoupération, sauvegarde et diffusion des données historiques françaises de météorologie et de climatologie: une composante Pacifique du programme international ACRE de reconstitution de la circulation atmosphérique terrestre)], the National Oceanic and Atmospheric Administration (support to A. M. L. for the South Pacific Rainfall Atlas), and the New Zealand Ministry for Science and Innovation (core funding to NIWA's National Climate Centre project Climate Present and Past).


Historical meteorological data are essential for increasing the level of understanding about past, present, and future climates. In the Northern Hemisphere, a significant amount of research has been dedicated to rescuing climate data from historical sources such as ship logbooks (e.g. RECLAIM, CLIWOC, and ICOADS). However, limited research in this field has focused on New Zealand, Southeast Australia, and the Southwest Pacific. Because these regions were colonized recently by Europeans (~200 years ago), only 50–100 years of land-based meteorological data exist for many locations. However, meteorological information contained in ship logbooks may extend and reinforce the existing historical climate record for these regions. The Log of Logs is a catalogue of ships that visited Australia, New Zealand, and surrounding waters in the th and 20th centuries of the Common Era. These volumes provide a record of the location of ship logbooks. This study extracted information from the Log of Logs for New Zealand, Southeast Australia, and the Southwest Pacific for 1786–1900. The purpose of this was to locate ship logbooks that may contain meteorological data. The next stage of this project is to gather, image, digitize, and to analyse the data from the prioritized logbooks. These data have application for local climate reconstruction, extension of regional circulation indices, and augmentation of the extended reanalysis without radiosondes effort.


Identifier: doi:10.5281/zenodo.6901

Creator: Ian Nicholson

Title: The Log of Logs Volumes 1, 2 & 3

Publisher: ZENODO

Publication year: 2013

Resource type: Book

Version: 1.0


1.1. The importance of historical climate data

Extending climate records further into the past is essential for establishing the range of natural variations that are possible and for placing current natural climate variability in context with past regimes. Future climate change scenario confidence can be improved by extending spatio-temporal coverage of historical surface observations that models are compared against (IPCC, 2007c; Sections 6.1 and 6.2). Obtaining numerous overlapping records within a region increases confidence in past climate and weather observations because the records can be corroborated. Therefore, recovering and analysing historical meteorological and climate records can provide a useful data source contributing to development of climate change mitigation strategies, risk assessments for insurance purposes, and improving the knowledge base that can assist in the adaptive capacity of communities vulnerable to climate and weather extremes (IPCC, 2007b).

Parts of Western Europe retain land-based instrumental meteorological records that extend to the mid-17th Century, and Eastern USA records date back to the mid-18th Century (Jones & Bradley, 1992; Dobrovolný et al., 2010; Luterbacher et al., 2010). Due to the relatively late arrival of Europeans (with instruments and written languages) in Australia, New Zealand, and the Southwest Pacific Islands around 200 years ago, the availability of land-based instrumental records are considerably more sparse than that of the Northern Hemisphere. Generally, high-quality and widely dispersed meteorological data in the Southwest Pacific, including Australia and New Zealand, are only available for the past 50–100 years (Gergis et al., 2010), with fewer records and fewer locations with observations for the 19th Century.

At present, climate change scenarios in the Southwest Pacific region have high uncertainty relative to other parts of the world, in part because of divergent climate model results about future El Niño-Southern Oscillation (ENSO) activity (IPCC, 2007a). In part, this is not only due to model differences but also because our understanding of ENSO and documented surface effects are temporally limited. Because ENSO drives a significant portion of inter-annual to decadal climate variability (along with extremes like tropical cyclones) for the island nations in the Southwest Pacific (Diamond et al., 2012), seeking additional weather and climate data useful for ENSO reconstruction is warranted. In addition, the dearth of available surface observations to test climate model simulations against also compounds the inadequacies of understanding decadal and longer climate variability and dynamics for the region.

1.2. Ship logbooks as a source of historical meteorological data

Recovery of historical climate data can certainly help to improve the situation described above. However, instrumentation was not widely distributed during the colonial era in the Southwest Pacific, which means that terrestrial observations are often sparsely located or temporally condensed (or sporadic). We also consider that the Southwest Pacific is largely oceanic, and thus the proportion of land relative to ocean is low, meaning there will have been fewer opportunities to establish station-based observations than in the Northern Hemisphere.

Despite the reduced land area of the Southwest Pacific, the age of exploration saw several voyages of significance occur (i.e. voyages of discovery; British and United States Antarctic Expeditions, etc.), which documented locations and observations of the physical environment, including the weather. The ‘discoveries’ of New Zealand in 1769 and Australia in 1770 by Captain James Cook were followed by European migration, convict transport, and trade voyages to these new islands, which were set up as colonies of several European nations. Many of these voyages occurred before land-based meteorological observation stations were established (Gergis et al., 2010).

For hundreds of years before instrumentation became widely available, comments about wind and weather conditions were recorded in logbooks (Wilkinson et al., 2010). With the initiation of instrumentation aboard ships in the early 19th Century, measurements of variables such as air and sea surface temperature and barometric pressure supplemented land-based observations of weather. The British East India Company was by and large the most consistent source of marine instrumental measurements prior to 1853 (Brohan et al., 2009). The 1853 Brussels Maritime Conference established a coordinated effort to systematically collect uniform and consistent instrumental meteorological observations aboard ships (Maury, 1854). US Navy Lieutenant Matthew Fontaine Maury led a delegation of naval representatives from ten nations (Belgium, Denmark, France, the United Kingdom, The Netherlands, Norway, Portugal, Russia, Sweden, and the United States of America) at that conference to agree on a protocol for collecting observations, which would lead to improvements in navigation and essentially our understanding of marine weather and climate that commonly affects seagoing vessels.

Due to the rigorous manner with which logbooks were kept, these sources provide an important and reliable archive of marine and atmospheric data (Küttel et al., 2010) to complement and extend the record of land-based data. Thousands of voyages and hundreds of thousands of logbook entries were made and subsequently archived during the colonial era in the Southwest Pacific, and therefore ship logbooks are a potential source of historical climate information for Australasia and the Southern Hemisphere.

Projects based in the United Kingdom, United States, and Europe have been involved in imaging, digitizing, and analysing meteorological data from 18th to 20th Century ship logbooks. These initiatives include ACRE (Atmospheric Circulation Reconstructions over the Earth;; ICOADS (International Comprehensive Ocean-Atmosphere Data Set; Woodruff et al., 2005), CLIWOC (Climatological Database for the World's Oceans; Garcia-Herrera et al., 2005), RECLAIM (RECovery of Logbooks and International Marine data; Wilkinson et al., 2010), and SEARCH (Southeast Australian Recent Climate History, So far, these projects have been successful in extending climate datasets back to the mid-18th Century for some regions. Although these on-going projects are wide in scope, there remains limited analysed data for Southwest Pacific, Southeast Australia, and New Zealand waters (most of the data focus on the Atlantic and Indian Oceans).

2. Aims and methodology

This study investigates the set of Log of Logs publications (Nicholson, 1990, 1993, 1999 – available online as Nicholson, 2013) and legitimizes their use as a resource for identification and prioritization of logbook recovery. The analyses undertaken highlight opportunities for directed meteorological data rescue based on the time and place that historical recordings were made in Australasia and the Southwest Pacific. A primary aim was to identify ship logbooks that potentially contain meteorological data from the late 18th and 19th Centuries that could supplement the temporal gaps in mid-to-late 19th Century land based data coverage, and extend the overall spatio-temporal coverage for Australasia and the Southwest Pacific from the mid-19th Century back in time. This undertaking is useful as a first step in contributing to a global data rescue effort (Wilkinson et al., 2010), because logbooks that contain meteorological data are widely dispersed, but are often grouped as special collections in common archives that can be readily accessed. As such, our analysis was done to assist historical climate researchers in their effort to locate (and subsequently digitally scan) these important documents. In addition, a critical first step was to identify several logbooks that may contain meteorological measurements for a common time period, because not all logbooks will contain relevant meteorological data, but specific time periods may be of interest to a researcher. An assessment of the availability of logs for different locations through time will help to facilitate recovery and subsequent transcribing of data that can be quality controlled and submitted to support the extended reanalysis without radiosondes initiatives (e.g. Compo et al., 2011), and support augmentation of local databases used to analyse climate variability and change. Here, we describe the analysis of the Log of Logs volumes with the purpose of identifying locations and outlining the temporal coverage of possible primary meteorological measurements as recorded by marine voyages for three broad areas: New Zealand, Southeast Australia, and the Southwest Pacific Islands.

2.1. The Log of Logs

The Log of Logs (Nicholson, 1990, 1993, 1999) is ‘…a catalogue of logs, journals, shipboard diaries, letters, and all forms of narratives, 1786–1998, for Australia and New Zealand, and surrounding oceans’ (Nicholson, 1999: iii). These volumes were compiled by Ian Nicholson, a commodore of the Royal Australian Navy, and were published between 1990 and 1999. Nicholson compiled the Log of Logs so the locations of records of ship voyages in Australasian and surrounding waters would be preserved and made accessible to historical and genealogical researchers, among others.

The Log of Logs volumes list ships alphabetically by name and include dates of voyages for which there are some form of narratives remaining, the nature of the narratives (such as logbooks, précis of voyages, passenger journals, among others), and where they were held at the time of publishing. Nicholson compiled records from worldwide sources, including libraries, universities, naval offices, public records offices and museums, with the majority located in the United Kingdom, Australia, New Zealand, and the United States. Volumes two and three update and supplement the records in the previous volumes.

The Log of Logs contains records of different types of ships (Table 1), many of which made multiple journeys to Australia, New Zealand, and to the Pacific. Some ships remained in one location for extended periods of time, such as ships that were anchored in New Zealand ports during the New Zealand Land Wars (1845–1872; King, 2003).

Table 1. Different types of ships recorded in Log of Logs
Type of shipExample
Merchant/trading shipsTrans-Tasman, from Europe, throughout Pacific
Warships/Naval vessels (including survey ships)Service throughout NZ, Australia, Pacific
Migrant shipsUnited Kingdom to Australia/NZ
Whaling/sealing shipsPredominantly from east coast USA to Pacific/sub-Antarctic
Convict transport shipsUnited Kingdom to Sydney and Hobart
Exploration shipse.g. HMS Endeavour (Captain James Cook)
Scientific shipse.g. HMS Beagle (Charles Darwin)
Postal service shipse.g. NZ to Panama postal run
Short trip passenger shipse.g. Wellington-Lyttelton, trans-Tasman
River boatsTrade on Murray, Darling, Derwent Rivers

2.2. Regions for analysis

Three regions were identified as targets for analysis of ship logbook data: New Zealand, Southeast Australia, and the Southwest Pacific region between latitudes 10oN and 30oS (see specific areas in Figure 1). These regions were identified because they are the main the focus of the Log of Logs. The Southeast Australian part of the analysis included the states of South Australia, New South Wales, Victoria, and Tasmania, as well as Norfolk and Lord Howe Islands. Analysis for New Zealand also considered Chatham Islands and subantarctic islands such as Macquarie, Auckland, and Campbell Islands. The Southwest Pacific analysis included all island groups selected in Figure 1. See Tables 2-4 for a more detailed list of locations.

Table 2. Keywords used in Log of Logs searches in New Zealand region
(New) Zealand (NZ)Tasman SeaAuckland (Auck)
Wellington (Wgtn)Christchurch (Chch)Dunedin (Dun)
OtagoLytteltonPort Chalmers
Port NicholsonKororarekaRussell
Bay of Islands (BOI)HokiangaMaori
TaurangaMarlboroughChatham Islands
NelsonKaikouraCloudy Bay
AkaroaHawkes BayMercury Bay
Banks PeninsulaInvercargillCanterbury (Cant)
Stewart IslandManukauWaikato
New PlymouthWhanganuiTaranaki
Hauraki GulfNapierPort Underwood
Macquarie IslandCampbell Island
Table 3. Keywords used in Log of Logs searches in Southeast Australia region
Sydney (Syd)DerwentPort Arthur
Melbourne (Melb)Australia (Aust)Geelong
NewcastleMacquarieHoldfast Bay
Hobart (HTn)Maria IslandPort Lincoln
New South Wales (NSW)Norfolk IslandPort Pirie
Tasmania (TAS)Great Australian BightJervis Bay
Victoria (VIC)King IslandMurrumbidgee River
South Australia (SA)Van Diemen's LandLaunceston
Western Australia (WA)Lord Howe Island (LHI)Bass Strait
AdelaideMurray RiverKangaroo Island
Port JacksonDarling RiverPort Dalrymple
Botany BayGeorgetownBroken Bay
Twofold BayPort PhillipTasman Sea
Port StephensVDL (Van Diemen's Land)
Table 4. Keywords used in Log of Logs searches in Pacific region
NukuhivaJarvis IslandApiaRaiatea
PalauPongopongoNew IrelandMalden
SamoaNew BritainPenrhyn IslandFiji
Navigator IslandsNadi (Nandi)RotumaAtiu
Line IslandsMitiaroMarquesasMangareva
MaukeCook IslandsLoyalty IslandsLifu
Port VilaIsle of PinesBougainvilleHoniara
OuveaBasilakiSolomon IslandsMoorea
Ellice IslandsPort MoresbyEimeoTorres Strait
Papua New Guinea (PNG)RabaulTikopiaPago Pago
(New) CaledoniaPhoenix IslandsFunafutiRaoul
Gilbert IslandsSanta CruzMehetiaMarshall Islands
TokelauPitcairn IslandKiribatiSociety Islands
FutunaNew HebridesAlofiTuamotu
Wallis IslandNauruKingsmill IslandsPapeete
Caroline IslandsBora BoraKermadec IslandsHuahine
TarawaRarotongaHowland IslandFriendly Islands
Baker IslandHiva OaKingmanAtuona
PalmyraGambier IslandsKiritimatiAitutaki
Christmas IslandVanuatuAvaruaBismark
Figure 1.

Map showing regions of interest to this study (SE Australia, NZ, SW Pacific). Source:

2.3. Analysing Log of Logs for relevant ship logbooks

The three Log of Logs volumes (each ~600 pages) were scanned and spliced into PDF files to facilitate Optical Character Recognition (OCR) with font calibration to enable keyword searching. Keywords for the target regions (Tables 2-4) were chosen because of their relevance to the maritime industry and the regions of interest in the 19th century (such as important ports, towns and landmarks). These keywords were mentioned frequently in ship logbooks and other sources, and thus appear in the Log of Logs.

Ships were required to match certain criteria during the OCR keyword searches in order to be considered as a source of possible meteorological data for the target regions. These criteria included the following:

  • Matching to one or more of the keywords from Tables 2-4.
  • Extending into the past beyond 1900. Voyages into the target regions after this date were not considered.
  • Logs that covered the earliest part of the Colonial period (prior to 1830) were noted based on any length of record they contained (provided they matched the other criteria).
  • After 1830, ships became more numerous in the region, and therefore the OCR search limited the findings to records of >1 year. The reason for only selecting ships that had a record of >1 year from this time forward was because longer time series of meteorological data are more desirable for climate reconstruction and the effort to recover longer and more complete records are deemed more valuable.

For the ships that matched the criteria above, the ship's name, year(s), location, and the sources of potential meteorological data were recorded in a Microsoft Excel spreadsheet to facilitate further analysis. This analysis involved separating out the ship records that contained logbooks, Royal Navy remark books and captain's diaries, as these are of higher priority when locating meteorological data than other sources such as passenger diaries and newspaper articles. Passenger journals and diaries are likely to be an account of shipboard life and the passenger's experiences, rather than a source of meteorological data. For ease of analysis, logbooks, Royal Navy remark books, and captain's diaries are all considered as ‘logs’. Figure 2 shows an excerpt from Log of Logs for the ship HMS Havannah, as an example of how the information from the Log of Logs was collected.

Figure 2.

Example of a ship's entry in Log of Logs volume 1 (Nicholson, 1990: 223).

The description of the entry of HMS Havannah in Figure 2 follows: From 1848 to 1851 the frigate HMS Havannah served for the Australian division of the East India Station in the Pacific, including New Zealand, New Caledonia, New Hebrides (present-day Vanuatu), Loyalty Islands, Santa Cruz Islands, Fiji, Samoa, and Tonga. The description also contains notation indicating location and form of the shiplog data: for instance, a logbook from 1849 to 1850 is available as a microfilm copy in the National Library of Australia, and seven remark books from 1848 to 1851 are available from the Royal Navy Hydrographical Department in England and as copies from AJCP (Australian Joint Copy Project, located as microfilms in National Library of Australia). A logbook from 1848 to 1851 by F Hixson (the master's assistant) is available at Mitchell Library in Sydney. Another log is also available for August to December 1851 at the National Maritime Museum in the United Kingdom. The journals documented in the Log of Logs entry may contain meteorological data, but this is relatively unlikely; so therefore the logs and remark books were considered as a higher priority to document for historical climate recovery. The second entry for HMS Havannah shows that the same ship sailed in the Pacific from 1855 to 1858, and one of its destinations was Hervey Island in the Cook Islands. The ship was not in Australian or New Zealand waters during this period. Remark books for the period are available from the Royal Navy Hydrographical Department in England, and will be copied by AJCP in due course. As this volume of Log of Logs was published in 1990, AJCP most likely has a copy of the remark books (the AJCP ended in 1997). Table 5 shows the information from HMS Havannah's entry in Log of Logs in the spreadsheet form used for analysis. This shows how the keywords from Tables 2-4 were used to identify ships of interest.

Table 5. Example of spreadsheet layout for Log of Logs analysis. Abbreviations are defined in appendix
Ship's nameStart yearEnd yearLocationLog?Sources
Havannah18481851Pacific, NZ, New Caledonia, New Hebrides, Loyalty Is, Santa Cruz Is, Fiji, Samoa, TongaYesNL, RN Hydro, AJCP, ML, NMM
Havannah (same)18551858Pacific, Cook IslandsYesRN Hydro, AJCP

After the three Log of Logs volumes were searched for relevant keywords and the ships that matched the criteria were recorded, the data were analysed to provide information on the frequency of ship voyage forays into the three target regions from 1788 to 1900. The total number of voyages per year for each region was calculated, and a subset of voyages where logs are available was also calculated. The number of logs in separate holding locations was also calculated in order to analyse the distribution of logs worldwide. These results are analysed in section 3.0.

3. Results

The results of the Log of Logs analysis show the numbers of ship logs (potential meteorological data sources) for each region (Figure 3). Figure 4 shows locations where logbooks are held. The three regions of interest are considered in detail below. It is worthwhile to note that these results show the number of ships with narratives that visited the regions of interest, not the total number of ships. Because only ships with narratives were documented in the Log of Logs, it is likely that there were more ships with no narratives (or that were inadvertently omitted from Log of Logs) that entered the region.

Figure 3.

Number of ships that passed through New Zealand, Southeast Australia, and the Southwest Pacific waters from Log of Logs keyword searches, 1788–1900, with all potential records prior to 1830 and records ≥1 year post-1830. The vertical dashed line indicates records pre- and post-1830. Note the different scales for the Southwest Pacific results.

Figure 4.

Number of logs held by different institutions, for ships that passed through New Zealand, Southeast Australia, and Southwest Pacific waters between 1788 and 1900. Numbers of logs are shown on a logarithmic scale. Coloured bars indicate country where source is held, and acronyms are defined in the appendix.

3.1. New Zealand results

The keyword searches of the Log of Logs found that between 1788 and 1900, 571 voyages into New Zealand waters were recorded (Figure 3). Many ships made multiple voyages (e.g. postal service ships, immigrant transport ships). The number of ships in New Zealand waters showed a net increase throughout the 19th Century until 1878. There was a strong peak of ship activity in New Zealand waters in 1878, with a total of 112 ships recorded in the region. After 1878, the number of ships recorded in New Zealand waters, as documented in the Log of Logs, rapidly declined to as low as 24 by 1900. Figure 3 also shows the number of ship records that have logs, as noted in the Log of Logs. It is evident that there are many less logs in existence at present than the total number of ships passing through New Zealand, as recorded in Log of Logs. There is a peak of 28 ships with logbooks in 1865. Before 1820, there are no more than three ship logbooks available for any year.

The locations of logs for New Zealand ships are shown in Figure 4. (See definitions of acronyms in Appendix.) Most of the logs are held in Australian and UK institutions, with fewer held in New Zealand and the United States. It should be noted that many logs are duplicated in different holding institutions (see appendix/MS Excel metadata). All Royal Navy remark books were copied by the Australian Joint Copy Project (AJCP) and are available on microfilm in the National Library of Australia.

3.2. Southeast Australia results

The keyword searches concerning Southeast Australia found that between 17861 and 1900, 1724 voyages into the Southeast Australia region retain some form of narrative (Figure 3). Figure 3 shows two distinct peaks in voyages into Southeast Australian waters, in the 1830s–1840s and 1840s–1850s. There is a peak of voyages in 1853, with 115 ships visiting Southeast Australia that year, according to the Log of Logs. The net increase of logbooks peaks in 1866 with 58 logbooks available for that year, with a subsequent decline to as low as 15 logbooks by 1900.

The location with the highest number of logbooks for ships that visited the Southeast Australia region, according to ships document in the Log of Logs, is Mitchell Library, Sydney (Figure 4). Other Australian sources such as AJCP and other state libraries account for most of the holdings of logbooks. Institutions in the United Kingdom also hold a number of these logbooks.

3.3. Southwest Pacific results

The results from the Southwest Pacific Log of Logs analysis show a different pattern to that observed for New Zealand and Southeast Australia. 586 different ships voyaged through the Southwest Pacific between 1788 and 1900, on 664 trips. Figure 3 shows two distinct peaks in the number of ships in Southwest Pacific waters, the first peak in the 1840s and the second in the 1880s. The greatest number of ships in the Southwest Pacific, according to the Log of Logs was 45, achieved in 1884. Like the New Zealand and Southeast Australia results, the number of ships with logs is much lower than the total number of ships (Figure 3). The year with the greatest number of logs is 1844 (with 25 logs).

The locations of logs for the Southwest Pacific are shown in Figure 4. Most of the logs are held in Australia or the United Kingdom, with less held in New Zealand and the United States. As stated above, there are many duplicates of logs in different sources.

4. Discussion

4.1. Limitations of using the Log of Logs

Although the Log of Logs is a significant compilation of ship logbook data for the three regions of interest for the Southwest Pacific, there are some limitations of only using the Log of Logs to find logbooks for meteorological data rescue. Because the Log of Logs was compiled by an historian, certain details important to climate and weather research may have been omitted. These issues are outlined below.

Ships that passed through the region of interest may have been omitted from the results of this study because only part of their journey was detailed in Log of Logs (e.g. the HMS Beagle passed through Tahiti on a voyage with Charles Darwin, but this is not stated in the Log of Logs, so this ship was not recorded in the Southwest Pacific results). In addition, because not every location where ships passed through was recorded in the Log of Logs, some locations within the target regions may have been omitted from analysis. This is because most of the keywords used for searches were obtained from within the Log of Logs. Augmentation of possible sites where data were recorded may be achieved by plotting ship track information (see Brohan et al. (2009) for examples).

The Log of Logs does not indicate which ship records have meteorological data and which do not. Inevitably, the period of time or place of interest will dictate further examination of possible data resources. Meteorological data could also be contained in media other than ship logbooks (for example, Navy remark books and captain's diaries). When analysing the Log of Logs, these three media were considered as ‘logs’ and potential sources of meteorological data (i.e. they were included in this analysis). However, other media (e.g. passenger diaries, surgeon's journals), could also contain useful meteorological data. These sources were deemed less likely to contain rigorously recorded meteorological data (in extensive time series length), so in this case were not prioritized for analysis.

In the different volumes of the Log of Logs, ships with the same name occur with similar or identical voyage dates. However, it is difficult to know whether these ships are the same ship, or whether they coincidentally have the same name. Because volumes two and three update the preceding volumes, it is unclear if ships/voyages mentioned in more than one volume are the same or different ships/voyages. In this study, the cautious approach was taken whereby if a ship record was under the same name, but different dates in the different volumes, all entries were recorded. If it was established that the ship's name, dates, and voyage were identical in the different volumes, then only the sources of logbooks were updated. Therefore, there may be duplicate ships in the results.

Long meteorological datasets from one location are the most desirable when prioritizing logbooks to rescue and digitize. Equally valuable are meteorological observations that were consistently recorded with time and location details. In general, the Log of Logs does not indicate the length of time ships were anchored in one location, or if they were in transit (e.g. carrying out surveys in Torres Strait area vs passing through the Strait on a journey from India to the Pacific). As such, further analysis is required, whereby rescued logs will be examined to verify the time spent in different locations, to further prioritize each log for analysis.

Because it was difficult to scan the Log of Logs volumes completely flat (each volume was ~5 cm thick with stiff binding), some words on the PDF files were on an angle, and therefore OCR could not identify them when carrying out the keyword searches. Therefore, some ships may have been omitted from the results because the keyword searches did not identify them as relevant to the regions of interest.

Some places had different names in the past (for example, Vanuatu was known as New Hebrides, and Tasmania was known as Van Diemen's Land until 1856). Also, spelling variations are apparent in the Log of Logs, where place names were spelt differently in the past (for example, Fiji was sometimes called Feejee). Some of these names may have been used in Log of Logs, but may not have been recognized in this study as being locations inside the regions of interest. Therefore, inadvertent omissions of locations and ships may have occurred. Where an unknown place name was found during the keyword searches, online search engines were consulted for information. Much of the Log of Logs was written in shorthand and used abbreviations (e.g. Syd =Sydney, Auck = Auckland, LHI = Lord Howe Island). These abbreviations were also deciphered and included in the keyword searches. As such, the general traits of the number of ships and possible logs should be a robust indicator of temporal availability.

Since the third volume of the Log of Logs was published in 1999, logbooks held by institutions indicated in Figure 4 may have been moved. This may create some problems in utilizing the results of this analysis to locate and digitize logbooks, however, given a total of 1601 (including duplicates) were listed, many will be found. In addition, some institutions have shifted locations or changed names. For example, the sources New Zealand National Archives and National Archives Wellington are now named an umbrella institution called Archives New Zealand. Name changes that occurred since 1999 are listed in the appendix.

4.2. Future directions

While this study provides a preliminary catalogue of potential historical climate and meteorological data sources from ship logbooks, future work will seek to prioritize and assess the worth of some sources identified.

  • Logs that cover data-thin years on land will be targeted first for rescue. This will involve searching for the log via online catalogues for the source locations (e.g. library catalogues). If this is unsuccessful or if there is no online catalogue available, direct contact with the institution will be made.
  • A preliminary search of some New Zealand libraries has shown that a number of logs, which did not have a New Zealand source cited in the Log of Logs are actually contained within these libraries. For example, copies of some AJCP microfilms are available at the Alexander Turnbull Library in Wellington, New Zealand. Elimination of duplicates will be undertaken, or gathered when opportunistic.
  • Once logs have been located and checked if they contain meteorological data, the researchers will contact leaders of projects such as ICOADS, RECLAIM, CLIWOC, ACRE, SEARCH, and the 20th century reanalysis project [project extension known as the Sparse Input Reanalysis for Climate Applications (SIRCA)], to gain guidance on rescue prioritization to ensure that the logs located have not already been digitized and entered into those databases. A preliminary comparison between potential sources identified in the Log of Logs analysis and logs identified by the RECLAIM project shows an overlap (105 years analysed, 18.8% mean overlap, 9.1% standard deviation), but a large number of logs that were mutually exclusive (Figure 5).
  • A large region of Australia (Queensland, Western Australia, and Northern Territory) was not analysed but has potential to be assessed in the same way the three regions have been in this analysis. In particular, an analysis of Queensland could reveal useful information about past tropical cyclone behaviour.
  • If the located logs contain meteorological data, they will be digitally scanned and transcribed. Following transcription and error checking, the data will undergo homogeneity testing, if possible. Eventually, it is envisaged that a suite of rescued historical climate logbook data will be provided to historical climate reconstruction projects (e.g. SIRCA, ICOADS, CLIWOC, ACRE, and RECLAIM, SEARCH). Thus far, the Southeast Australian analysis presented in this study has been provided to SEARCH.
Figure 5.

Intercomparison of Log of Logs (LoL) and RECLAIM archive searches for logbooks covering Australasian sub-regions.

5. Summary

The research presented here is the first part of an ongoing initiative driven by ACRE Pacific (the Australasian and Southwest Pacific branch of ACRE) to identify and recover ship logbooks across or in New Zealand, Southeast Australia, and the Southwest Pacific. The major purpose of obtaining historical meteorological data is to employ it in climate reconstructions and the extended reanalysis projects. The three Log of Logs volumes were searched to obtain a list of ships that voyaged through New Zealand, Southeast Australia, and the Southwest Pacific, and indicate far fewer logs are in existence today than the number of ships that voyaged to the target regions. However, the logs that are available have the potential to provide meteorological data for locations and time periods where no or limited data exist in historical climate reconstruction databases (i.e. land-based measurements). Most of the logs are held in institutions located in the United Kingdom and Australia, with some in New Zealand and the United States. Certain years are logical to prioritize for logbook rescue, while some years have no or limited data. While there are a number of issues with the approach of using the Log of Logs to locate relevant logbooks, a valuable number of resources along with metadata about the archive location and the spatio-temporal distribution of possible non-land-based meteorological observations across much of Australasia and the Southwest Pacific have been summarized in this work. While this study is expected to ease the process of data rescue in forthcoming efforts to augment and extend the globally crucial reanalysis without radiosondes weather and climate reconstructions, we expect the Southwest Pacific island data sources can potentially be assessed, rescued, and used to extend important regional climate indices (i.e. Southern Oscillation Index, Trenberth Indices, SAM Index, IPO Index, etc.) back in time.


We thank Dr Rob Allan of the UK Met Office for identifying the Log of Logs as a potential source of information for past Australasian and Southwest Pacific climate and weather data, and for encouraging this analysis. This work was supported in part by the Government of New Caledonia and the French Pacific Fund [Recovery, safeguarding, and dissemination of historic French meteorological and climatological data: a Pacific component of the international ACRE programme of Atmospheric Circulation Reconstructions over the Earth (Récupération, sauvegarde et diffusion des données historiques françaises de météorologie et de climatologie: une composante Pacifique du programme international ACRE de reconstitution de la circulation atmosphérique terrestre)], the National Oceanic and Atmospheric Administration (support to A. M. L. for the South Pacific Rainfall Atlas), and the New Zealand Ministry for Science and Innovation (core funding to NIWA's National Climate Centre project Climate Present and Past). Dr Yves Lafoy is thanked for his assistance in setting up a bilateral science activity between New Zealand and New Caledonia that supported part of this work.


Table A1. Table of institution acronyms and locations
Institution acronym (as at 2012)Institution acronym/name (in Log of Logs, if different to 2012)Institution name and location (as at 2012)
ACLAuckland Central Library, NZ
AHRRAustralian Historic Records Register (part of NL)
AJCP (including AJCP PRO, AJCP M)Australian Joint Copy Project (PRO = Public Record Office series, M = Miscellaneous series). Held in NL.
ANMMNMM SydAustralian National Maritime Museum, Sydney, NSW
ANZNZNA, Wellington ArchivesArchives New Zealand, Wellington, NZ
AOTTAS ArchivesArchives Office of Tasmania, Hobart, TAS
ATLAlexander Turnbull Library, Wellington, NZ
BLBritish Library, London, UK
BLARBedfordshire Records OfficeBedfordshire and Luton Archives and Records Service, Bedford, UK
BMBritish Museum, London, UK
CBSBuckinghamshire Records OfficeCentre for Buckinghamshire Studies, Aylesbury, UK
CLCrowther LibCrowther Library, Hobart, TAS
CMCant MusCanterbury Museum, Christchurch, NZ
DLDixson Library, Sydney (part of State Library of NSW)
Fisher LibraryFisher Library, University of Sydney, NSW
FNAParis National Archives du MarineFrench National Archives, Paris
HLHocken Library, Dunedin, NZ
HRAHistorical Records of Australia database
India Office RecordsIndia Office Records, British Library, London, UK
JOLOLJohn Oxley Library, Brisbane, QLD (part of State Library of QLD).
LTLLa Trobe Library, Melbourne, VIC
MLMitchell Library, Sydney (part of State Library of NSW)
MOD Library, UKMinistry of Defence Library, UK
NAANational Archives of Australia, Canberra, ACT
Nat Met Archive UKNational Meteorological Archive, Exeter, UK
NBWMNew Bedford Whaling Museum, MA, USA
NHMNatural History Museum, London, UK
NLNational Library of Australia, Canberra, ACT
NMMNational Maritime Museum, Greenwich, UK
NWMNantucket Whaling Museum, MA, USA
OSMOESMOtago Settlers Museum, Dunedin, NZ
PEMEssex Institute MuseumPeabody Essex Museum, Salem, MA, USA
PLRIProvidence Library, RI, USA
PMBPacific Manuscripts Bureau, Canberra, ACT
PRO LondonPublic Record Office, London, UK
PROVPRO VICPublic Record Office Victoria, Melbourne, VIC
RGS LondonRoyal Geographical Society, London, UK
RHSQRoyal Historical Society of Queensland, Brisbane, QLD
RHSVRoyal Historical Society of Victoria, Melbourne, VIC
RN HydroRoyal Navy Hydrographic Department, Taunton UK (now the UK Hydrographic Office)
RN Library UKRoyal Navy Museum Library, Portsmouth, UK
RNMRN MuseumRoyal Navy Museum, Portsmouth, UK
Royal Society LondonRoyal Society of London, London, UK
SAMMSouth Australian Maritime Museum, Adelaide, SA
SLSAState Library of South Australia, Adelaide, SA
SLVState Library of Victoria, Melbourne, VIC
SPRIScott Polar Research Institute, University of Cambridge, UK
SRNSWPRO NSWState Records Office of New South Wales, Sydney, NSW
SRSAPRO SAState Records of South Australia, Adelaide, SA
TAS MMMaritime Museum of Tasmania, Hobart, TAS
UHUniversity of Hawaii
USNAUS National Archives, Washington DC
WAMMWestern Australian Maritime Museum, Fremantle, WA
WMMWellington Maritime Museum, NZ


  1. 1

    While Nicholson's Log of Logs covers 1788–1999, the First Fleet voyages extend to 1786.