IT WAS a defining moment in British history. The Royal Navy’s victory over the French and Spanish fleets at the battle of Trafalgar elevated Admiral Nelson from national hero to legend, and gave its name to London’s grandest square and to hundreds of pubs up and down the country. Two centuries later, it has provided climatologist Dennis Wheeler with a victory of his own. The logbooks of the ships of Nelson’s fleet have helped him prove that seafarers’ observations of wind and weather, made long before ships carried reliable meteorological instruments, can give us a precise, daily record of the weather hundreds of years ago. That could be a key to understanding what is happening to the world’s climate today.
Climatologists and politicians are desperate to know whether the climatic changes the world is seeing now really are unprecedented, or whether there have been other times when change was just as swift. “If we go back to the past and find there really hasn’t been anything like this before, then that strengthens the case that human activity is responsible for the rapid changes we are seeing now,” says Wheeler, who shuttles between the University of Sunderland in north-east England and the National Maritime Museum at Greenwich in London.
But the past is tricky territory for climatologists. While tree rings and ice cores chart annual changes in climate going back millennia, they cannot show what happened from one month to the next. There are a few long-running records of temperature and air pressure dating back to the 17th century, but reliable data about air movements over the oceans – which govern most of our weather – were almost non-existent before 1850, by which time the industrial revolution was well under way and factory chimneys were pouring carbon dioxide into the atmosphere. To find out how human activity is altering Earth’s climate, we need detailed records from earlier times. Wheeler is convinced that the multitude of jottings in old ships’ logs can provide them as far back as the mid-1700s, and perhaps even earlier.
Advertisement
By the late 17th century, the great seafaring nations of Europe had fleets of ships criss-crossing the seas, all keeping meticulous accounts of their voyages. In 1731, it became compulsory on British naval vessels to keep a log, and by 1750 the practice was universal. Hundreds of thousands of these logs survive in archives in the UK, Spain, the Netherlands and France. While they do contain occasional records of momentous events – a sea battle, an exotic landfall, perhaps a mutiny – for the most part logs are filled with observations of wind and weather. Unlike tree rings and ice cores, these are first-hand accounts of what the weather was like on a particular day that could reveal in unprecedented detail how wind patterns varied, rainfall fluctuated and the northern sea ice advanced and retreated. But is the data good enough to reconstruct past climate?
Sailors had every reason to make accurate observations, not least because they were vital for navigation at a time when few ships had the means to determine longitude with any accuracy. Officers needed to know the force and direction of the wind to calculate their speed and location. “If they got it wrong then they had big problems,” says Wheeler. On every watch – once every 4 hours – or sometimes more often still, they checked the strength of the wind and its direction, and recorded other observations about the weather: the state of the sea, whether it was fine or foggy, wet or sunny, or if there were icebergs on the horizon.
Historians set great store by seafarers’ records, but scientists have always been reluctant to use them. “There is still prejudice against using historical documents in science,” Wheeler says. Though there are some poorly kept logs, and sometimes confusion over the author (see “Ghost writers”), he considers the prejudice unwarranted. “If you treat the data carefully and verify as much of it as you can there is no reason why you can’t use historical information for profound scientific research.”
In 2000, Wheeler began a project that had teams from the Netherlands, Spain and Argentina – where many Spanish logs were left after the collapse of the Spanish empire – sifting through 2000 logbooks. With funding from the European Union, the researchers began to create a climatological database for the world’s oceans (CLIWOC) from 1750 to 1850, by which time ships were equipped with instruments to collect weather data.
Fair wind or foul
To reconstruct past climate, modellers need to know details of wind strength and direction at as many locations as possible. These allow them to reconstruct air pressure fields and track highs and lows as they move across the sea. The CLIWOC teams pored over the logs, picking out daily observations of a ship’s position, wind direction and strength, plus any general observations about the weather. The result was the equivalent of 2 million days of information. “You could see the wind backing or veering over the course of a day. In fact there was too much data, so we restricted ourselves to the midday observations,” says Clive Wilkinson of the Greenwich museum, the project’s research officer.
But plentiful as it is, this data needed translation. Until 1805, even a day wasn’t a day as we know it but a nautical day, which started at noon. Dates needed to be treated carefully too. Before 1752, the year began on 25 March, and until September that year the English calendar lagged 10 days behind the one used in the rest of Europe.
Even the wind’s direction posed a problem. It took a long time to work out if the direction noted in the log was based on magnetic or true north. Get it wrong and the data could be out by as much as 30 degrees. On British vessels, most observations were magnetic. “Once a day, officers would convert the readings to true in order to plot the ship’s true course on the map – but the logbooks don’t tell you that,” says Wheeler. “It took some detective work to find out what was going on.”
Harder still was interpreting observations on the strength of the wind. Without instruments, officers had to rely on their judgement and experience. They did not write clear, unambiguous numbers in their logs, but words – and often archaic or peculiar ones. The Beaufort scale of wind strength did not become official until 1838. Before that, officers had a rich and colourful vocabulary to describe the wind. What precisely did they mean?
The earliest logbooks were the most perplexing. “Before the 1750s, anything above a calm was a gale of some sort,” says Wheeler. “There were no breezes; the term came later.” Descriptions of these “gales” were many and varied: soft or feint, strong or hard, fair and brave, handsome and even indifferent. A sample of English logs from 1685 to 1700 produced 73 different ways of describing the wind’s strength. To everyone’s relief, closer scrutiny showed that many terms were used just once or twice: most seafarers stuck to a similar set of terms.
So how strong was a wind described as “moderate” or a gale described as “fine”? And did all seafarers have the same thing in mind when they used them? This can be tested by checking the distance ships sailed each day, another fact carefully recorded in the logbook. The speed a sailing ship travelled was directly related to the force of the wind, at least up to Beaufort force 7, so the stronger the wind, the further the distance “made good” each day. “If you have a large enough sample you can pick out all the days when there was a ‘small gale’ or a ‘fine gale’ and check the distance covered,” says Wheeler. “That gives a measure of the relative wind strength.”
The wealth of data Wheeler was able to call on allowed him to show statistically that these terms were not vague approximations but fell into distinct categories. The researchers have now put a Beaufort number to almost all the terms used by 18th-century sailors, and even checked their data against one talented sailor’s paintings of the sea (see “Art and science”).
Having to deal with logs from four nations, from both naval ships and merchantmen, made the task of translating archaic terms into hard numbers doubly daunting. The Spanish seafarers, like the English centuries before, had a multitude of often poetic words to describe the wind force, although most officers seemed to stick to just a few terms. The French too, mostly limited themselves to a small set of phrases. But the Dutch had a different system, describing wind strength in terms of the amount of sail they were carrying: as the wind blew harder, they began lowering sails, starting with the topmost. Even tougher to unravel were the terms used aboard merchantmen. “The East Indiamen’s vocabulary had a more complicated range of gales and winds, trade and monsoons, each in several grades,” says Wilkinson. By checking distances sailed, and cross-checking with the logs of occasional naval escorts, the team eventually worked out which terms meant what. Pairs of ships in convoy, or squadrons sailing in the same area, should have the same records. “If there are enough of them you can check statistically for consistency,” says Wheeler. And this is where Nelson’s famous victory became so helpful.
The British fleet at Trafalgar presented the perfect opportunity to cross-check the records. In October 1805, Nelson had 30 fighting ships and support vessels in the Atlantic off south-west Spain, each with a captain and several lieutenants all keeping logs. Taking a sample of 13 ships, Wheeler picked out records of wind force, direction and weather for each day that month. Except for the day of battle, when officers had more pressing concerns, the entries were in agreement for around 70 per cent of the time.
With Wheeler’s early hopes for the data’s potential confirmed, it is now up to the climate modellers to see what they can make of it. At the University of East Anglia in the UK, Phil Jones and his team have begun to use the data to reconstruct wind flow and air pressure over the North and South Atlantic and the Indian Ocean. Data for the Pacific, where only the Spanish had an interest, is too sparse to be very useful. Jones’s team expects eventually to produce synoptic weather maps for every month between 1750 and 1850. “You can reconstruct the weather month by month – daily where there’s enough data. You can look at seasonal cycles and differences between winter and summer,” he says. “We will be able to see how different weather patterns were then, and look at the general circulation to see if it’s very different from recent patterns.”
Rebuilding the past
Jones is particularly interested in reconstructing the behaviour of the North Atlantic Oscillation, El Niño’s less well-known Atlantic cousin. The NAO determines whether Europe’s winters are mild or frigid, wet or dry, and how the weather changes between southern and northern Europe, leaving some areas drought-ridden but others deluged.
Global warming sceptics like to think that the vagaries of the NAO are responsible for much of the warming ascribed to greenhouse gases. Many climate scientists, however, see things the other way around and are concerned that global warming might be interfering with the NAO’s cycles. Data from ships’ logs might help to settle the matter.
Like Wheeler, Jones can check the accuracy of his reconstructions because they have a good idea of the sort of atmospheric circulation required to cause certain events. “We know what sort of pressure patterns were responsible for 1816, the ‘year without a summer’,” Jones says. “And we have a rough idea what sort of circulation causes very cold winters such as those of 1795 and 1812, when the Thames froze over.” If the reconstructions show those features, Jones can have confidence in his models.
The CLIWOC database will go online next year, opening up access to detailed weather reports from the pre-instrumental age. And there are plenty more where they came from. The Spanish have scoured all the logs in their archives, the Dutch have gone through more than half of theirs, but the UK has hardly made a dent in its great hoard. “There are perhaps as many as 200,000 English logbooks from the age of sail,” says Wilkinson “We’ve looked at about 5 per cent – just the tip of the iceberg.”
Art and Science
Today ships carry a set of flip cards showing photographs of the sea for each step on the Beaufort scale. The old logs record the state of the sea using subjective descriptions that have scientists sucking their teeth with disapproval. But one man’s logbooks left little doubt what he meant when he described the wind and weather.
Nicholas Pocock was master of a snow, a small two-masted trading ship. He was also a brilliant artist, and later became court painter to King George III. He was famous for the accuracy and fine detail in his seascapes. Pocock’s logbooks (above), which span the decade from 1766 to 1776, are works of art. Like other officers, he noted wind and weather conditions as well as the crew’s complaints about the food or the state of the ship’s supplies. But every day he also made a pen-and-ink sketch of what the sea looked like. If they were as accurate as his later paintings, they could verify that a sailor’s description of the sea tallied with the wind force.
Taking the logs of two voyages from Bristol to the West Indies in 1771 and 1772, Wheeler compared Pocock’s drawings with a set of modern flip cards, assigning them a Beaufort number. Only then did he look up Pocock’s estimates of wind force, translated into Beaufort numbers, for those days. “They matched pretty well,” he says.
Ghost writers
Not all ships’ logs are quite what they seem. A few officers, for instance, were lazy and wrote much the same thing every day. “You get a feel for the bad ones and discard them,” says Clive Wilkinson, who has spent long hours in the library of the National Maritime Museum in London examining logbooks. Then there are those who paid others to write their logs. “Some lieutenants seem to have paid the ship’s clerk,” Wilkinson says. “I’ve seen two different logs by two different officers in the same hand.” But that doesn’t make a log suspect, he says. “It just means the ship’s clerk probably had neater handwriting.” Robert Parker, captain of the 64-gun Intrepid, clearly kept a clerk to write up his log. The entry for 22 November 1797, which notes that the captain fell overboard and drowned, was written in the same neat hand as all the previous ones.
There are also occasions when a log changes abruptly mid-voyage. In May 1764, the captain of the sloop Liverpool locked himself in his cabin and shot himself. Lieutenant John Okes had the job of sailing the ship home from the Indian Ocean. “What was a beautifully kept log deteriorated in content and tidiness,” says Wilkinson. “You get the impression that Okes, who was probably a young man, was getting very stressed by his responsibility and the prospect of facing a court of inquiry. He may have found the duty thrust upon him too much – which would explain why he remained a lieutenant until he retired in 1805.”