Rain Read online

  Fitzroy hired Symons as an assistant in 1860. But the young man was so obsessed with his British Rainfall Organisation that Fitzroy felt it took away from his official duties. Symons lasted only three years with the government’s meteorological office. Fitzroy did not last much longer. He committed suicide in 1865 by slashing his throat with a razor. Fellow scientists felt sure it was the pressure of trying to make accurate weather predictions in the face of constant criticism and ridicule. His suicide sharpened the view that forecasting was an immoral pseudo-science. It helped lead to a ban on any public forecasts in England for the next thirteen years, allegedly because of inaccuracies. Darker motives were also at work. For one, the large ship-salvage companies of Cornwall and Devon complained to Parliament that the forecasts were putting them out of business.

  Symons went on to collect and publish his rain data privately, which may have been the wiser route given the birthing pains at the British Meteorological Office, known today as the Met. He placed newspaper ads for rainfall observers in every corner of the British Isles. He offered to buy the instruments and train observers “of both sexes, all ages, and all classes.” By 1865, the British Rainfall Organisation had a thousand reporting stations, by 1876 two thousand, and by 1898 three thousand. The dedicated volunteers checked their gauges at 9 a.m. every day, logged inches of rainfall on charts supplied by Symons, and sent them in once or twice a year.

  Symons also dug up every historical record he could find, appealing for ships’ logs and old weather journals so that he could reconstruct past weather. He ultimately gathered some seven thousand sets of records that let him build a reliable picture of rainfall all the way back to 1815.

  With kind, crinkly eyes and a full Victorian beard, Symons was “a man of singularly genial manner making a friend of almost everyone with whom he came in contact, even those with whom he differed,” a fellow meteorologist said of him. He maintained a patient and masterful correspondence with his rain army, sometimes cajoling, reprimanding, or polling the observers on various weather questions. He painstakingly recorded their thousands of annual readings, with town, observer’s name, elevation, and total inches, in his annual publication British Rainfall.

  The loyal rain-gauge readers in turn helped fund the organization with donations and subscriptions to his magazines. In addition to the measurements, British Rainfall carried anomalies and records from the year and exhaustive explanations of how to measure rain—on different terrains, for light rains, snows, and heavy rains, and how to note the total in decimal points by hundredths of an inch: “Vulgar fractions should never be employed.”

  Another publication, Symons’s Monthly Meteorological Magazine, covered an astonishing variety of phenomena. The first volume, in 1866, ran features on a “new enemy to rain gauges”—the leaf-cutter bee; meteor showers; the migration of swallows; various floods and droughts; and an investigation into black rains, which we will encounter later.

  By 1870, Fitzroy’s successors realized what the Met had lost in Symons and his rain obsession. But Symons rejected offers to bring his network back to the public sphere. He told his observers their important rain work would not be pushed into “an obscure corner in some Government office.” Such a move would undermine his own “enormous expenditure of money, time and physical and mental energy,” and moreover mean the rain group’s “esprit du corps would be extinguished.”

  When Symons died in 1900, he left no survivors, having lost his father when he was a young boy, then his mother, his only child in infancy, and his wife fifteen years before. But he was mourned by a family of thousands of fellow rain watchers. By all accounts he died a happy man, for having devoted his life to British rainfall.

  —

  Americans did not have the same qualms as the Victorian Brits about the voodoo of weather forecasting. Thomas Jefferson dreamed of setting up stations of weather instruments and deputies around the nation, but he was foiled by the Revolutionary War. Later, the telegraph made scientific forecasting tangible. Then and now, the best way to know how the weather is going to behave at your place is to ask its most recent host. In 1847—three years after Samuel F. B. Morse electrically transmitted his famous message, “What hath God wrought?” from Washington to Baltimore—America’s first government meteorologist, the “Storm King,” James Pollard Espy, pitched the idea of a national weather network connected by telegraph lines. Espy sold the notion to Joseph Henry, director of the brand-new Smithsonian Institution, who sold it to his board by stressing that telegraphing weather from the far western and southern reaches of the nation would “furnish a ready means of warning the more northern and eastern observers to be on the watch for the first appearance of an advancing storm.”

  By 1860, five hundred stations across the United States telegraphed weather reports to Washington. War, again, set back the effort; the network crumbled after the secession of the southern states. At the Cincinnati Observatory in Ohio, a young astronomer named Cleveland Abbe became so frustrated by the lack of public storm and flood warnings that he took on forecasting as a personal mission. With funding from the Cincinnati Chamber of Commerce, Abbe developed a system of telegraphic weather reports, daily weather maps, and predictions he compiled and shared via the Western Union Telegraph Company.

  Western Union supplied outline maps, to which Abbe or a telegraph clerk could add symbols showing wind direction, areas of high and low pressure, and other details. Abbe designed the familiar triangular arrows and codes on hand-drawn forecast maps, with the “R” for rain. For the first time, Americans could read not only current weather conditions, but “probabilities,” as Abbe called his forecasts, for the days to come.

  Abbe’s service was immediately popular. Only thirty years old, he became known as “Old Probabilities” or “Old Prob.” It also became clear that forecasting could save many lives and ships. A petition from the Great Lakes region—which suffered 1,914 shipwrecks in 1869 alone—urged Congress to establish a meteorology agency and national telegraphic weather service to track “the origin and progress of these great storms.”

  Congress approved, and in early 1870, President Ulysses S. Grant signed a resolution that the secretary of war should head up meteorological observations in military stations across the nation, as well as “notice on the northern lakes and seacoasts, by magnetic telegraph and marine signals, of the approach and force of storms.” Gen. Albert J. Myer, the army’s chief signal officer and a personal friend of Grant, had pulled strings to secure weather in the War Department. Now, instead of shrinking in peacetime, the Signal Office could grow. Myer turned out to be as dedicated to forecasting as to devising the signal system. He quickly hired Abbe away from the Cincinnati Observatory. Abbe cranked out his daily Probabilities and trained a generation of meteorologists in the art of forecasting. Myer’s signaling, developed for war, now grew into a system of flags that flew in cities and seaports to warn of coming storms. When he died in 1880, one newspaper tribute said “no careful seaman ventures out of port when the red light is burning or the red flag flying.”

  To this day, if you see a red flag flying with a black square in the center, be warned of the storm to come—and say a little thanks to General Myer and the U.S. Signal Service.

  —

  In a storm-gray federal government building in downtown Asheville, North Carolina, 160 scientists and other employees of NOAA’s National Climatic Data Center work with weather records as new as satellite images transmitted a minute ago, and as old as the 1740s. The 20 million pages of observations, many handwritten, and billions more computerized records here represent the largest archive of weather data anywhere in the world. Logs of ship captains and military officers reporting in from around the globe are filed twenty-five feet high in the basement. “Your faithful servant,” many of them sign off, in elegant, rounded script.

  “A dash of rain” is an oft-repeated entry from the nineteenth century, the measure then as common as our dash of salt.

  “The rain was dr
iven with the force of arrows into my face, and the oppression was similar to what one feels riding on a fast horse at a riding pace,” reads a harsh 1831 observation by the naval engineer William Redfield of New York, who subjected himself to storms as he worked to figure out the circular nature of cyclones.

  If there is a common link in rain science past and present, it is a certain mystification. For countless reasons beyond the storm warnings that save lives, we humans have an insatiable will to divine rain—to manage the water supply of major cities, to figure out when to plant winter wheat, to plan an outdoor concert or choose a wedding date. But even as modern meteorology improves upon the old probabilities, rain is eminently difficult to predict or to count on. It’s the classic example of chaos theory; as pontificated by Jeff Goldblum to Laura Dern in Jurassic Park: “A butterfly can flap its wings in Peking, and in Central Park you get rain instead of sunshine.” The real scientist behind the butterfly theory was the late MIT meteorologist Edward Lorenz, the first to recognize that tiny, faraway triggers can change the weather in ways that make it impossible for mathematical models to predict. Your weather app can’t tell you that it’s about to rain old women and walking sticks in front of your house—while your backyard remains completely dry. It may not have predicted any rain at all.

  Meteorologist Scott Stephens, a child cloud-watcher who knew his life’s work by the time he was five years old, is charged with answering the National Climatic Data Center’s public requests for weather knowledge past and predicted. Beyond the billion hits on the agency’s website, he fields calls from police detectives and insurance investigators who need to know how much rain fell at the hour of a crime or accident scene; dam engineers seeking annual averages; construction bosses deciding whether to hire a crew; and the occasional crank who has figured out how to control hurricanes. (We’ll get to that a little later.) These and the ten thousand other interested parties who call the center every year could find the historical data and predictions online, but they want the human insight. “Weather models do a great job forecasting temperatures, they do a great job of forecasting wind speed and direction,” Stephens tells me. “The models have a more difficult time with precipitation.”

  This is true even in the Big Data age of supercomputers that crunch billions of global weather readings every day—from Earth-orbiting satellites; radars; ground sensors at thousands of stations; thousands more buoys and ships at sea; aircraft weather chasers; and one thousand beige weather balloons with white boxes known as radiosondes attached, launched every morning and afternoon from points around the world. (If ever you stumble upon a large, deflated balloon tethered to a plain white box, possibly making disconcerting noises and giving off a sulfuric stink, don’t be alarmed, and return to sender in the postage-paid mailer inside.)

  In the late 1940s, the military gave the Weather Bureau what became one of its most valuable tools for predicting rainfall—25 surplus radars, launching the age of weather surveillance. That network has grown to include 155 Doppler radars that let meteorologists peer into the clouds, and LIDAR, remote-sensing laser beams that can simulate floods. The United States and Japan jointly operate a satellite mission dedicated to measuring global rainfall. All of this data pours into NOAA’s superspeed weather-prediction computers, which can crunch trillions of calculations a second, making forecasting more accurate all the time. Today’s four-day rain outlook is as accurate as the one-day forecast of thirty years ago. The satellites and supercomputers have particularly sharpened forecasts for tropical storms; the National Hurricane Center nailed Hurricane Sandy’s southern New Jersey landfall five days out.

  Yet rain continues to defy Big Data—routinely washing out zeropercent predictions as well as the pronouncement by the former Wired magazine editor Chris Anderson of “a world where massive amounts of data and applied mathematics replace every other tool that might be brought to bear.” What’s fascinating about high-tech rain forecasts is the degree to which they are improved by human meteorologists like Stephens. National Weather Service statistics show meteorologists improve rain-forecast accuracy 25 percent over computer guidance alone.

  —

  Just as computer rain forecasts still need the instincts of human meteorologists, they benefit from the physical collection of rain—measured in gauges around the world—over radar and satellite imagery alone. For the Colorado state climatologist Nolan Doesken, this truth hit achingly close to home nearly twenty years ago. The National Weather Service’s Nexrad (next-generation Doppler weather radar) failed to detect freakish rainfall variation in a storm that developed directly over Fort Collins, where Doesken lives. On July 28, 1997, many residents were lulled to sleep by the sound of rain drumming on rooftops, with no warning of what was to come. Before midnight, the storm poured two inches of rain over much of the region, but hurtled fourteen inches onto southwest Fort Collins. The rains swelled docile Spring Creek to a flash flood that swamped homes and sent families scrambling up trees and rooftops in their pajamas. Five people were washed to their deaths. Doesken has never forgiven himself for not making a call that night to the Weather Service; he never imagined the radar was not picking up such fierce rains.

  The disaster led Doesken to launch a network of volunteer rainfall observers, known as the Community Collaborative Rain, Hail and Snow Network, who take rain gauge readings at home and report them over the Internet. What began as a small local project has grown to 30,000 weather watchers around the United States who are building a database of highly local precipitation measurement. (About half of them consistently check their gauges every morning.) The physical data has proven invaluable for scientists, especially when it comes to rain’s variability. A Texas volunteer once reported seven inches of rain in his Comal County gauge, while nearby volunteers reported no rain or a couple hundredths of an inch. Meteorologists were sure it was a typo until they looped back through the radar to find a tiny, short-lived convective storm that had formed and died in the same spot, circling the dedicated rain volunteer like the rain clouds chasing lorry driver Rob McKenna.

  In April 2014, a storm carrying more rain than had hit the region in any hurricane of the past century swamped Florida’s Panhandle, turning parts of Pensacola into sea and eating away chunks of the scenic coastal highway. The storm knocked out power to the official weather station at the Pensacola airport, but a number of the network’s volunteers corroborated the extreme rainfall—reporting as much as 20 inches in one day—allowing scientists to confirm a new record. “The high-tech stuff will increasingly win the day,” Doesken told me, “but it will win it better if it’s ground-truthed” with credible measurement of rain falling in the backyards of tens of thousands of citizen scientists.

  Doesken is a modern-day G. J. Symons (with salt-and-pepper mustache rather than Victorian beard), corresponding with his thousands of volunteers and meeting them in person when he travels. Most are older than fifty-five and so committed to the work that their families contact Doesken when they pass away. He writes back letters of condolence, not a duty he expected when he became a meteorologist. “These volunteers don’t realize just how important their records can be,” Doesken says. “Most of them find it interesting, but they don’t realize that it can also be lifesaving.”

  —

  Because we so urgently need to know what they know—every day, or even every hour—meteorologists have had to share their knowledge more publicly than most experts. From the earliest days of Fitzroy’s forecasts in Britain and Abbe’s Probabilities in America, a more parasitic group of weather watchers stuck close to the meteorologists like barnacles. Journalists were eager for weather news, the number one interest of most of their newspaper readers then; readers, listeners, viewers, and Internet users now.

  My colleague Bill Kovarik, a professor of journalism and environmental history at Unity College in Maine, says early American journalists devoted ink to the weather long before the U.S. Weather Bureau existed. The Niles’ Weekly Register, the most popular
publication in the nation before the New York Times debuted in 1851, and then the Times, were both committed to weather reporting in the nineteenth century. Niles’ Weekly ran lengthy reports on the atmosphere in its science section; in 1849, the magazine published an analysis of national rainfall patterns atop an update on aerial navigation and another on “the importance of fresh air.” The Times was covering weather regularly by 1857, and had a daily weather report by the 1870s.

  The addition of a newspaper weather map was a coup for newsmen who wanted to hook their readers, and weathermen who wanted their forecasts to reach a larger audience. With the help of the telegraph, meteorologists began making national weather maps in 1848; Joseph Henry hung one up for visitors each morning on the wall of the Smithsonian. In 1875 The Times of London became the first paper to print one daily. The New York Times began running one in 1934, and the next year, the Associated Press started to transmit a national map to its member papers. The early newspaper weather maps contained much more meteorological science than today’s: isotherms—the lines that connect points with the same temperature—to help readers see where weather fronts were moving in; Abbe’s little arrows to show direction; areas of high and low pressure. Like most things in newspapers, the maps were dumbed down steadily over the century, until they carried little besides temperature—and of course, the rain.

  —

  By 1900, the U.S. Weather Bureau had moved from the War Department to the Department of Agriculture, which helped set up the first radio weather broadcasts at the University of Wisconsin in Madison. Most fledgling radio stations shuttered during World War I, but UW’s stayed on air, broadcasting weather news to ships sailing on the Great Lakes. During the New Deal, the Weather Bureau got much more involved with local radio, beginning a long history of love-hate relationships between the agency’s meteorologists and local broadcasters who shared or hyped their forecasts.