"A unique historical account of meteorological cartography."—Science News

An excerpt from
Air Apparent
How Meteorologists Learned to Map, Predict, and Dramatize Weather
by Mark Monmonier

American weathercasting began on October 14, 1941, on WNBT-TV, an experimental station (later WNBC) serving at best a few thousand viewers in New York City. Hawking Botany Wrinkle-Proof ties, a cartoon character named Woolly Lamb introduced the forecast by singing, "It's hot. It's cold. It's rain. It's fair. It's all mixed up together. But I, as Botany's Woolly Lamb, predict tomorrow's weather." The forecast that followed consisted of a single screen with several lines of text, but no map. As far as I can tell, the distinction of telecasting the first weather map belongs to the British Broadcasting Corporation, which for a brief period in 1936 transmitted to far fewer viewers simple weather charts with isobars and pressure centers as well as temperature, sky conditions, and precipitation for a handful of unnamed cities in the British Isles and the adjacent continent. As Woolly Lamb presaged the weathercaster's role as an entertainer, the BBC experiment demonstrated television's ability to deliver timely weather maps to a mass audience. Although World War II interrupted the new medium's development, postwar prosperity stimulated a new industry in which news programs with weather reports became a standard feature.

Weathercasting's first full decade could be called the turbulent '50s. Annual sales of millions of new receivers made television attractive to advertisers, whose lucrative support triggered intense competition at local and national levels. With no immediate consensus on how to report weather, broadcasters experimented with formats ranging from terse Weather Bureau reports read by the station-break announcer to mini-lectures by a long-winded meteorology prof from the local university. Where weather was an especially fickle antagonist of farmers or travelers, larger stations hired their own full-time meteorologist, typically a weather service retiree or war veteran. National networks did likewise, following the lead of CBS, where John Clinton Youle reported the weather each weekday evening on the Camel News Caravan. More common, though, was the performer selected less for meteorological savvy than for an engaging personality, good looks, or fondness for comic costumes—an umbrella and slicker for rain, a bathing suit and beach ball for summer sun, and the mandatory parka and snowshoes for anything from light snow to a raging blizzard. Alarmed by ill-informed weather clowns and air-head weather girls, the American Meteorological Society established its Seal of Approval program in 1957 to promote understanding and reliability, if not decorum.

Since 1959, the AMS has certified over 900 weathercasters. Requirements are rigorous—too rigorous for many wannabe's without a degree in meteorology or earth science. In addition to core courses in atmospheric science and mathematics, applicants must submit videotapes of on-camera reports for three consecutive days. The Board of Broadcast Meteorology—its members are working weathercasters—evaluates submissions in four areas: technical and scientific validity, informativeness, explanatory value, and communication skills. In addition to informing viewers about "recent, current, and anticipated weather conditions locally and nationally," seal holders must explain the processes that produced these conditions. Lapses following certification are taken seriously: if a seal holder "clearly fails to conduct himself or herself in a manner that reflects the dignity and honor of the profession or . . . fails repeatedly to adhere to the four criteria," certification can be revoked or suspended.

For many stations, an AMS seal holder on the news team was once a mark of distinction, especially where competing stations lacked the society's implied endorsement. Now though, the specialized formal training required for certification is of little interest to a station manager who can pay a technically adequate but engagingly relaxed performer appreciably less than a card-carrying meteorologist. Small wonder then that a 1985 survey revealed that none of the managers of small-market stations considered the seal of approval "very important," while almost half rated it "not important." Among respondents at large-market stations, which often face keen competition, 20 percent rated the seal very important, but 39 percent strongly disagreed. As evidence that stations managers don't confuse competence with certification, the qualifications most valued at all levels were, in order, knowledge of weather, personality, and broadcast experience.

Many of today's weathercasters began their careers in college. Prestigious research-oriented meteorology departments at Penn State and Florida State University offer at least one course in broadcast meteorology, while smaller schools like Mississippi State University and Vermont's Lyndon State College run media-tailored meteorology programs for undergraduates eager to combine science and show biz. At Mississippi State, for example, students in broadcast meteorology supplement an atmospheric science major with required courses in TV production, acting, and voice and articulation. At both MSU and Lyndon a well-equipped lab linked to the college television station helps students develop forecasting and presentation skills. Even so, the college laboratory is no substitute for the TV newsroom. A few lucky graduates land positions at the Weather Channel, but most fledgling weathercasters begin their careers in low-paying jobs at small-market stations, which have the greatest turnover.

These days competence in weathercasting implies skill in computer graphics as well as an understanding of atmospheric images. It's been that way for nearly two decades—even longer at large-market stations better able to afford new technology. In the 1985 survey, for example, the most prevalent graphics were satellite images, used at 46, 51, and 96 percent of the small-, medium-, and large-market stations responding. Color radar—pre-NEXRAD, of course—was a close second, with corresponding adoption rates of 40, 56, and 79 percent. Less common, except at large-market stations, were computer-generated maps, with rates of 29, 37, and 80 percent, respectively. By the late 1990s, though, any station wanting lively graphics had them: a gift of powerful, low-cost personal computers and fierce competition among software vendors and service bureaus eager to cover every corner of a specialized but profitable market.

If today's TV weather maps are largely automatic, their counterparts in the turbulent '50s were purely manual. With tiny labels and complex symbols, faxed images from the Weather Bureau were unsuitable even for close-ups, and had to be redrawn and simplified. As stage props with a sparse geographic framework of state boundaries, weather maps were huge—about four feet high by five feet wide, typically. The weathercaster stood in front of the map, just to the side of where he or she needed to add symbols or point out features. The more cartographically industrious presenter might draw up several charts in advance and clip them to an easel in sequence (first in front, last in back) so that the drawing just discussed could be quickly ripped off and discarded. Some stations attached maps to panels that could slide up and out of the way on vertical tracks, like windows. (Sets were small, and unless the program director provided a second camera, space and equipment discouraged moving about.) Gary England, who broke into weathercasting in the early 1970s, described his first live broadcast at an Oklahoma station that mounted maps on miniature Ferris wheels:

The metal weather maps on the large four-sided drums somehow looked larger that night. Each drum weighed 180 pounds but felt much heavier. Every time I turned a drum, some the letters and numbers would fall off or would assume a crazy tilt and have to be rearranged. It was frustrating those days, the norm.

Some presenters drew in their map symbols on camera: an impressive act to a home audience unable to see the thin red lines penciled in as guides. Skill as an illustrator was a marketable asset for the weathercaster who could quickly sketch clouds, lightning bolts, or a radiant "Mr. Sun."

Competition for ratings inspired clever embellishments. Artistically challenged performers could polish their presentations with a metal-backed outline map and a set of magnetic stick-on symbols representing highs, lows, fronts, clouds, and suns. Moved about carefully, these symbols provided a crude but sometimes instructive animation. But when coated with material responsive to a rotating source of polarized light, apparent movement within selected symbols mimicked falling rain or diverging solar rays. Equally ingenious was the boundary map painted on a large sheet of plate glass or Plexiglas: a solution that not only intrigued viewers with the picture of a map drawn from behind but also got the performer's hand and arm out of the way of the map's most immediately relevant symbols. When I first saw this ploy, I wondered how the weathercaster trained herself to print letters backward on a map with New England on the left and California on the right. Quite a feat it seemed, until I realized how easily video electronics could reverse left and right.

Among the many electronic gimmicks that enliven today's weathercasts, none is more fundamental than chroma key, which integrates electronic maps, singly or in sequence, with a human interpreter. The concept is quite simple: the weathercaster stands in front of a large screen coated bright blue. A camera trained on the screen captures an image, which the chroma key unit splits into two layers: bright blue and everything else. For the blue layer the system substitutes a radar or satellite image, a computer-drawn forecast map or animation, a "title graphic" with numbers and simple symbols describing temperatures and sky conditions for next few days, or an outside shot of flowering shrubs or falling snow. Assuming none of the weathercaster's clothing is the same vivid blue, the "everything else" is the map's narrator, whom viewers perceive as standing in front of the display, as in plate 14.

Plate 14

Chroma key requires careful staging. Marks taped or painted on the floor indicate preferred positions for camera and reporter. To avoid pointing at Minnesota when describing a storm centered over Kentucky, the weathercaster cautiously eyes a monitor just out of camera range. Just above the monitor a prompter screen displays the script the weathercaster composed earlier while analyzing the latest data and selecting appropriate images. Another video prompter mounted above the lens makes it easy to remember lines and look at the camera.

Weather graphics follow an established sequence, with tomorrow's weather as its climax. According to Robert Henson, author of a comprehensive history of television weathercasting, the succession of "current weather conditions, previous highs and lows, current map, forecast map, and local forecast" was not only well entrenched by the mid-1950s but "carried such strong inherent logic that only the most daring of programmers altered it." Although recent modifications include a state or regional map that follows the national map and complementary views from GOES, NEXRAD, and fly-through animation software, the typical sequence remains a blend of chronology and cartographic scale: the coherent conflation of past-present-future with continent-region-place.

Geographer Jim Carter, who examined television weather presentations by local stations and national networks, noted the prevalence of forecast maps with three different themes: high temperatures, low temperatures, and the juxtaposition of fronts and areas of precipitation. Equally common are dynamic maps called loops, in which a rapid sequence of snapshots pauses for a moment at the most recent frame before starting over again. Typically a "satellite loop" of GOES cloud-cover images recapitulates the past 12 to 24 hours at the national or continental level, and a "radar loop" of NEXRAD images covering the last four to six hours describes developing thunderstorms or passing fronts. To make certain viewers see an important feature or relationship, the weathercaster can cycle through the sequence several times as well as pause at an especially revealing frame. Local stations make effective use of regional radar loops, while weathercasts from the major networks as well as the Weather Channel and CNN (Cable News Network) favor the national radar loop, a composite of smaller-scale images from individual radars.

According to Carter, television weather reports provide a "unique viewing environment" in which familiar base maps and predictable graphic sequences as well as a personable narrator make complex information intelligible to nontechnical viewers. Equally important is the diversity of formats, tailored to time of day as well as to market size. Morning weathercasts, for example, are shorter and less detailed than their evening counterparts. According to Tal White, the morning and noon weather anchor at WWBT-TV in Richmond, Virginia, morning viewers have a "fast-food mentality": eager to eat and run, they "listen rather than watch." By contrast, evening viewers have the patience and leisure to watch three to four minutes of weather—even more when a storm threatens or severe recent weather warrants a fuller explanation. Because morning viewers tune in at different times, White's a busy guy: "From 5:30-7 a.m., I do five full [90-second] forecasts, five 45-second quick forecasts, [and] four weather teases . . . From 7-9 a.m., I do four live Today cut-ins [and] four prerecorded Today cut-ins." Cut-ins are important to local viewers because network morning shows like NBC's Today offer little more than a short satellite loop and a fleeting glimpse of the national forecast map—less than a quarter of a weather slot in which the amiable Al Roker walks outside, looks at the sky, and chats with tourists.

Plate 15

The Weather Channel (TWC), a 24-hour cable-only service, must meet a different challenge: grabbing "grazers" who flip channels when the networks break for commercials. Because viewers are most likely to graze three minutes before and after the hour and half hour, TWC runs a six-minute, uninterrupted summary of national and regional weather in these slots. Summaries include a satellite loop, which the weathercaster can halt—often with a close-up—to point out a significant feature, as well as a national-radar loop covering the past 90 minutes. Forecast maps show weather predictions for the next three to five days, while carefully tailored graphics warn of severe weather (plate 15) and recapitulate notable storms (plate 16). Although the six-minute roundup uses a generally familiar sequence of maps, TWC revises the content as conditions change and breaks the monotony by rotating throughout the hour among three or four reporters with similar presentations but different styles. Backing up a staff of 30 weathercasters are 50 off-camera meteorologists, who analyze data and work with TWC's graphics specialists to generate a thousand unique images every day.

Plate 16

Founded in 1982 by Landmark Communications, Inc., in partnership with John Coleman, formerly the weather reporter on ABC's Good Morning America, the Weather Channel survived a shaky start-up by offering itself free to major cable systems for the first few years: a gamble needed to build a subscriber base attractive to advertisers, who now account for 40 percent of its revenue. Available in 97 percent of all cable TV households, TWC attracts a wide range of advertisers, including Michelin Tires, sponsor of the Michelin Travel Watch. According to company officials, the Weather Channel pulls in three distinct groups of viewers: weather enthusiasts, who watch regularly; people planning trips, work, or recreation; and severe weather viewers, who tune in when storms threaten or strike.

Jim Carter, who figures the ads consume a fifth of TWC's total airtime, noted the added variety of "international weather" segments focused on Europe as well as frequent local weather reports, which account for another fifth of total programming. Little longer than a minute, these local segments appear several times an hour, usually in a predictable "local on the eights" time slot. To compile customized packages for local cable markets, a Weather Channel computer canvasses current National Weather Service forecasts for more than 800 local zones, extracts information for each cable system's central city and its neighbors, and transmits the modules by satellite to the entire country. Location codes allow the local cable system's computer to grab the right module for automatic insertion in the prescribed time slot. In addition to a concise text forecast of the area's weather for the next 36 hours, an extended three-day outlook, and a few statistical and astronomical facts, each report includes a regional forecast map and a radar loop recapitulating precipitation throughout the area over the last 90 minutes. Accompanied by background music or an automatic voice-over, these customized local updates are a key attraction for viewers, who stay tuned an average of nearly 14 minutes. Cable viewers can catch weather maps on two other 24-hour information channels: Cable News Network (CNN), founded by Ted Turner in 1980, two years before TWC, and MSNBC, a joint venture of Microsoft Corporation and NBC that went on the air and on the Internet in July 1996.

Cable channels and local stations depend heavily on specialized commercial intermediaries for sophisticated graphics software as well as for NEXRAD and satellite images. A consequence of the increased complexity of computer technology, these intermediaries are comparatively recent. In the 1950s and 1960s, TV weather reporters relied on government facsimile charts in working up locally tailored forecasts and crude supporting graphics. The picture changed in the early 1980s, when digital electronics created a demand for animated forecast maps and other engaging weather graphics. Intermediaries acquired an even larger role when the Reagan administration pushed privatization and free-market competition. A single tax-supported agency still gathers raw data and makes predictions, but today's National Weather Service lacks the budget if not the inclination to satisfy viewers' demand for vivid and lively weather maps. Free enterprise kicks in as various private firms disseminate the information in raw and enhanced form to television, newspapers, and thousands of other consumers with specialized needs. Among the few exceptions are dial-up telephone forecasts and NOAA Weather Radio, which broadcasts local and regional summaries over seven high-band FM channels accessible only to receivers with a "weather band."

The lords of meteorological dissemination are the four "private-sector data providers" with wires running out the back of every NEXRAD receiver. If the NEXRAD Information Dissemination Service (NIDS) has a charter member, it's Unisys, the contractor that developed NEXRAD's Principal User Processor (PUP) and now delivers radar imagery to the government's principal users: the National Weather Service, the Federal Aviation Administration, and the military. In July 1990, nearly three years after the government advertised for pre-proposals, the Department of Commerce selected three additional NIDS contractors to service other government agencies as well as numerous external users: Alden Electronics Incorporated, Kavouras Incorporated, and WSI Corporation (Weather Services International). An annual access fee of approximately $1,350 per NEXRAD site entitles NIDS data providers to sell radar data to whoever wants to buy it, including firms like Accu Weather and EarthWatch Communications, which market enhanced weather graphics to much the same clientele. Restrained by federal antitrust regulations and NOAA guidelines, Adam Smith's invisible hand thus promotes wide dissemination at an arguably fair price.

Specialized vendors offer broadcasters and cable networks a full range of studio-ready weather products, including GOES imagery from NOAA's National Environmental Satellite, Data, and Information Service (NESDIS), which remains in government hands. (In the early 1980s, the government tried to sell the nation's weather satellites but backed off when commercial weather services and other users complained loudly.) Key attractions include integrated data and graphics software, which allow the operational convenience of one-stop shopping and guaranteed compatibility as well as a broad range of products and services readily tailored to clients' needs. Experience in developing weather software and computer workstations for television stations and various government and commercial customers accounts for the eagerness of Alden, Kavouras, and WSI to become NIDS providers: they knew their market and its potential for profit. In mid-1997, for instance, WSI's Web site boasted of a healthy list of media clients, including the Weather Channel, CNN, NBC's Today Show, and over 325 U.S. television stations.

Attesting to television's addiction to trendy technology, most of these customers eagerly bought into three-dimensional animation in the mid-1990s, when WSI and its competitors took advantage of faster, less costly computers to implement earlier advances in digital cartography. Now widely affordable in large and medium-size markets, 3-D weather systems enliven radar and cloud-cover presentations with spectacular fly-throughs and dynamic simulations of falling rain. Although the rapid succession of images can be overwhelming, animated graphics provide a vivid integration of continental weather systems with local landmarks. With 3-D—actually 4-D, if we recognize time as a dimension—a weather reporter can move seamlessly from a bird's-eye view of cyclonic circulation to a cross-sectional below-the-clouds description of an approaching cold front (plate 17). The result is radical if not revolutionary, with abstract map symbols attaining a "virtual reality" of sorts as the barbed blue line tracking a cold front's forward progress triggers thunderstorms, which highway symbols and labels then relate to recent reports of flooding and traffic accidents. This flair for engaging the local audience is 3-D's strongest selling point.

Plate 17

Hyperreality took the Syracuse weathercasting market by storm in October 1995, when two of the three stations affiliated with a major network went 3-D. My graduate cartography seminar had just visited WIXT-TV, where head meteorologist Dave Eichorn (plate 17) enthusiastically described his new system, recently purchased from WSI for around $200,000 by a supportive management, which further affirmed its commitment to weather reporting by hiring a third meteorologist. With "3-D SkyTrack," WIXT's weather team could design its own fly-throughs and make changes minutes before going on the air—a distinct advantage over its rival's 3-D simulations, composed hours earlier by a weather graphics service in California.

A tape I made a few weeks later captured Eichorn's appreciation of coherence: an otherwise baffling SkyTrack segment added a different but welcome perspective when preceded by the more familiar national satellite and regional radar loops. Having presented an overview of recent weather, Eichorn explained his forecast by taking viewers beneath the clouds and moving westward through rain and snow falling over Lake Ontario, southern Canada, and Michigan. "It's kind of nice," he observed, "to get into a storm in 3-D like that, and take it apart, and see what's coming our way." Yes, indeed—as long as simpler 2-D sequences provide an introduction.

Like many television stations, WIXT supplements its NEXRAD loop with a lightning-stroke display. Blips on this dynamic map record the locations of short but powerful cloud-to-ground discharges, often but not necessarily associated with thunderstorms. As a complement to radar and satellite imagery, lightning-detection systems help the Forest Service and electric utilities monitor the likely locations of forest fires and power outages. WIXT's display taps into the high-resolution lightning detection network developed by the Niagara Mohawk Power Corporation to help dispatch repair crews as well as redesign transmission and distribution lines. Receiving directional data from 20 lightning detectors covering 24,000 square miles, the system correlates tripped circuit breakers with lightning discharges and triangulates damage locations. Because electrical storms threaten many other industries, including air-traffic management, casualty insurance, and construction, television broadcasters comprise a small part of the market for lightning data offered by Accu Weather, Alden, and similar weather services.

There's little limit to the weather images and programming a television station can buy. As I write, Syracuse's NBC affiliate rents production facilities to the area's Fox Network station, which broadcasts a local news program at 10 P.M., an hour before the three major-network stations go head-to-head with half-hour late-evening newscasts. The Fox outlet provides its own news anchors, while the electronic host supplies video footage, production staff, and some abbreviated, almost-as-nice weather graphics as well as its well-known weather personality. There's another option, though: small stations can avoid the expense of generating their own weather reports by buying a complete localized weather segment from the National Weather Network, whose five weathercasters serve over 50 small stations throughout the country from a satellite uplink in Jackson, Mississippi. Broadcasters willing to pay extra for a special feed can even warn their viewers of approaching tornadoes or severe thunderstorms.

Among our four local stations, I most enjoy watching weather on WIXT. In addition to impressive graphics and generally accurate forecasts, Eichorn and his colleagues faithfully identify sources of uncertainty—valuable information for folks like me, who believe that understanding weather is the best alternative to controlling it. Complementing this naturally triggered discourse on jet streams, blocking highs, and lake-effect snow is WIXT's "Weather School," which challenges and entertains viewers with jargon-free multiple-choice questions about atmospheric physics, synoptic meteorology, and weather maps. Most appealing, though, is Eichorn's ill-disguised lust for tumultuous days with "lots of lines on the weather map."

 

---

Mark Monmonier
Air Apparent: How Meteorologists Learned to Map, Predict, and Dramatize Weather
©1999, 340 pages, 22 color plates, 4 halftones, 89 line drawings
Cloth $27.50 ISBN: 0-226-53422-7
Paper $17.00 ISBN: 0-226-53423-5

For information on purchasing the book—from bookstores or here online—please go to the webpage for Air Apparent.

---

See also:

• Monmonier's list of ten risky places, a web feature for Cartographies of Danger: Mapping Hazards in America
• "Eleven Ways You Are Being Watched", a web feature for Spying with Maps: Surveillance Technologies and the Future of Privacy
• Spotting Bushmanders, a web feature for Bushmanders and Bullwinkles: How Politicians Manipulate Electronic Maps and Census Data to Win Elections
• All our books by Mark Monmonier
• Our catalog of Cartography titles
titles
• Other excerpts and online essays from University of Chicago Press titles