WEATHER SATELLITES: A 20-YEAR SUCCESS STORY
Washington — When dust from Mt. St. Helens boiled into the stratosphere May 18, an ever-watchful weather satellite quickly spotted it. Meterologists could see the winds grab that dust and spread it eastward hour by hour in a series of pictures that helped them forecast fallout and assess possible climatic consequences. (They turned out to be minor.)
We take this all-seeing perspective for granted these days. No hurricane bears down upon as undetected, no tornado-spawning front or thunderstorm phalanx eludes surveillance. On millions of television screens daily, the mystical symbols of the weather map are transformed into "living" cloudscapes.
We forget that it was only 20 years ago that the first experimental weather satellite -- TIROS, meaning Television and Infrared Observation Satellite -- was launched, to a chorus of skepticism from "practical" forcasters.
Would these expensive gadgets ever produce anything but pretty cloud pictures? And what would you do with them anyway when they arrived by mail several days late? Weather people need timely, reasonably accurate measurements of pressure, temperature, humidity, and winds at several levels throughout the atmosphere -- not just "scenery." Against this need, all that the Weather Bureau could promise was a description of the cloud pictures in laconic teletype code, which would still be many hours late.
Weather satellite prophets were undaunted. The included such farsighted scientists as the late Harry WExler, Weather Bureau research director; Morris Tepper (now retired), who would shepherd the development of much weather satellite technology at the National Aeronautics and Space Administration (NASA); and David S. Johnson, who now heads the National Environmental Satellite Services (NESS). They knew that, once meteorologists had the Olympian perspective of seeing weather in large scale from above, they would never be satisfied with the worm's-eye view of looking up t it station by station from below.
Wexler drew a picture of what a satellite might see -- a tantalizing scene of North America with those now-familiar cloud patterns sweeping over it -- and remarked:
"There are many things that meteorologists do not know about the atmosphere; but of this they are sure: that the atmosphere is indivisible, and that meteorological events occurring far away will ultimately affect local weather."
By filling in the blank oceanic areas on weather maps and covering the "data poor" Southern Hemisphere, satellites, he believed, would at last enable meteorologists to deal with weather in this comprehensive perspective. This remains the justification and fruitage of what has become a global weather satellite network in which most nations participate. Europe, Japan, the Soviet Union, and the United States maintain satellite systems, primarily for their own purposes but providing benefits that are widely shared.
Yet even the prophets had twinges of doubt when NASA launched that first TIROS experimental satellite April 1, 1960. Although the "superb" pictures it sent back seem fuzzy and inadequate today, Harry Wexler and his colleagues felt they confirmed their faith in the weather satetllite's potential. They also knew they now had to come through on their promises, and the technology for doing it was not all in hand by any means nor even particularly well defined.
Besides a comprehensive view of cloud patterns and early warning of ocean storms, they were promising timely, reasonably accurate data throughout the atmosphere, which forecasters demanded -- and on a global basis. They were dreaming of imaging things the eye cannot see, such as large-scale movements of invisible water vapor and of measuring the upward flow of heat radiation, which is a major element in the income-outgo (sunshine-heat radiation) energy budget of the atmosphere. They spoke of someday "stripping away" the clouds to survey the surface everywhere, and of monitoring snow and ice cover, ocean waves and currents, and sea surface temperatures, which are an important determinant of large-scale weather patterns. In a burst of enthusiasm, Weather Bureau Director Francis W. Reichelderfer had forecast savings of "many lives and many dollars" a year from improved forecasts due to data from satellites -- $4 billion a year to be specific, and that in 1958 dollars.
Yet in 1960, engineers were still discovering what it meant to build space hardware that would work reliably day after day for several years, let alone inventing new sensors. And rocket launches were still touch and go. It's little wonder that the "practical" forecasters didn't take the space talk too seriously. To their credit, the prophets never did suggest they would do it all right away. Morris Tepper repeatedly and patiently explained to overeager reporters the difference between long-range hopes and the nitty-gritty of turning the TIROS series of experimental satellites into a workaday weather observing system. Such caveats were often lost in the media hype of the day.
Well, after 20 years, were the skeptics justified or have promises been kept? "We did all we said we'd do and then some," David Johnson says proudly.
It's easy to believe him. On my dsk are pictures of "invisible" water vapor flowing across the North and South Atlantic, moving from tropical regions to higher latitudes. They were made by an infrared (heat) radiation sensor on Meteorsat, a geostationary satellite launched by NASA in September 1977 for the multinational European Space Agency. Geostationary means the satellite, in an equatorial orbit 35,800 kilometers high, moves at the same rate at which Earth turns. Thus the satellite appears to hover over a given location. Its infrared (IR) sensor monitors frequencies at which the water vapor absorbs IR radiation in the middle levels of the troposphere, the part of the atmosphere below the stratosphere. It images represent humidity in the middle of this important part of the atmosphere were most weather phenomena occur.
There's also a technical paper comparing "soundings" (temperature vs. height measurements) made by the latest TIROS-class satellites with those of conventional radiosondes. A radiosonde is a balloon-borne instrument pack the measures temperature, humidity, and pressure. Tracking of the balloons also indicates winds at different heights. On the satellite, sensors monitor a range of frequencies at which water, carbon dioxide, or other atmospheric constituents absorb or transmit IR radiation. They provide data from which computers construct temperature-height profiles.
The fit between the satellite product and conventional readings looks pretty good, although the radiosonde plots are more accurate and show more detail. Forecasters won't abandon the radiosondes yet. However, the technical paper concludes, the satellite soundings now "provide useful and valuable meteorological information, especially in data-sparse regions. . . ." Temperature measurements that skeptics once thought could be made usefully only by instruments hanging from balloons are now made routinely from orbit.
These satellites can also measure the total water vapor content in the column of atmosphere beneath them, although they cannot give specific humidity readings at specific heights. However, engineers know how to do this by measuring heat radiation at IR or radio frequencies. The degree of absorption at specific frequencies represents humidity at different altitudes. Future US weather satellites will likely carry sensors to to do this.
No weather satellite can yet measure winds directly. But by following clould movements, computer analysis can estimate winds at cloud height. AT NESS, this means making routine wind measurements by tracking cirrus clouds whose heights can be estimated from the cloudtop temperatures measured by the satellite, and following low-level clouds assumed to be at about 1,000 meters. A film loop of a sequence of pictures made by a geostationary satellite over a given time interval is projected onto a plotting table. An analyst picks a distinctive cloud pattern and follows its movement, taking care not to confuse growth or dissipation of the clouds with true motion. The measurement is automatically transmitted to a computer, which computes wind speed and velocity. It's not as good as baloon tracking by eye or radar, and it doesn't indicate wind at intermediate levels. But the data are useable and where there are no other data , over oceans in particular, they can be vital.
Dr. Johnson has this kind of general achievement in mind when he says promises have been kept. He is also thinking of the initial set of formal goals that were set in 1961 when Congress authorized the Department of Commerce to establish and run a global environmental satellite system. Working through what now is the National Oceanic and atmospheric Administration (NOAA) and its predecessor agencies (the Weather Bureau, which later became part of the Environmental Science Services Administration), the Commerce Department formed a continuing partnership with NASA in which the space agency has done, and still does, much of the experimental work. There was a series of simply stated, but challenging, goals:
First, "establish an operational system for viewing the atmosphere regularly and reliably on a global basis, both day and night, with direct readout to local ground stations within radio range of the satellite."
This goal began to be realized in 1966 with the launch of the first two operational weather satellites, ESSA 1 and ESSA 2, whose designs were proved in a series of 10 TIROS experiments. Nighttime coverage was not yet possible. But one of the satellites carried an automatic picture transmission system that broadcast satellite pictures to anyone with the relatively inexpensive equipment needed to receive them. Today some 120 nations take advantage of these transmissions from NOAA satellites -- to say nothing or private groups, schools, and even Scout troops, some of which hve built their own receivers. By 1970, with the launch of NOAA 1, an improved TIROS system, infrared scanners that form picturelike images inaugurated nighttime coverage on a global basis. Placed in a nearly polar orbit, this satellite could cover virtually all the Earth.
A second formal goal was to "establish an operational system for sounding the atmosphere regularly and reliably on a global basis and for providing quantitative inputs to numerical weather predictions." That meant providing data that working weather people would feel were good enough to feed into their forecasting computers. NOAA 2, launched in 1972, began to satisfy that objective by taking temperature soundings on nearly a global basis.
A third goal called for "an operational system for continuous viewing of weather features [rather than just a few times a day] and for collecting and relaying environmental data from remote platforms such as buoys, ships, automatic stations, aircraft, and balloons." Continuous viewing meant geostationary satellites. NASA began developing them in 1966. But NOAA's system didn't become fully operational until GOES 1 (Geostationary Operational Environmental Satellite) was launched in 1975.
Today NOAA's National Environmental Satellite Service has three GOES spacecraft in orbit, plus one of the NASA prototypes -- Synchronous Meteorological Satellite (SMS 2) -- in service. These complement the polar-orbiting NOAA 6 and TIROS N satellites, which represent the third generation of the TIROS system. Both GOES and the polar orbiters can read out data from remote platforms and relay it. The level of a mountain stream or air temperature in the remote Arctic now can be read as easily as though the instruments were downtown.
The equipment is reliable; launches of replacement satellites are routine, if not error-free. Booster trouble May 29, 1980, put what would have been the NOAA 7 polar orbiter into the wrong orbit and NESS will have to try again in four or five months. Nevertheless, Dr. Johnson points out, the present system has amply fulfilled those 1961 goals.
What of those parts of the dream that have not been realized? Winds are still not well measured from orbit. However, Dr. Johnson notes that some experts are beginning to see ways to track wind flow even where there are no clouds. T. Rhidian Lawrence at NOAA'S Boulder, Colo., laboratories, for example , thinks that lidar (laser radar) on a polar orbiting satellite could measure winds to within one meter per second in speed and 10 degrees in azimuth. Dust particles, water drops, and other light scatterers in the air would give the lidar beam trackable targets. NOAA should have a prototype ready for testing by 1984.
Satellite scientists have yet to "strip away" the clouds, but they now know how to do it. Microwave radiation easily penetrates clouds. By sensing heat radiation at microwave frequencies and using radar, many important surface features could be constantly monitored. Experimental meausurements by the Seasat satellite launched two years ago show what can be done. For 90 days, before Seasat failed, radar measured surface winds over the ocean, wave heights, and the varying slope of the sea which is related to currents. Radio sensors measured surface temperatures and ice and snow coverage, and even detected differences between new ice and ice more than a year old. It even seems possible to determine the characteristics of snow below the surface and of underlying ground. Ice and snow monitoring is important for shipping and for managing water supplies. NOAA hopes to follow up the Seasat work later in this decade with a satellite to be launched by the Space Shuttle.
Last year, W. John Hussey and E. Larry Heacock of NESS came up with an estimate of how much the weather satellite system saves US industries and government. It came to around $172 million a year just for jobs such as crop protection, more efficient ship management, or location of promising fisheries, items that could be reasonably pinned down. The scientists added that many more millions are probably saved by storm warning and better weather forecasting. While it's a respectable figure, $172 million is far short of Reichelderfer's $4 billion ($10 billion to $11 billion in 1979 dollars.)
Reichelderfer was thinking in terms of dramatic improvement in weather forecasting. White satellite data have brought substantial forecasting gains, especially in terms of early warnings, neither satellites nor the computerized forecasting that their data support has revolutionized weather prediction. That's especially true of the local forecasts that affect people directly. However, Vincent Oliver at NESS sees a true breakthrough here -- a breakthrough that depends on making everybody his or her own forecaster.
In the early 1960s, he was urging professional meteorologists to wake up to what the satellite view could do for them. Now, he says, it's time to make that view directly available to the public. He explains that NESS has a wealth of up-to-the-minute satellite pictures. These could be disseminated by communications satellite to local television outlets (most probably cable-TV outlets). Fed into home television sets, these pictures could give a view of general weather that would be continuously available and up to date. Then, if thunderstorms threatened, viewers could see whether places they planned to visit were in their path. Local clear areas could also be spotted. And so forth.
Just as soon as commercial television is ready to try this, Dr. Oliver says, NESS could make the pictures available. He thinks it would have a bigger impact than all of the fancy new satellite sensors. "Most meteorologists don't believe the public can understand whether maps; and they are hard to understand," he explains. "But they can understand clouds. Let's give them what they can understand. I think the biggest improvement in weather forecasting will be in this one - to two-hour, short-term area, and it will be the public that will do it."