EACH morning, millions of us turn on our radios and televisions for the weather forecast. Do the cloudy skies mean rain? Will the early sunshine last? Will rising temperatures bring a thaw to melt snow and ice? Once we hear the forecast, we decide what clothes to wear and whether to carry an umbrella or not.
From time to time, though, weather forecasts are conspicuously off the mark. Yes, though the accuracy of forecasts has improved dramatically in recent years, predicting the weather is a fascinating mixture of art and science that is far from foolproof. What is involved in predicting the weather, and how reliable are weather forecasts? In answer, let us first explore how weather forecasting developed.
Measuring the Weather
In Ancient times weather forecasting was primarily based upon observations made with the naked eye. Today meteorologists have an array of sophisticated tools at their disposal, the most basic of which measure air pressure, temperature, humidity, and wind.
In 1643, Italian physicist Evangelista Torricelli invented the barometer—a simple device that measures air pressure. It was soon noted that air pressure rises and falls as the weather changes, a drop in pressure often signaling a storm. The hygrometer, which measures atmospheric humidity, was developed in 1664. And in 1714, German physicist Daniel Fahrenheit developed the mercury thermometer. Now the temperature could be accurately measured.
About 1765, French scientist Antoine-Laurent Lavoisier proposed that daily measurements of air pressure, moisture content, and wind speed and direction be made. “With all this information,” he declared, “it is almost always possible to predict the weather for one or two days ahead with reasonable accuracy.” Unfortunately, doing so proved to be anything but simple.
Tracking the Weather
In 1854 a French warship and 38 merchant vessels sank in a fierce storm off the Crimean port of Balaklava. The French authorities asked Urbain-Jean-Joseph Leverrier, director of the Paris Observatory, to investigate. By checking meteorologic records, he discovered that the storm had formed two days before the disaster and had swept across Europe from the northwest to the southeast. Had a system of tracking the movements of storms been in place, the vessels could have been given advance warning. A national storm-warning service was thus established in France. Modern meteorology had been born.
Needed, though, was a quick way for scientists to receive weather data from other locations. And Samuel Morse’s recently invented electric telegraph was just the means to do so. This made it possible for the Paris Observatory to begin publishing the first weather maps in modern format in 1863. By 1872, Britain’s Meteorological Office was doing the same.
The more that meteorologists acquired data, the more they became aware of the enormous complexity of the weather. New graphic devices were thus developed so that weather maps could convey additional information. Isobars, for example, are lines drawn to link points that have the same barometric pressure. Isotherms connect locations that have the same temperature. Weather maps also use symbols that show wind direction and force, along with lines that depict the meeting of warm and cold air masses.
Sophisticated equipment has also been developed. Nowadays hundreds of weather stations around the world release balloons carrying radiosondes—instruments that measure atmospheric conditions and then radio the information back. Radar is also used. By bouncing radio waves off raindrops and ice particles in clouds, meteorologists can track the movement of storms.
A leap forward in accurate weather observation came in 1960 when TIROS I, the world’s first weather satellite, rocketed heavenward equipped with a TV camera. Now weather satellites orbit the earth from pole to pole, whereas geostationary satellites maintain a fixed position above the earth’s surface and continuously monitor the part of the globe in their field of view. Both types beam down pictures of the weather, which they view from above.
Forecasting the Weather
While it is one thing to know exactly what the weather is right now, it is quite a different matter to predict what it will be in an hour, a day, or a week. Shortly after World War I, British meteorologist Lewis Richardson reckoned that since the atmosphere follows the laws of physics, he could use mathematics to predict the weather. But the formulas were so complicated and the number-crunching process so time-consuming that weather fronts were gone before forecasters could complete their calculations. Besides, Richardson used weather readings taken at six-hour intervals. “An only marginally successful forecast requires that measurements be taken at intervals of thirty minutes at the most,” observes French meteorologist René Chaboud.
With the advent of computers, however, it became possible to perform the lengthy calculations speedily. Meteorologists used Richardson’s calculations to develop a complex numerical model—a series of mathematical equations that encompass all the known physical laws governing the weather.
To employ these equations, meteorologists divide the earth’s surface into a grid. Currently, the global model used by Britain’s Meteorological Office has grid points spaced about 50 miles [80 km] apart. The atmosphere above each square is called a box, and observations of atmospheric wind, air pressure, temperature, and humidity are recorded at 20 different levels of altitude. The computer analyzes the data received from the observation stations throughout the world—more than 3,500 of them—and then produces a forecast of what the world’s weather will be for the next 15 minutes. Once this has been done, a forecast of the following 15 minutes is rapidly produced. Repeating this process many times over, a computer can make a six-day global forecast in a mere 15 minutes.
For greater detail and accuracy in local forecasting, the British Meteorological Office employs the Limited Area Model, covering the North Atlantic and European sectors. It uses grid points spaced at intervals of about 30 miles [50 km]. There is also a model that covers only the British Isles and surrounding seas. It has 262,384 grid points ten miles [15 km] apart and 31 vertical levels!
The Forecaster’s Role
Predicting the weather, however, is not all hard science. As The World Book Encyclopedia puts it, “the formulas used by the computers are only approximate descriptions of the behavior of the atmosphere.” Furthermore, even an accurate forecast for a large area may not take into account the effect of local terrain on the weather. So some degree of art is also necessary. Here is where a weather forecaster comes in. He uses his experience and judgment to determine what value to place on the data he receives. This allows him to make a more accurate forecast.
For example, when air cooled by the North Sea moves over the European landmass, a thin cloud layer often forms. Whether this cloud layer heralds rain in continental Europe the following day or simply evaporates in the sun’s heat depends on a temperature difference of only a few tenths of a degree. The forecaster’s data, along with his knowledge of previous similar situations, enables him to offer good advice. This mixture of art and science is critical to producing accurate forecasts.
Presently Britain’s Meteorological Office claims 86-percent accuracy for its 24-hour forecasts. Five-day estimates from the European Centre for Medium-Range Weather Forecasts achieve an accuracy of 80 percent—better than the reliability of two-day forecasts in the early 1970’s. Impressive but far from perfect. Why are forecasts not more reliable?
For the simple reason that weather systems are enormously complicated. And it is not possible to take all the measurements needed to make foolproof predictions. Vast areas of the ocean have no weather buoys to beam data via satellite to ground stations. Rarely do weather-model grid points correspond exactly to the location of weather observatories. Besides, scientists still do not understand all the forces of nature that shape our weather.
But improvements are constantly being made in weather forecasting. For example, until recently, forecasting the weather depended mainly on observation of the atmosphere. But with 71 percent of the globe’s surface covered by ocean, researchers are now focusing attention on the way energy is stored and transferred from the ocean to the air. Through a system of buoys, the Global Ocean Observing System provides information about slight rises in water temperature in one region that can have dramatic consequences on the weather far away.