How Meteorology Works

Pattern recognition is one of the key properties of intelligence. Your dog knows she'll get a treat if she does a trick because you always feed her after she rolls over. You know your aunt will probably give you pajamas for Christmas because your closet's stuffed with hideous, plaid pajamas. Our minds learn and act based on the patterns we perceive all around us. By knowing what came before, we know what the future will hold.

­Early humans observed patterns in the weather and learned to anticipate changes that affected their food supply and well-being. They created names for the seasons and even forms of calendars to guide their wanderings and, later on, the growing cycles of their crops. People knew what the weather would be based on their own cumulative experiences, as well as those passed onto them by their ancestors. For instance, certain Australian Aboriginal tribes carry with them 18,000 generations worth of local weather observations [source: BBC]. Outside of their territory, their knowledge gradually fizzles out, but their understanding of weather incorporates a great deal of local detail. A tribe may recognize as few as two or as many as six seasons, depending on local precipitation and temperature factors.

Early humans knew that cooling temperatures meant the coming of winter. They knew the sights and smells that preceded a rainstorm. And where their own senses failed them, they turned to those in nature: the life cycles of various vegetation and the migration of other animals. Plus, many animal species are far more attuned to changes in air and water pressure that often signal storms and other atmospheric changes.

Long before humans invented devices to measure these conditions, they simply looked to the skies and fields. Animals recognized subtle patterns in the atmosphere, and we recognized patterns in their responsive behavior. These traditions continue to this day in the pages of The Old Farmer's Almanac and with practices such as Groundhog Day. To learn more about animals and weather, read Can animals predict the weather?

Following the weather in a region or tribal territory is one thing, but charting atmospheric patterns on a global scale is a different undertaking altogether. Over the last few centuries, synoptic meteorology, or the idea of comprehensively charting the weather over a large area, has emerged. By comparing simultaneous weather conditions in adjacent areas, scientists were able to better understand prevailing conditions over a larger area and to provide the kinds of weather maps we view on TV and the Internet every day.

­How do meteorologists record current weather conditions? Read the next page to find out how we go about determining what our weather is doing right now.

You've probably heard the saying "too many cooks spoil the broth," most likely in reference to an album, business venture or even a sports team. It means that when too many people work on a project, the results are often confusing, lackluster or just plain awful. In a way, the atmosphere is a broth stirred by way too many cooks: gravity, sunlight, rotation, conflicting pressure zones, cool oceans, hot deserts, mountain ranges and colossal air currents, just to name a few. These forces constantly stir the atmosphere to movement, and just figuring out what it's doing at any given second requires a great deal of study and observation.

Three of the key atmospheric properties are air pressure, air temperature and humidity. To really understand what's going on, you have to be able to measure these conditions. For this reason, meteorology didn't really emerge as a science until the 17th-century inventions of the barometer, which measures air pressure, and a reliable thermometer, which gauges temperature. Before the 1600s ended, scientists also developed reliable hygrometers to measure humidity. These instruments, along with rain gauges, allowed for improved agricultural planning and sea travel.

But to gain a truly synoptic view of current weather conditions, you need a way to communicate with other observers in other regions. The 1837 invention of the telegraph made this possible. By the mid-19th century, meteorologists at various weather stations were able to communicate quickly with each other and assemble the big picture.

Toward the end of the 19th century, meteorologists used weather balloons to study the upper atmosphere. In doing so, they made crucial discoveries about upper-air pressure and wind patterns. Through this, meteorologists were able to discover the role low-pressure centers play in the weather. You've probably seen these centers pointed out on a map in weather forecasts. Cooler, denser air spirals into warmer, low-pressure areas from the surrounding regions. This, in turn, pushes the warm air up into the upper atmosphere, where the air spirals out in all directions. These formations are called cyclones (not to be confused with hurricanes and typhoons, which are called cyclones in some regions).

But this elevation of air doesn't just happen at low-pressure centers. It also happens when two air masses meet each other in a front. In either case, elevated air often forms clouds and storm sy­stems. With these discoveries, meteorologists were better prepared to forecast weather. They weren't just making educated guesses based simply on pattern recognition; they were understanding how the atmosphere works.

In the 20th century, advancements in aviation made it possible for us to better study the upper atmosphere, and new radio technology allowed meteorologists to send sensitive equipment up with balloons to high altitudes -- a practice that continues today. Similarly, radio weather buoys communicated back conditions at sea, including water temperature, wind speed and wave height. Following World War II, scientists began using radar to study the weather since the technology made it possible to detect rainfall in addition to aircraft.

In 1960, another advancement added to our ability to observe and measure Earth's atmosphere: the weather satellite. By placing these automated observatories in north-to-south polar orbits and east-to-west geostationary orbits, humans were able to view their atmosphere from the outside and observe storm systems from a truly synoptic vantage point. Weather satellites provide more than just an extraterrestrial view of weather; they also carry sensors to measure temperature, moisture and solar radiation.

­It's one thing to know what's happening now, but how do meteorologists transform all this data into an idea of what tomorrow's weather will bring? Read the next page to find out.

Modern technology allows meteorologists an unprecedented understanding of Earth's atmosphere and an excellent vantage point from which to observe its weather. But how do meteorologists translate this into a reasonable prediction of what the weather will continue to do?

­Instead of simply looking at current conditions and estimating based on past observances, meteorologists create numerical weather prediction (NWP) models. These models are objective, physics-based calculations which, when processed through a computer, predict what current weather will look like in the future. The equations involved in these models are complex and involve multiple atmospheric variables. These variables leave room for error, so the further out meteorologists attempt to forecast, the greater the potential for error becomes.

Look at any hour-by-hour weather forecast: Each hour's forecast is a step into the possible future. An initial forecast (such as what the weather will be an hour from now) is the result of running a computer model based on what the weather is doing now. Then, to run a model of what the weather will be doing in two hours, you apply the various equations to the model that came before it. So while the first forecast was based on actual conditions, the second was based on forecasted conditions that might have been less than accurate. Each subsequent forecast compounds the possibility for error. For this reason, NWP models become increasingly flawed the further ahead you try to look.

Meteorologists have steadily improved NWP models since the 1980s. By constantly tweaking them, they've created more accurate equations with fewer errors. Another technique, called Model Output Statistics, improves weather forecasts by taking the NWP model, which is based on current conditions, and extrapolating it by comparing it to past surface conditions for a particular region. This method essentially uses past weather readings to balance some of the errors inherent in an NWP model.

­Despite continued advancements in meteorology, don't expect a flawless weather forecast anytime soon. When considering the numerous variables in an NWP model, it's important to realize just how much difference even a small discrepancy can make. In 1961, meteorologist and chaos theory founder Edward N. Lorenz took a close look at the drastically different models a difference of a single decimal point could make. Based on these findings, Lorenz coined the term butterfly effect, in which the question is asked, "Does the flap of a butterfly's wings in Brazil set off a tornado in Texas?"

But while forecasting the weather is far from flawless, meteorology has saved countless lives by allowing scientists to predict where destructive weather will strike and warning people ahead of time. Your five-day forecast may not be perfect, but neither is our understanding of the complex array of movements that fill the atmosphere we live in.

Explore the links on the next page to learn more about the weather.