Weather refers to various phenomena that occur in the atmosphere of a planet. On Earth the regular events include wind, storms, rain and snow, which occur particularly in the troposphere, the lower part of the atmosphere. Weather is driven by energy from the sun, with key factors such as temperature, humidity, atmospheric pressure, cloud cover, wind speed, and elevation.
Weather stations all around the world measure these conditions on a continuous basis. The weather system is an inherently chaotic system; practically this means that meterologists can only predict the weather a few days in the future. In fact it was through studying weather systems that chaos theory first emerged, pioneered with the aid of early computers by Edward Lorenz. Recently, though, it has been shown that the theoretical time ahead that it is possible to forecast is not as short as previously thought. Research suggests that problems with the formulas used in current models remains a limiting factor in the accuracy of predictions, rather than the inheritantly chaotic nature of the weather. However this still only extends the forecast time by at most a week.
Weather prediction is an area of much active research, due to the advantages that an accurate forecast of the weather would bring. Chaos theory says that small changes in the initial conditions increase [exponentially?] over time, until the weather is very different from what it would have been under slightly different circumstances. Forecasters are able to increase the quality of their predictions by calculating a number of different scenarios making small changes to each's initial conditions. These end results are then compared, and the most common scenario is chosen as the forecast. This also has the advantage that forecasters have an estimate of the reliablity of their forecast - if different models produce wildly incompatible results the forecast is known to be unreliable.
See also: climate.