Weather Forecasting

The complexity of atmosphere conditions is reflected in the fact that none of the forecasting methods outlined above is dependable for more than a few days at best. This reality does not prevent meteorologists from attempting to make long-term forecasts. These forecasts might predict the weather a few weeks, a few months, or even a year in advance. One of the best known (although not necessarily the most accurate) of long-term forecasts is found in the annual edition of the Farmer's Almanac.

The basis for long-range forecasting is a statistical analysis of weather conditions over an area in the past. For example a forecaster might determine that the average snow fall in December in Grand Rapids, Michigan, over the past 30 years had been 15.8 in (40.1 cm). A reasonable way to try estimating next year's snowfall in Grand Rapids would be to assume that it might be close to 15.8 inches (40.1 cm).

Today this kind of statistical data is augmented by studies of global conditions such as winds in the upper atmosphere and ocean temperatures. If a forecaster knows that the jet stream over Canada has been diverted southward from its normal flow for a period of months, that change might alter precipitation patterns over Grand Rapids over the next few months.