|MadSci Network: Earth Science|
A good question Becky. Forecasting what the weather will be like in a few
hours, or the next day, or the next week is a challenging task...but one
that meteorologists tackle every day. There are many different approaches
to predicting the weather.
One is called persistence forecasting. This means using the probability that what happened yesterday is likely to happen again today. This method works fairly well in periods of slow changing weather patterns...like during the summer months, or in the tropics.
Another method often used in short-term forecasting, for the next 6-12 hours, involves looking "upstream" at what is happening and moving that weather into the area of interest. An example of this might be tracking a line of thunderstorms moving across an area and timing its arrival in your city, or estimating the time a layer of clouds will arrive and turn sunny skies cloudy. Satellite and radar images are used routinely for this kind of forecasting.
For longer-term forecasts, meteorologists around the globe rely on numerical models of the atmosphere. The atmosphere moves and changes in accordance to known laws of physics. Numerical models have been developed that can predict the future state of the atmosphere based on these laws. Observations at thousands of places on the globe are taken simultaneously. Observations above the surface are taken with weather balloons at the same time. This data is fed into supercomputers that then run one of these atmospheric models. The forecaster makes a prediction of what conditions at a particular location will be like based on the model's output.
Many different models of the atmosphere have been developed for forecasting...and many more are used in research. Some are global scale models and include input over the entire earth. Some are "coupled" ocean and atmospheric models that include ocean water conditions. Some are regional models that provide greater detail over a smaller area. Then there are models developed to forecast very specific types of weather, like hurricanes. The National Hurricane Center uses as many as six different models to forecast tropical storm development and movement.
The challenge, (or hard part), for the forecaster is to determine which method and which model to use on any given day. This requires a knowledge of the local geography and the local climatology (what usually happens here in this situation), knowledge of how small scale weather systems (thunderstorms, fog, etc.) behave, and knowledge of how the numerical models work, what they consider and what they leave out. Sometimes it's not too hard. Sometimes there is lots of conflicting information and the forecaster has to evaluate each piece of information and make the best educated guess he or she can as to what the atmosphere will do. A professor I once had likened it to playing chess with Mother Nature! If you've ever played chess you know its a game with very rigid rules, but two games are rarely ever exactly the same.
Check out these links from USA Today on weather forecasting:
How Computer Models Aid Forecasters
Some Operational Forecast Models In Use
Try the Earth Science links in the MadSci Library for more information on Earth Science .
Page generated by
MODERATOR_1.2b: Tools for Ask-An-Expert websites.
© 1997 Enigma Engines for a Better Universe: We are forever combustible, ever compatible.