The principle of weather forecast

Weather forecast is based on meteorological observation data, using the principles and methods of meteorology, dynamics and statistics to make a qualitative or quantitative forecast of the weather conditions in a certain area or a certain place in the future. Accurate weather forecast has always been an important goal of atmospheric science research. The history of weather forecast can start with observing the weather with clouds and forecasting the weather with objects at the earliest, and then go through the stages of single station forecast and weather map forecast. Up to now, advanced detection data such as meteorological satellites and weather radars are used to forecast the weather with computers. With the continuous progress of science and technology, weather forecasting has developed rapidly.

The types of weather forecast can be roughly divided into: immediate forecast (1 ~ 2 hours), very short-term forecast (2 ~ 12 hours), short-term forecast (12 ~ 48 hours), medium-term forecast (3 ~ 10 days) and long-term forecast (/kloc-0 days). According to different clients, it can be divided into daily weather forecast and professional weather forecast (such as aviation weather forecast, etc.). ); According to the forecast range, it can be roughly divided into regional forecast and site forecast. Due to different customers, there are some differences in forecasting items, forecasting timeliness and forecasting period.

At present, weather forecasting mainly adopts synoptic forecasting method, statistical forecasting method and dynamic forecasting method, and the weather-statistical forecasting method, dynamic statistical forecasting method and weather-dynamic forecasting method formed by combining these three basic forecasting methods.

Weather forecast method (or weather map method): it is a method of making weather forecast by using weather map as the main tool and satellite cloud map and radar map. Analyze and study the changing law of weather with the principle of meteorology. This method is mainly used for short-term forecasting.

Numerical forecast method (also called dynamic forecast method): it is a method to make weather forecast by solving dynamic equations describing atmospheric motion with a large-scale fast computer. This method can be used for both short-term and medium-and long-term forecasting. In recent years, it has also been used for climate prediction.

Statistical forecast method: it is a method to find out the statistical law of weather change by using a large number of long-term meteorological observation data, establish a statistical model of weather change according to the principle of probability statistics, and make weather forecast. This method is mainly used for medium and long-term forecasting and meteorological element forecasting.

The main ideas of these three methods of making weather forecast are different. The occurrence of weather phenomena (or weather processes) contains inevitability and contingency. The statistical forecasting method is based on the contingency of weather phenomena (or weather processes), and thinks that weather changes are a random process. Under the same conditions, the same weather changes may not necessarily occur, only the possibility or probability of a certain weather can be found. Based on the inevitability of weather phenomena (or weather processes), synoptic methods and numerical forecasting methods believe that weather changes are not random, and they satisfy certain laws (such as momentum balance, energy balance, mass balance, etc.). ), the same change should happen under the same conditions. According to the state of the atmosphere at a certain moment, the definite state of the atmosphere at the next moment can be calculated.

At present, three methods are commonly used to make weather forecast together, and the weather map, satellite and radar images, dynamic analysis and statistical analysis, numerical forecast products are comprehensively analyzed, and finally the weather forecast is made.

References:

http://2 10 . 45 . 66 . 147/chapter 09/dqhy 9 _ 0 . html