Weather forecast technology

The skills and methods of weather forecasting are as follows;

(1) weather chart forecasting method

Weather chart forecasting method is a method to reasonably infer the future weather situation and the changes of related meteorological elements according to the meteorological principle by using weather charts and other auxiliary charts, and to predict the weather conditions in a certain period in the future. Its working procedures include meteorological observation, data transmission, drawing weather map, analyzing and inferring weather map, making weather forecast, etc.

1. Meteorological observation

Meteorological observation includes ground meteorological observation, high-altitude meteorological observation, atmospheric remote sensing detection and meteorological satellite detection, which are collectively called atmospheric detection. The meteorological observation system composed of various means can monitor the atmospheric state and its changes from the ground to the sky and from the local to the global.

2. Data transmission and weather map drawing

At the same time, the observed meteorological data are coded by various observation stations in China and transmitted to the National Meteorological Center by telegraph. The meteorological center will fill in the blank map with meteorological data from all over the country according to the prescribed format. This kind of special map, which is filled with simultaneous meteorological records in different places and can reflect the weather conditions in a certain area, is the weather map. Weather maps are mainly divided into ground weather maps and high-altitude weather maps, and there are some auxiliary maps.

3. Weather analysis and forecast

Forecasters draw isobars according to the air pressure values of each station on the ground weather map, analyze the distribution of high and low pressure systems, and determine the position of the front according to the air temperature, dew point temperature and weather distribution. This surface weather map comprehensively shows the weather systems such as surface front, cyclone, anticyclone, precipitation, gale, thunderstorm, hail and their influence areas at a certain moment. At the same time, contour lines and isotherms are analyzed, and various weather systems are marked on the upper-air weather map. The contour line reflects the position and influence range of weather systems such as high-altitude trough and ridge. According to several weather maps, the continuous change of weather system is analyzed, and its strength, position, moving speed and direction are judged.

(2) Numerical weather forecast

Numerical weather forecast is one of the important ways to realize objectification, quantification and automation of weather forecast, and it is also an important symbol to measure the development level of meteorological science in a country.

Numerical weather forecast is a method to predict the future weather by solving the atmospheric dynamic equations describing the weather evolution process through numerical calculation under certain initial values and boundary conditions. The atmospheric dynamic equation is based on the principles of fluid mechanics and thermodynamics.

Numerical forecast service is an automatic engineering system consisting of data collection, decoding, objective analysis, prediction model calculation, statistical test, product graphic output, data archiving, fax broadcasting and so on.

1. Data acquisition and decoding The National Meteorological Center collects and transmits real-time meteorological data by using high-speed supercomputers. All meteorological data collected by nearly 100 international and domestic routes are preliminarily processed and stored for further analysis. Generally speaking, more than 90% of ground observation data in China can be collected within 60 minutes from the observation time, and most high-altitude data in the northern hemisphere can be collected within 240 minutes. In addition, the collected unconventional data, such as temperature, humidity, pressure and clear sky emissivity, are also processed.

2. Objective analysis Objective analysis is the process of analyzing and sorting the observation data with irregular distribution into the dot data with regular distribution, which is suitable for the computer of numerical prediction model. It is different from manual analysis. Once a plan is given, the analysis results will not vary from person to person, so it is called objective analysis.

3. Forecasting model With the continuous upgrading of numerical forecasting model and the use of new high-speed supercomputers, the numerical forecasting business of the National Meteorological Center has developed rapidly. At present, the short-term circulation situation forecast adopts advanced models, which increase the forecast time limit from 60h to 72h, expand the forecast range from Eurasia to the northern hemisphere, and increase the forecast of upper-air physical quantities.

The update of the medium-term circulation situation forecast model has expanded the forecast range from the northern hemisphere to the whole world, increased the forecast time limit from 5 days to 7 days, and increased the precipitation forecast. 1992 updated limited area precipitation forecast model can forecast the precipitation and related physical quantities in China and its adjacent areas in the next 48 hours. In addition, the newly established rainstorm forecast model can forecast the precipitation and related physical quantities in the next 72 hours, and the typhoon model can forecast the typhoon track, sea level pressure, altitude field and 850 hPa wind field in 48 hours.

4. Products and their Exports About 70 kinds of numerical forecast products are used for meteorological facsimile broadcasting every month by the National Meteorological Center. Among them, more than 40 species are sent to stations all over the country through the first-class line, and the rest are sent to most stations in northern China through the second-class line. These products not only provide real-time weather service for local stations all over the country, but also can be used for weather analysis and research.

The application of local stations all over the country shows that the 500hPa situation forecast in the northern hemisphere has certain forecasting ability for the movement, development and transition evolution of the mid-high latitude weather system, especially the long wave trough ridge. It also has certain reference value for precipitation forecast in North China, Northeast China and Jianghuai area.

(3) Probability statistical prediction method

The occurrence of weather phenomena is inevitable and random. Probability and statistics method is based on the randomness of weather phenomena, and explores the hidden inevitability through a large number of historical data, thus making weather forecast. In other words, it is to analyze the statistical law of weather evolution and the quantitative relationship between forecast factors and forecast quantities by probability statistics method, and establish a mathematical model to forecast the future weather, which is probability statistical forecast.

There are many mathematical models for probabilistic statistical forecasting, including statistical forecasting methods suitable for central and provincial meteorological stations with advanced calculation tools, such as linear (nonlinear) multiple regression, balanced time series, discriminant analysis, Markov chain, orthogonal function expansion and so on. There are also simple statistical forecasting methods suitable for the vast number of stations lacking advanced calculation tools, such as contingency table, multi-factor correlation, simple regression and so on.

In recent decades, most meteorological stations in China have adopted the probabilistic statistical forecasting method, which has played a certain role in improving the accuracy of weather forecasting by combining numerical forecasting with weather map forecasting method. However, because the physical mechanism of atmospheric random phenomena, which is the basis of probability statistical forecast, is still unclear, and the meteorological data series in China is short, the simple probability statistical forecast is somewhat blind.

At present, the dynamic statistical forecast widely used in meteorological stations still belongs to the category of probability statistical forecast, such as the model output statistical forecast method (MOS method for short). With the development of numerical weather forecasting and the popularization of microcomputer, people regard long-term accumulated data (such as situation field and various physical quantities fields) and essential data as forecasting factors and objects respectively, and establish their mathematical relationship by statistical methods. When making daily forecast, the forecast value of future meteorological elements can be obtained only by substituting the numerical forecast products at that time into the relational expression.