How can the current weather forecast be so inaccurate?

How can the current weather forecast be so inaccurate? All kinds of forecast analysis made by weather forecast are just a kind of "probability". For example, under the premise that the ambient temperature, standard air pressure and ambient humidity are certain, if it rains in 20% cases in history, the Meteorological Bureau will think that the probability of precipitation is 2%, and the probability of rainy days is not great. Then, combined with other data and information, when the weather forecast is broadcast, it may be said to be cloudy for the convenience of everyone, but whether it will rain in the end depends on God's mentality.

The meteorological department introduced in detail that the weather forecast seems simple, but it is actually a huge project. Weather forecast should be based on the basic theory of atmospheric science, various meteorological detection methods, standard weather forecast as the key, and comprehensive analysis by meteorological forecasters. At the same time, authoritative experts stressed that every stage of weather forecast has certain variability, and it is impossible for every weather forecast conclusion to be consistent with the specific situation. Improving the accuracy of weather forecast is still a global problem.

First, people's understanding of the principle of atmospheric motion is still relatively limited. Cold, hot and cloudy eyes, rainy and snowy weather, mixed wind and rain, and various temperatures are all caused by continuous aerial fitness exercises. Because of the diversity of atmospheric motion, experts can't really describe the tiny structure of atmospheric motion.

Second, the meteorological monitoring Internet can't "go from evil to collapse". Meteorological observation has developed into a magnificent three-dimensional observation system software covering subgrade, empty foundation. Road observation points, high observation points, automatic weather stations, radar observation points and climate stars constitute an observation network for monitoring atmospheric movement and change at all times. However, this Internet will ignore the small-sized weather system, just like a big website fishing for small fish, which is very easy to miss. Moreover, the observation data may be biased, for example, the wind frequency and wind observation results are the average of 2 minutes observation, which may be biased.

Thirdly, the physical model of standard weather forecast can't fully simulate the air evolution. The change of weather is influenced by the change of atmospheric motion around the earth. The change of atmospheric motion should conform to some basic laws of fluid mechanics and heat, which can be expressed by the language of mathematics class and written mathematical equations. We use high-performance computers to turn the problem of weather forecast into the problem of solving equations in math class. This method is called standard weather forecast, and it is also the key of contemporary weather forecast. But at this stage, all the physical models can't really simulate the air evolution, they are just similar, and there must be deviations.

Fourth, there will be differences in working experience and standards among meteorologists. The meteorological forecast conclusion calculated by calibration method can not be used as the meteorological forecast result at the same time, and it needs to be expressed and used by meteorologists and revised according to local conditions. For example, there is Yanshan Mountain in the north and Taihang Mountain in the west of Beijing, so the weather forecast must take into account the geomorphological hazards. The personal experience of meteorologists also plays an important role in the prediction of complex temperature and the overall management decision.