What's the weather forecast?

Modern weather forecast has five components:

gather data

The most traditional data are data such as air pressure, temperature, wind speed, wind direction and humidity collected by professionals, enthusiasts, automatic weather stations or buoys on the ground or at sea. The World Meteorological Organization coordinates the timing of these data collections and sets standards. These measurements are made every hour (METAR) or every six hours (SYNOP).

Using weather balloons, meteorologists can also collect the temperature, humidity and wind force values in the sky. Weather balloons can rise all the way to the tropopause.

Data from meteorological satellites are becoming more and more important. Meteorological satellites can collect data from all over the world. Their visible photos can help meteorologists study the development of clouds. Their infrared data can be used to collect the temperature of the ground and cloud top. By monitoring the development of clouds, we can collect the wind speed and direction at the edge of clouds. However, because the accuracy and resolution of meteorological satellites are not good enough, ground data is still very important.

Weather radar can provide information about precipitation area and intensity. Doppler radar can also measure wind speed and direction.

Data assimilation

In the process of data assimilation, the collected data are combined with the numerical model used for prediction to produce meteorological analysis. The result is the best estimate of the current atmospheric state, which is a three-dimensional representation of temperature, humidity, air pressure, wind speed and direction.

Data weather forecast

Digital weather forecast uses computers to simulate the atmosphere. Based on the results of data assimilation, it calculates the changes of the atmosphere with time according to the results of physics and fluid mechanics. Because the equations of fluid mechanics are very complicated, digital weather forecast can only be carried out by supercomputers. The output calculated by this model is the basis of weather forecast.

Output processing

The original output of model calculation generally needs to be processed to become a weather forecast. These treatments include using statistical principles to eliminate the deviation in known models or making adjustments with reference to the calculation results of other models.

In the past, meteorologists had to do their own processing work, but today's weather forecast over 24 hours mainly uses a variety of different models to synthesize the results. Meteorologists must also analyze the predicted model data so that end users can understand it. In addition, the resolution of weather forecast models is generally not particularly high. Local meteorologists must also consider the influence of the region through local experience to make the local weather forecast more accurate. However, with the continuous refinement of the weather forecast model, this workload is getting smaller and smaller.

show

For end users, the display of weather forecast is the most important in the whole process. Only by knowing what information the end user needs and how to convey it to the end user easily can this task be completed.