Big data helps us visualize the sky.

Http://www.aliyun.com/zixun/aggregation/32255.html "> Weather forecasts are now like daily necessities, the information people need to master every day." But how do you predict weather conditions after 40 days? How to make the weather forecast more accurate? In the past, such problems are difficult to solve, but now different, big data can help. With the improvement of the global Meteorological Observation Network and the increasing number of numerical prediction products, the meteorological data has reached the threshold of large data in volume, and the meteorological field has entered into the big data age quietly.

The lack of meteorological data and the error of observation accuracy restrict the meteorological forecast level. The traditional numerical prediction theory and method can only obtain the basic credible forecast within 10 days, and the accuracy of the time is decreasing rapidly. Under the support of abundant meteorological information, people can directly analyze the correlation between different time, different place and different weather phenomena through large data, and establish specific models to make predictions. Last year, a technology company in the United States fully used the large data analysis method to predict the future weather. It builds a weather forecast model based on 82 billion calculations and meteorological history data from 60, and then forecasts the weather by comparing the model to the current climatic conditions. This is the longest way to predict the weather in the next 40 days, including a certain temperature, air pressure, humidity, wind and precipitation and other detailed meteorological elements.

In the future, smart phones and smart TV will become large data sources of information, and even have some data processing capabilities. In the future, meteorological information collection and disaster monitoring warning is no longer the "patent" of the Government and meteorological Department, and everyone may play an unexpected role in predicting the weather and conveying the disaster. Recently listed smartphones are equipped with barometer, hygrometer ﹑ thermometer and illuminance meter, these sensors can obtain important meteorological data. In 2013, the British Meteorological Office used personal mobile terminals to collect meteorological data to track the geographical location and scope of snowfall. A large amount of data collected by smartphones may change the way people understand and predict the weather, according to the journal Scientific American.

With large data This tool, the numerical weather forecast will be more accurate, more timed, fixed-point and quantitative weather forecast products will play an important role in the major tasks such as combat. For example, the spacecraft launch on the meteorological conditions are very high, the cloud, precipitation, thunderstorms, wind and other meteorological elements of the forecast requirements to be refined to 1 km. With the support of large meteorological data, the use of computer and cloud computing, the refinement of numerical computing is no longer a problem. Meteorological scientists can also assimilate a large number of global meteorological satellite imagery and meteorological radar data into high-resolution numerical forecasting models, shortening the forecast time from every 3 hours to rolling forecasts per hour.

In the

Big Data age, weather forecasts not only affect people's travel but also bring about great changes in people's lives. The combination of large meteorological data with many areas of social science or business behavior will show new value and dramatically change the way people live. For example, some Web sites that rely on the weather forecast for directional precision marketing can use the forecast to put insecticide ads in dry areas, add compliant shampoo ads in cold and humid areas, put on shampoo ads that increase hair elasticity in dry areas, and produce foods that people prefer in hurricanes before hurricanes, These attempts have been very good results.