“Tianqing” — the data engine for the future of Meteorology

It integrates 100 million megabytes of data into one, integrates business systems into one, integrates hardware, data, process, platform and monitoring, and realizes second level synchronization of data and second level response of demand For a long time, such a platform that can integrate the information resources of meteorological departments and greatly improve the efficiency of business scientific research exists in the concept. Now, that dream is coming true. At the beginning of this year, the No.1 document issued by China Meteorological Administration focused on promoting the key reform of meteorological operation technical system, and clearly required to build a “cloud + end” Meteorological business technical system with meteorological big data cloud platform as “cloud” and meteorological business system as “end”. < / P > < p > < p > the construction of “Tianqing” is related to the key tasks of the Party group, such as research-oriented business pilot, intensive development of meteorological business, business technology system reform and so on. It can be described as the key common core technology platform of China Meteorological Administration. So what changes will the platform bring to the existing business? Let’s get closer to Tianqing and have a look at the scenery on the cloud. < / P > < p > the era of big data has begun. At present, the meteorological department has accumulated massive data resources. With the more precise monitoring and extensive collection of multi-source data, the meteorological data continues to grow explosively. It is preliminarily estimated that the meteorological data inside and outside the Department will reach 45pb by 2022, with a daily increment of 40tb. At the same time, the rapid aggregation and processing of massive data, online analysis of long series data, comprehensive application of multi-source data inside and outside the Department, intelligent prediction and service algorithm based on data mining and machine learning need big data technology support. < / P > < p > the current meteorological operational system and mechanism urgently need to connect with the era fast track. Previously, the national integrated meteorological information sharing platform has solved the problem of unified data source application in the country and province, but the “information island” and “application chimney” problems among business units and business systems have not been fundamentally solved. The construction of meteorological big data cloud platform is an important starting point to promote the integration and sharing of meteorological data resources, promote the intensive development of meteorological business, and promote the optimization and reengineering of business processes one of. In this context, “Tianqing” emerges as the times require. It will be responsible for solving the problems of data intensive and business intensive, effectively supporting the meteorological business system and various data applications, and promoting the transformation and upgrading of meteorological business and government management technology system. “Tianqing” will adopt cloud computing, big data, artificial intelligence and other new information technologies to provide cloud computing services of “digital computing integration”, support the development of intelligent forecast and service system, and comprehensively support the ecological operation of “cloud + end”. The infrastructure cloud platform built in the same period will provide hardware support for Tianqing. < p > < p > in terms of inheriting the achievements of standardization construction and innovating the technical level, “Tianqing” is born out of cimiss, but it has greatly expanded its business scope, and has become the general engine of meteorological business with a more open attitude of “integration of digital computing” and “cloud + terminal”. The essential feature of “Tianqing” is “cloud + terminal”, which is also the development direction of meteorological operational technical system. “Cloud” refers to the meteorological big data cloud platform, which manages the complete and authoritative earth system big data of observation, forecast, service, government affairs, industry and society, integrates the whole process business product algorithm of quality control, statistical processing, forecast and early warning, unifies the product processing pipeline, and provides the platform service of “data, computing power and algorithm”. < / P > < p > < p > “terminal” includes all kinds of business systems such as observation, prediction, service and management, as well as hardware terminals and software terminal programs such as computers, mobile phones and various sensors. By promoting the deep integration of meteorological business system and meteorological big data cloud platform, reducing intermediate links, optimizing the process, realizing the integration of data, technology and business, promoting the efficient collaboration of observation, forecast, service and management business, and promoting the high-quality development of meteorological cause. < p > < p > < p > “integration of number and calculation” is the biggest feature of “Tianqing”. “Number” refers to meteorological data resources, “calculation” refers to algorithm resources, and “integration of data and calculation” refers to the integration of data and algorithm resources. In the past, it is very common for each meteorological operational system to call its required data and store and transmit repeatedly. The “integration of data and calculation” makes the algorithm close to the data, so as to reduce the transmission of data on the network. Algorithm developers and operators don’t care where the data is, they just need to develop and submit algorithms to run the product. With the support of computing power and pipeline scheduling, the global business process will be optimized and application collaboration will be promoted. < / P > < p > the pipeline scheduling processing of meteorological products enables users to focus on algorithm research and development, without paying attention to the operation environment, data resources, operation efficiency, product management, operation monitoring and other affairs. < / P > < p > in the future, with the integration of isolated data islands in the meteorological field and the application of chimneys to break down one by one, “big” and “fast” are no longer “you can’t have both”. At the same time of 100pb level massive data online storage, the national and provincial core real-time data can realize second level synchronization; the processing algorithm of the whole business process will be unified and integrated, which can run according to the business natural time sequence and pipeline scheduling; real time interactive application can realize sub second level response, historical long sequence analysis and multi-dimensional spatial analysis second level response; new information technology support can drive meteorological forecast Intelligent development of newspaper and service model. < / P > < p > in such an open, application integrated and clear intellectual property platform, the meteorological department will form a benign business ecology of co construction and sharing, greatly reduce the R & D threshold and workload, and promote the development of research-oriented business. < p > < p > in July 2019, the newly formed version 1.0 “Tianqing” was put into operation in the national meteorological service. In July 2020, the provincial “Tianqing” will be deployed and will be put into trial operation nationwide by the end of this year. < / P > < p > although this version of the sky engine is not “complete body”, it has fully covered the data management and service functions of cimiss, and realized the whole process speed-up. On this basis, “Tianqing” added an algorithm library and processing pipeline to extend the historical sequence of key data and improve the quality of real-time and historical data. Version 1.0 of “Tianqing” has the ability of massive data storage, full business connection and efficient data application, which can directly support the cloud integration of various meteorological business applications such as weather and climate prediction, atmospheric detection, public services, weather modification, etc., and form a “cloud + terminal” ecological environment. < p >, social and economic data from domestic and foreign industries. Data integrity authority. < p > < p > and “Tianqing” will not stop here. It is characterized by openness, and will be able to continuously improve with the progress of the times, accept the latest technology and meet the latest needs. Artificial intelligence, blockchain, 5g All kinds of new and cutting-edge technologies that trigger topics are likely to show their talents on “Tianqing” through open interfaces. < / P > < p > take artificial intelligence as an example, “sky engine” can support the training and construction of meteorological AI algorithm, and provide visual training and evaluation process. At present, the National Meteorological Information Center is building the AI training data set for severe convection prediction. In the foreseeable future, AI will play an increasingly important role in meteorology through the incubation of “Tianqing”. In addition, 5g communication technology, which has not been applied on a large scale, also has a corresponding interface in “Tianqing”; blockchain technology may also achieve breakthroughs in copyright protection, statistical analysis of user needs, etc Driven by demand, “Tianqing” is constantly improving and evolving. Thanks to the large scale of data storage and the design concept of efficiency first, “Tianqing” can respond quickly. On the one hand, it does not consume additional resources, on the other hand, it builds the database and even algorithm library that is most suitable for users’ needs. Looking forward to the future, “Tianqing” is on the road of high-speed development together with the meteorological industry, and is firmly moving forward towards a more intelligent, efficient and intensive tomorrow. Continue ReadingVideo Number assistant internal test online! Four functions let you send 1g video on the computer