Source: China Meteorological News Press
For a long time, a platform that can roll hardware, data, procedure, platform, monitoring into one is only a conception. However, currently, this dream is on track to become a reality.
It is the meteorological big data cloud platform “Tianqing” under construction.
At the outset of 2020, China Meteorological Administration (CMA) has released an official document, which has outlined the construction of meteorological operational technical mechanism. The construction of Tianqing has bearings on research-oriented operation pilot, intensive development of meteorological operation, and operational technical mechanism reform. It can be hailed as the key core technical platform of CMA. What changes will be brought about after this platform is constructed? Here’s an overview of this platform.
Tianqing is designated to adapt to the era of the big data and promote the upgrade of the meteorological operational system.
The era of the big data has been unveiled. Currently, meteorological departments have amassed huge rafts of data resources. As monitoring becomes more refined and multi-source data are widely collected, meteorological data are on constant explosive increase. According to preliminary estimate, meteorological data inside and outside of meteorological departments will total 45PB in 2022, with a daily added volume of 40TB. In the meanwhile, the swift collection and processing of substantial data, long series data online analysis, multi-data integrated application, smart prediction and service algorithm based on data exploration and machine learning all call for the support from the big data technology.
The current meteorological operational mechanism needs to be in step with the era. Previously, national integrated meteorological information sharing platform (CIMISS) solved the problem of applying unified data source at the national and provincial levels. However, the information isolated island between all operational institutions and operational systems has not been radically addressed. Meteorological big data cloud platform construction can underpin the promotion of meteorological data resources integration and sharing, meteorological operation intensive development, and the optimization of operational procedures.
Against this backdrop, Tianqing has come into being. It will be responsible for solving the problem of intensive data and operation, effectively supporting meteorological operational system and application of all kinds of data, and promoting the transformation and upgrade of meteorological operations and administrative management technological system. Tianqing will harness new type information technologies like cloud computing, big data, and artificial intelligence (AI), support smart prediction and service system R&D, and fully underpin the operation of “cloud plus terminal”. And the infrastructure cloud platform under construction at the same time will provide hardware support for Tianqing.
Tianqing has largely expanded the scope of operation and become the main driving force of meteorological operations.
The essential feature of Tianqing is “cloud plus terminal”, which is also the development trend of meteorological operational technical mechanism.
Cloud refers to the meteorological big data platform, which takes charge of the comprehensive and authoritative earth system big data of observation, prediction, service, administration affairs, industry, and society, and provides the platform service unifying data, computing power, and algorithm.
“Terminal” encompasses operational systems of all kinds of observations, prediction and projection, services, and administration, as well as hardware terminals and software terminals like computers, mobile phones, and all kinds of sensors. Through pushing the profound integration of meteorological operational system and meteorological big data cloud platform, cutting back intermediate links as well as optimizing the procedures, it will promote the efficient coordination of observation, prediction, service and administration as well as the high quality development of meteorological cause.
The most salient feature of Tianqing is the integration of data and algorithm. Data refers to meteorological data resources and algorithm means the algorithm resources. The integration means the combination of data and algorithm resources. Previously, all meteorological operational systems invoked needed data respectively, and there were pervasive phenomena of duplicate data storage and transmission. And this integration can make algorithm be more connected with data, and decrease the data transmission on the internet. The computing support and operation of pipeline of the integration will optimize the overall operational procedures and promote the coordination of application.
In the future, the construction of Tianqing can fully demonstrate its strength. National and provincial core real-time data can realize minute scale synchronization alongside the online storage of 100PB data; the processed algorithm of all operational procedures can be integrated; the support of new information technologies can promote the smart development of meteorological prediction and service models.
In this open platform, meteorological departments will form a benign co-building and sharing operational ecology, which has dramatically lessened the workload and promoted research-oriented operational development.
In July, 2020, the 1.0 version of Tianqing was in operation at the national level meteorological operations. In July, 2020, provincial-level Tianqing was completed and will enter trial operation across the nation at the end of 2020.
This version of the Tianqing has fully covered the data management and service functions of CIMISS, and accelerated all procedures. On this basis, Tianqing has added algorithm database and processed assembly lines, prolonged the historical sequence of key data, and ramped up the real-time and historical data quality.
1.0 version of Tianqing is capable of large quantities of data storage, all operations run-through, and highly efficient data application, which can directly underpin the cloud integration of all kinds of meteorological applications like weather and climate prediction, atmospheric sounding, public services, and weather modification.
The development of Tianqing is well positioned to keep up with the times, accept new technologies, and meet the latest demands.
New technologies spurred by AI, blockchain, and 5G are all on track to flex their muscles in Tianqing.
For instance, Tianqing can underpin the training and construction of meteorological AI algorithm, and provide visualization exercise and assessment procedures. Currently, National Meteorological Information Center (NMIC) is constructing the training dataset of AI prediction of severe convection. In foreseeable future, AI will play an increasingly important role in meteorological cause with the support of Tianqing.
Driven by the demand, Tianqing is in constant evolution. It will embrace a better prospect as time goes by.
Editor: Liu Shuqiao