Abstract: In the analysis of high-impact extreme weather events, particularly in historical extreme value statistics, challenges such as inconsistent data sources, non-unified statistical methodologies, low computational efficiency for long time-series data, and overly technical data representations often lead to significant analytical errors, inefficiencies, insufficient impact assessments across sectors, and inadequate public communication. To address these bottlenecks, this study leverages the Meteorological Big Data Cloud Platform (Tianqing) to develop Extreme Value One-table Visualization Service System, a unified, visualization-driven analytical platform dedicated to historical ground-based meteorological extremes.
The system focuses on core meteorological variables, including precipitation, temperature, and wind, from long-term hourly and daily observations at 2400 national ground stations across China. By integrating GBase 8a, an analytical database optimized for large-scale data processing, the platform enables efficient retrieval and statistical analysis of terabyte-level meteorological datasets, supporting real-time, rapid queries over extensive observational records. Its hierarchical query architecture allows seamless integration of real-time observations, facilitating station-wise extreme value ranking since each station’s inception. Moreover, the system supports flexible queries by arbitrary spatial domains (e.g., administrative regions, river basins, or user-defined areas) and temporal windows (e.g., the past 5 or 10 years), catering to diverse research and operational needs. To enhance accessibility and visualization performance, the system adopts a browser/server (B/S) architecture. It features a responsive user interface built with the Vue.js framework, immersive 3D spatial pattern rendering via advanced web graphics technologies, and dynamic map interactions powered by WebGIS integration. These innovations significantly improve user experience, enabling the generation of high-resolution dynamic spatial distribution maps, multi-dimensional comparative analyses, and temporal trend visualizations, all with millisecond-level query response times.
Since its nationwide operational deployment in 2023, the system has served both national and provincial meteorological agencies, supporting post-event analyses of major weather events. By the year of 2025, the platform has accumulated 9.42 million visits, demonstrating its scalability, stability, and practical value in meteorological operations. By synergizing big data infrastructure with user-centered visualization design, this research provides robust support for scientific decision-making on climate change and extreme weather among researchers and the public alike. Future development will focus on extending statistical services to massive gridded datasets and deeply integrating cutting-edge AI technologies to enable conversational, question-answering-style user interactions.