Based on the existing concept and architecture of the smart runway, this paper addresses the challenge of managing the smart runway data flow. A systematic analysis of the data characteristics reveals that the data are massive, multi-source, heterogeneous, and multi-frequency. Based on this analysis, it proposes a multi-level data integration framework, a multi-platform physical structure, and a comprehensive data flow architecture for the smart runway. To support efficient and reliable data utilization, as well as classified and long-term data preservation, it develops detailed methods for data preprocessing and standardization. Additionally, it examines the characteristics of smart runway data storage, leading to the design of a corresponding multi-level storage architecture and appropriate storage strategies. The implementation of the proposed system at Rikaze Dingri Airport shows a reduction of over 95% in invalid data without any loss of effective information, achieving unified management and multi-source data sharing. This paper supports the enhanced capabilities of the smart runway, including real-time performance evaluation, advanced risk warning, and intelligent maintenance decision-making.