Multi-level grid management and retrieval model for rainfall data using sequential encoding

The increasing severity of extreme weather events necessitates efficient spatiotemporal data retrieval technologies for rapid response. This study introduces an innovative model for rainfall data management and retrieval using sequential encoding based on the GeoSOT framework (RDMR_GeoSOT). The model incorporates a multi-level grid management method (GMM-RD) and a grid retrieval method (GRM-RD) to enhance data retrieval efficiency during extreme weather events. To evaluate the feasibility and effectiveness of the GRM-RD, it was applied to mainstream databases and compared with their spatial expansion modules. The experiments utilized simulated rainfall datasets of varying sizes, 0.5 million, 5 million, and 50 million records, to assess retrieval performance across different target areas. The results demonstrate that the GRM-RD effectively addresses the challenge of real-time retrieval for large-scale spatiotemporal data in extreme weather emergency response scenarios, while simultaneously reducing spatial database indexing complexity, with an overall performance improvement of 41.70 times compared to PostgreSQL + PostGIS and 28.01 times compared to Oracle Spatial. Moreover, sensitivity analysis further shows that the GRM-RD consistently achieves retrieval efficiency improvements exceeding 11 times, even with increased computational demands from finer spatial partitioning. These findings offer valuable insights for advancing spatiotemporal data management frameworks and practical pathways toward addressing climate-related issues.