Abstract: Spatiotemporal analysis of recurring slope lineae (RSL) heavily relies on the effective integration and correlation mining of multi-source heterogeneous data to provide a reliable scientific basis for understanding RSL formation mechanisms. Based on this, an RSL knowledge graph covering the Coprates and Melas Chasma within the Valles Marineris canyon system was constructed using the Neo4j graph database, with HiRISE (high resolution imaging science experiment) high-resolution remote sensing images as the primary data source. Subsequently, key topographic and seasonal factors controlling RSL distribution were identified through graph queries and spatiotemporal correlation analysis. Furthermore, a quantitative relationship model between RSL activity, slope aspect, and solar longitude was established. Based on 364 time-series images from 101 RSL sites, the spatiotemporal distribution patterns of RSL were systematically analyzed. Finally, by integrating previous observational results and climate event records, the effectiveness of the constructed knowledge graph in revealing RSL spatiotemporal characteristics was validated. The results indicate that RSL distribution exhibits a significant westward preference, and seasonal activity is closely coupled with slope aspect, forming an alternating intra-annual activation pattern. Following the MY34 global dust storm, RSL activity was observed on all slope aspects in Garni Crater, marking a stark contrast to previous years when activity was limited to specific aspects, confirming the significant impact of climatic events on RSL scale.

Key words: Mars, recurring slope lineae (RSL), knowledge graph, spatiotemporal characteristics, seasonal factors, climate events