Cross-scale studies in geomechanical and hydrological systems employ a variety of approaches, either experimental, simulation or theoretical, each characterized by corresponding scale-specific methodologies. This perspective identifies and discusses challenges encountered at various scales, ranging from molecular to field scale, and examines issues related to integrating these scales. It highlights discrepancies in resolution and data compatibility, emphasizing the necessity for improved scale transition techniques. Insights and recommendations are proposed for future research to enhance multiscale modeling frameworks. These suggestions are crucial for bridging knowledge gaps on geological systems and improving the analyses accuracy for better engineering applications or earth system modelling.