Abstract: Under the combined influences of global climate change and human activities, land surface hydrological processes exhibit pronounced nonlinearity and scale complexity, and the individual components of surface water, soil water, and groundwater, as well as their interactions, have become key factors constraining the accuracy and predictive capability of hydrological models. This study adopts the hydrological cycle as a central perspective and systematically reviews the interaction mechanisms among surface water, soil water, and groundwater, with particular emphasis on key processes including infiltration, evaporation, lateral recharge and discharge, and transitions between saturated and unsaturated zones. Major hydrological models in the field of water resources and hydrology are comparatively analyzed in terms of model structure, coupling strategies, and scale adaptability. Furthermore, development pathways of hydrological modeling are discussed from the perspectives of physically based models and artificial intelligence based approaches. The review indicates that existing models still exhibit notable limitations in multi media coupling representation, cross scale parameter consistency, and characterization of nonlinear processes. Trade offs among model complexity, computational feasibility, parameter identifiability, and cross scale adaptability remain prominent, while data driven approaches still face challenges related to insufficient physical constraints and limited extrapolation capability. Future advances in hydrological modeling require strengthening physical process representation while integrating multi source observations and intelligent algorithms to develop comprehensive modeling frameworks that balance physical consistency, computational efficiency, and scale adaptability, thereby better supporting hydrological research and decision making under complex environmental change.