Abstract Hydrological processes involve multi‐time scale interactions among different variables, and the interactions are driven, in part, by climate and landscape factors. Knowledge on the relationships of hydrological interactions with climate and landscape factors is crucial for modeling the hydrological processes, including the model structure determination and parameter regionalization. To quantify such multi‐scale hydrological interactions, the spectral analysis technique, which can capture the phase synchronization among variables at an interested frequency from a periodically related perspective, was adopted in this study. The hydrological interactions are from precipitation ( P ), soil moisture ( SM ), evapotranspiration ( E ), and snow water equivalent ( SWE ) to runoff ( Q ) in 447 catchments across the contiguous United States at various time scales. Correlation and interpretable machine‐learning methods were used to explore their relationships with climate and landscape factors. Results show that the complex spatial patterns of hydrological interactions at multi‐time scales are jointly influenced by aridity and snow fraction. Climate attributes are the most influential factors for various hydrological interactions while the landscape attributes also play a significant role under dry and snow conditions, especially at seasonal and annual scales. Based on the global wavelet coherence curve between effective precipitation ( P minus E ) and Q , we further specified the steady time scale for water balance and found a higher value in dry conditions, and the dominant influencing factors vary across different climate conditions. These findings deepen our understanding of climate and landscape control on hydrological processes across multiple time scales, providing valuable insights for hydrological modeling.
Liang et al. (Sun,) studied this question.