Biotic and abiotic complexity and heterogeneity of the landscape plays critical role in governing and influencing the ecohydrological processes on the landscape. These processes operate at a range of spatial and temporal scales, appropriate representation of which are critical for ecosystem modeling studies that focus on understanding the current and future state of the ecosystem. Capturing the subgrid heterogenity and their affects across scales is important for understand the current and future state of the landscape in changing climate. However, characterizing and modeling these processes are especially hard in data limited and ungauged landscapes like Arctic tundra landscape at Seward Peninsula of Alaska where our study region is located. To characterize and model this topographically complex mountainous landscape we developed a physics-informed data-driven approach. We characterize the heterogeneity of the landscape using a multivariate approach to hierarchically segment the landscape units across spatial scales. Using physics-based hydrology models we simulate and analyze the dynamics of the complex heterogeneous landscape mosaic at and across various scales to understand. Goal of the study is to understand the landscape organization and their role and influence in modulating the ecohydrological processes to improve our ability to accurately model them in data poor Arctic tundra landscape.