Land surface models are increasingly used in simulating land surface processes at hyper-spatial resolutions (e.g., ~1 km). As spatial resolution increases, grid-scale topographic effects on surface radiation fluxes and interactions between adjacent grids become more and more pronounced, e.g., the slope/aspect effects, self-shadowing and shadowing cast by adjacent grids, and multi-scattering from neighboring grids. However, current land surface models neglect the fine-scale topographic effects on surface radiation balance. This study developed a physically-based parameterization (fine-TOP) that explicitly resolves the fine-scale topographic effects on downward shortwave radiation, downward longwave radiation, and land surface albedo and emissivity, and considers the impacts of surface area enlargement on surface radiative fluxes in the state-of-the-art Energy Exascale Earth System Model (E3SM) Land Model (ELM) version 3. The fine-TOP successfully captures the topography-driven variations in the diurnal cycles of downward shortwave radiation as well as the diurnal asymmetry of land surface albedo. The 1-km ELM simulations over the California Sierra Nevada show that fine-scale topography has significant impacts on surface energy, water, and carbon cycles. ELM with fine-TOP shows similar aspect-dependence of snow cover fraction and land surface temperature found in MODIS data, while the default ELM fails to capture this phenomenon. This study enhances the ELM’s capability to accurately represent fine-scale topographic effects on surface radiation balance and advances our understanding of the role of fine-scale topography in land surface processes across mountainous regions.