The energy in solar radiation heats up the land surface, evaporates water, and stimulates photosynthesis. Most of this energy is released back into the atmosphere and then transported vertically by turbulent motions, which enables the formation of clouds. In turn, clouds strongly modify the amount of solar radiation reaching the surface by scattering and absorbing the incoming radiation. While the propagation of radiation through the atmospheric is a complex three-dimensional process, radiation computations in atmospheric models are usually highly simplified. Such simplified computations are computationally affordable, but result in very unrealistic spatial patterns of solar radiation below clouds. In this thesis, we therefore combine realistic three-dimensional radiation computations with high-resolution atmospheric simulations to understand the tight coupling between solar radiation, clouds, and land surface processes with a great level of detail. Furthermore, we explore a couple of pathways for speeding-up radiation computations in atmospheric simulations.