Water, through its phase changes and cycling through the Earth system, holds a major key to understanding climate variability and change. Even as the basic view of the global water cycle was established over a century ago, gaps in understanding, quantifying, and predicting the water cycle and its changes remain large. Numerical models have been used prevalently for hypothesis testing, prediction, and projection but modeling water cycle processes is intractably difficult, owing to their multiscale, multiphase nature. With recent advances in computing, high resolution models are becoming important discovery tools for climate science, even if their use in climate prediction and projection is limited by computing resources at this time. Realistic simulations of mesoscale water cycle processes have important implications for large-scale circulation and hydrologic impacts. Improving modeling of mesoscale processes may be a useful target for bridging weather and climate modeling, but achieving this goal would also require focused efforts in improving observation of mesoscale processes and computational advancements.