Hydrologic research is increasingly data and computationally intensive, and often involves hydrologic model simulation and collaboration among researchers. With the development of cyberinfrastructure, researchers are able to improve the efficiency, impact, and effectiveness of their research by utilizing online data sharing and hydrologic modeling functionality. However, further efforts are still in need to improve the capability of cyberinfrastructure to serve the hydrologic science community. This dissertation first presents the evaluation of a physically based snowmelt model as an alternative to a temperature index model to improve operational water supply forecasts in the Colorado River Basin. Then it presents the design of the functionality to share multidimensional space-time data in the HydroShare hydrologic information system. It then describes a web application developed to facilitate input preparation and model execution of a snowmelt model and the storage of these results in HydroShare. The snowmelt model evaluation provided use cases to evaluate the cyberinfrastructure elements developed. This research explored a new approach to advance operational water supply forecasts and provided potential solutions for the challenges associated with the design and implementation of cyberinfrastructure for hydrologic data sharing and modeling.