Warning systems with the ability to predict floods days in advance can benefit tens of millions of people. Because of these potential impacts there have been efforts to improve prediction systems such as the United States’ Advanced Hydrologic Prediction Service and European-developed Global Flood Awareness System. However, these projects are currently limited to relatively coarse resolutions. This thesis presents a method for downscaling and routing global runoff forecasts generated by the European Centre for Medium-Range Weather Forecasts using the Routing Application for Parallel computatIon of Discharge program that make possible orders of magnitude increases in the density of the resolution of stream forecasts. The processing method involves using the Amazon Web Services to distribute execution in a cloud-computing environment to make it possible to solve for large watersheds with high-density stream networks. Using the Amazon Web Services, the number of streams that can be used in the downscaling process in a twelve-hour period is approximated to be close to five million. In addition, an application for visualizing large high-density stream networks has been created using the Tethys Platform of water resources modeling developed as part of the CI-WATER NSF grant. The web application is tested with the HUC-2 Region 12 watershed network with over 67,000 reaches and is able to display analyzed results to the user for each reach.