There is an increasing need for quantitative climate projections at a local scale. Within municipalities and utilities, water infrastructure planning in some regions is required to be performed with some perspective given to climate outlook data. This climate perspective is being encouraged since climate change brings potential risks to municipalities and water utilities associated with increasing local precipitation including more frequent combined sewer overflows, increased erosion, and increased riverine flooding. Global Climate Model (GCM) derived precipitation data is the typical tool for future quantitative assessments of physical climate impacts including precipitation changes, but raw GCM data cannot appropriately characterize local climate impacts due to its coarse resolution. Thus, downscaling of the raw GCM data to the local level is necessary
This paper will focus on evaluating the efficacy an event-driven method to execute dynamic downscaling of GCM data for use in defining and evaluating high resolution municipality and utility-scale precipitation data under multiple climate change scenarios. In this method, the entire future climate period is not modeled, but only that period predicted by the GCM to have significant precipitation. By mining the GCM data and picking a critical precipitation event over an asset area to simulate, the dynamic downscaling approach would change from a daunting task of modeling multiple GCM outlooks under multiple scenarios out to 100 years in the future to modeling fewer targeted simulations that capture the envelope of the future climate event.
