Engineers typically rely on the assumption of nonstationarity to apply historical information to future periods during which the transportation infrastructure will serve its intended purpose. However, with projections of a changing climate, transportation agencies and policy makers are asking engineers to consider future scenarios without relying on nonstationarity. To accomplish this, engineers using rainfall/runoff models need estimates of future daily and sub-daily precipitation as inputs at watershed-level spatial scales.
This presentation addresses the research project findings and recommendations for actionable guidance for estimating projected daily precipitation data for use in rainfall/runoff modeling, including a 10-step procedure that involves selection of a portfolio of downscaled GCM outputs using a range of future scenarios. The procedure describes a method for using GCM-based information in conjunction with historical precipitation for estimating projected precipitation and estimating mean values and confidence limits for each scenario.