Using the past to predict the future is a primary raison d’etre for many monitoring networks in support of engineering design. This worked well for the 20th century resulting in efficient and water-resilient of most of our built infrastructure. However, the future is not what it used to be. The authors compare three methods of using observed data for predicting future rainfall intensities. Combining observational data with modeled futures will almost certainly need to be incorporated into 21st century engineering design standards.