Adapting Precipitation from Climate Models for Use in Stormwater Modeling
Image credit: AGU WaterSciConAbstract
A multitude of climate models are available to simulate future weather events using different drivers, climate change scenarios, time steps, grid sizes, and bias correction methods. Model intercomparison projects and other guidelines can help researchers select which of these models are most appropriate for a given study site, but model outputs cannot easily be used for stormwater engineering tasks without manipulation. This is because stormwater models perform best with time steps smaller than the smallest time step available from climate models (1 hr). In addition, it is often desirable to isolate the impact of increases in precipitation magnitude, but this can be difficult to isolate from changes in precipitation patterns in historical vs. future outputs from climate models.
As part of a larger project assessing the resiliency of coastal military bases to climate change, we developed an approach to take precipitation data outputs from climate models and downscale the model timestep, making the temporal resolution high enough to be used as an input for stormwater modeling. We use python to process outputs from past and future climate models and calculate scale factors representing the changes in precipitation volume due to climate change for storms of representative sizes. We then applied appropriate scale factors to storms within a historical dataset with a high temporal resolution. This method 1) isolates the impact of climate change on precipitation magnitude, 2) preserves the historically observed pattern of precipitation, and 3) maintains the high-resolution time step of the historical data and may be useful for other studies with one or more of these goals.
Austin K. Farnum
PhD Student
