Quantifying Climate Change Impacts on Infrastructure: A Statistical Downscaling Application

English Abstract:

Over the coming century, climate change has the potential to impact infrastructure in many different ways, particularly in population-dense areas that depend on transportation and built environments. Many of these impacts may occur via changes in the frequency and magnitude of extremes: high and low temperature, precipitation, coastal flooding, and storm events.

In some cases, it is not yet possible to quantify how global change will affect local extremes, and/or how climate change might impact society. At the same time, resilience can be built into planning through preparing for conditions that already occur: tornadoes, derechoes, or even extreme thunderstorms. For other applications, simply the direction of change is sufficient to inform future planning: will hurricanes become stronger, or winter storms more or less frequent? For some impacts, though, particularly those that show significant historical trends as well as consistent future change and quantifiable impacts, it is possible to resolve both projected changes as well as impacts under a given scenario.

In this presentation, we describe a framework for developing and implementing future projections using the example of a set of infrastructure-relevant climate indices for the state of Delaware, USA. The indices are based on mean and extreme temperature, precipitation, humidity, and combinations thereof calculated from GCM simulations from nine models corresponding to a lower and a higher scenario, statistically downscaled to 14 individual station records using the ARRM model. We also describe the process of selecting an appropriate subset of GCMs and future scenarios, as well as how to extract useful information from a large selection of downscaled output with associated uncertainties. The results highlight the importance of including future climate projections into infrastructure planning, as opposed to using the past as a guide when planning for the future.