Assessing the Value of Higher-Frequency Modeled Precipitation Data for Quantifying Risk from Changes in Extreme Precipitation

The NOAA Atlas 14 has long been considered the definitive resource to quantify probabilistic risks associated with extreme rainfall for infrastructure and engineering design. These risks are graphically characterized in intensity-duration-frequency (IDF) curves that indicate the probabilistic occurrence (i.e., frequency, or return period) that a measured amount of precipitation (i.e., intensity) could occur over a period (i.e., duration) and at a given location based on its historical observation record. Although jurisdictional planning units often are instructed to build to a 25-year, 24-hour storm, the IDF curves are produced for return periods out to 1000 years and durations ranging from 5 minutes to 60 days. Data that are not directly supported by the observational records (such as sub-hourly data and very long return periods) are imputed from statistical distributions that have been fitted to observational data.

As the Atlas 14 is being updated to Atlas 15, the new Volume 2 will be developed to consider climate change. In some methodologies, projected changes to precipitation IDF curves use statistical functions for sub-daily data when only daily projections are available. However, the current work explores using dynamical downscaling to produce climate projections with a finer temporal granularity, including hourly and sometimes 5-minute increments. This presentation will explore whether additional value is added by using high (sub-hourly) frequency precipitation data to derive IDF curves from modeled data. In this work, IDF curves computed from modeled historical simulations will be developed at selected U.S. cities, with a focus on comparing finer temporal increments of data.

Disclaimer: The views expressed in this presentation are those of the authors and do not necessarily reflect the views or policies of the U.S. EPA.