Criterion for Analyzing the Ability of Climate Models to Represent Major Extreme Events

Abstract:

Detection of extreme events with high accuracy is a major challenge in modeling towards understanding the potential impacts of the changing climate on society. The purpose of this study is to examine major extreme events in the U.S. Midwest and Northeast over the last 40 years (1980-2019) to determine how well a regional model using reanalysis input can represent observed events that have had a large impact. In the process, we consider five examples each of four types of major observed extreme events – heat waves, extreme precipitation events, droughts and floods. Each of these events have different criteria like intensity, duration, region of influence, and others to determine ranking. These criteria are used to compare the observed and simulated events over these regions. For this study, we used the NCEP-R2 reanalysis data as the initial and boundary conditions in the Weather Research and Forecasting (WRF) model over this time period. The model was run at a grid spacing of 12km and covers most of North America. Initial studies have shown that the WRF datasets can identify the temperature based extreme events with a greater accuracy compared to the precipitation based events, but this study represents a different way of examining the capabilities of such models to represent extreme events. The aim is to develop a new metric for such analyses. This study is part of an interdisciplinary project funded by NSF to understand the effects of the changing climate on the food-energy-water system.