Analyzing the Extreme Behavior of Large-Scale Meteorlogical Variables Found To Have Influence on Severe Storms and Tornadic Events Using Global Reanalysis Data

• Activity Number: 530
• Type: Topic Contributed
• Date/Time: Thursday, August 10, 2006 : 10:30 AM to 12:20 PM
• Sponsor: IMS
• Author(s): Eric Gilleland*+ and Matt Pocernich
• Companies: National Center for Atmospheric Research and National Center for Atmospheric Research
• Keywords: extreme values ; climate change ; reanalysis data ; spatial ; temporal ; GEV

Abstract

Concurrently high values of convective available potential energy (cape) and 0--6 km vertical gradient of horizontal weight (shear) are found to occur with severe storms and tornadic events, which occur at fine scales that cannot be resolved by climate models, unlike cape and shear. Difficulties arise in analyzing these events (e.g., cape is typically zero, high amounts of both do not occur often at the same time, the spatial structure is highly nonstationary). Initial analysis is on global reanalysis data on a 1.9 by 1.9 degree grid (approx 17,000 points) every 6 hours from June 1, 1957, through December 31, 1999. The relationship of cape and shear is inspected and significant changes over time are investigated, accounting for the false discovery rate. Finally, data are fit to the GEV distribution individually at grid points using constrained MLE.