Insights into Building Damage and Vulnerability from the Recent Severe Thunderstorm Outbreaks: Analyses of Post-damage Surveys and Insurance Claims Data

摘要

The frequent occurrence of severe thunderstorms (ST) in recent years has caused substantial life and property losses in the United States. The 2011 severe thunderstorm outbreaks alone caused a record-breaking 26 billion dollars in losses for the North America insurance industry- as per Property Claim Services (PCS). In 2012, total insured losses reported by PCS were on the order of 15 billion dollars. The majority of the time, these storms affected the non-hurricane prone regions of the US - "Midwest and Central Plains". Hail, tornadoes, and straight-line winds were the main contributors to damage from such events. Apart from significant levels of wind and hail hazard from these events, vulnerability of buildings in these areas also exacerbated the level of damage and losses. Buildings in these regions are generally designed to a lower than wind speed than those imposed by extreme wind events such as Hurricanes. Experimental and numerical techniques are generally employed by engineers to study the impact of extreme weather events on buildings. However, tornado wind fields and falling hail are difficult to accurately simulate. Hence, engineers have mainly depended on post-disaster survey data to rate the hazard intensity and assess the level of damage. In catastrophe modeling - a study area in which risk from natural catastrophes is assessed - damage survey data can also be supported by actual insurance claims data and used to inform the vulnerability assessment of buildings. hi this paper, blending of the insurance claims data and post-damage survey data for the assessment of building vulnerability from tornadoes and hail is studied. The relationship between the damage ratio (repair cost divided by the replacement value of a building) from insurance claims data of various companies and the EF-scale's degree-of-damage from the field surveys, along with the progression of damage across the tornado path, and the impact of several building characteristics on the damage at a given hazard intensity level were investigated. In general, the data and the practical experience gained from the damage surveys proved to be instrumental for validation and calibration of resulting vulnerability functions.