The effectiveness of measured and derived tropical cyclone parameters in predicting coastal damage

Recent hurricanes have exposed the inadequacy of the Saffir-Simpson hurricane scale (SS) in gauging the severity of tropical cyclone (TC) impacts at landfall. The SS uses a single intensity measurement taken at one location within a TC to describe a storm covering thousands of square miles. Considering multiple variables such as TC intensity, size, forward speed, strength, and the persistence of these variables all prior to landfall, might prove crucial in building a more effective rating scale.

Pielke and Landsea's (1998) methodology of normalizing hurricane damages to correct for annual changes in wealth, inflation, and coastal population is utilized in order to compare TCs on an equal dollar value. Observations of thirteen parameters from the Extended Best Track (EBT) dataset are organized by time prior to TC landfall, then plotted against the normalized damage data in a series of scatterplots. A new duration parameter for each of the EBT parameters is created by averaging observation values over varying time intervals prior to landfall. These duration parameters are also plotted against damage totals, and all scatterplot data is subjected to regression analysis to determine the strength of correlation between each atmospheric parameter and resultant storm damage.

The goal of this study is to isolate those atmospheric parameters whose values remain most consistent up to three days prior to TC landfall. Such parameters, once isolated, would not only serve as reliable predictors of TC damage, but they could also provide a basis for a more accurate future TC rating scale.