Are Tornadoes Becoming Stronger? A Statistical Model for Tornado Intensity

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Monday, 3 February 2014: 11:15 AM
Room C205 (The Georgia World Congress Center )
James Elsner, Florida State University, Tallahassee, FL; and T. H. Jagger and I. J. Elsner

A cumulative logistic model for tornado intensity is developed and examined. Damage path length and width are significantly correlated to the odds of a tornado receiving the next highest damage category. Given values for the cube root of path length and square root of path width, the model predicts a probability for each category. The length and width coefficients are insensitive to the switch to the Enhanced Fujita (EF) damage scale and to distance from nearest city although these variables are statistically significant in the model. The width coefficient is sensitive to whether or not the tornado caused at least one fatality. This is likely due to the fact that the dimensions and characteristics of the damage path for such events are always based on ground surveys.

The model predicted probabilities across the categories are then multiplied by the center wind speed from the categorical EF scale to obtain an estimate of the highest tornado wind speed on a continuous scale in units of meters per second. The estimated wind speeds correlate at a level of .82 (.46, .95) [95% confidence interval] to a small sample of wind speeds estimated independently from a doppler radar calibration. The estimated wind speeds allow analyses to be done on the tornado database that are not possible with the categorical scale. The modeled intensities can be used in climatology and in environmental and engineering applications. More work needs to be done to understand the upward trends in path length and width. The increases lead to an apparent increase in tornado intensity across all EF categories.

Supplementary URL: http://rpubs.com/jelsner/TornadoIntensity