No Explanation of the Evaluation Methodology
On the surface, the evaluation of risk appears questionable. For example, while the yearly risk of occurrence of tornados would surely be higher in the central plains (Dallas), the risk of damage from a nearby event is relatively low. On the other hand, yearly risk of occurrence of a significant earthquake event would be much lower, but the chance that damage would occur from a nearby event would be much higher than with a tornado. I would guess a hurricane would fall in between in terms of both risk of occurrence and area affected by an event. Furthermore, it would be helpful to understand what risk is being described - permanent damage, repairable damage, loss of life, etc. Perhaps a better measure would be normalized insurance claims, though you would have to address how to normalize for dollar value, population density etc.
