Estimating the Expected Number of U.S. Lightning Fatalities During a Year, Throughout a Year, and Comparison to Other Storm Phenomena

The number of fatalities from lightning has been steadily declining in the United States for many decades. This makes it difficult to estimate the expected number of lightning fatalities in a particular year and whether the actual fatality rate is statistically significantly above or below the expected value. One common practice for weather fatalities is to use the most recent 30-year average. Unfortunately, that can be misleading if the fatalities from a particular weather phenomena have a significant trend during that period. One common remedy is to use the most recent 10-year average, to make the estimated rate more representative of the current time. Unfortunately, that has the drawback that a single extreme event, large or small, can skew the results. In addition, a 10-year average doesn't completely eliminate the lag so the estimate is still not completely representative for the current year. A better approach is statistical curve fitting of the fatality rate over many years to estimate the fatality rate and its error bars for the current year or any year in the near future. This allows hypothesis testing to see if the actual rate is significantly different from the expected rate. A percentile regression of the expected number of lightning fatalities throughout the current year or near future year is also presented. Statistical models will also be developed for the other sources of storm deaths in the U.S. in addition to lightning: floods, tornadoes, and hurricanes. The fatality rate of lightning will be compared to these other weather phenomena. Lightning has now dropped to the third leading source of storm deaths in the U.S. with tornadoes rising to second place, and lightning may even be falling below the fatality rate of hurricanes and straight-line wind.