Tornadoes getting stronger and stronger??? – Is it good?

This study, not too surprising based on the title, claims that tornado strength has increased by an average of 5.5% per year between 1994-2016 as shown in Figure 1 below: https://myweb.fsu.edu/jelsner/PDF/Research/ElsnerFrickerSchroder2019.pdf

I find this result at least intriguing, especially considering that intense tornadoes (EF3-EF5) appear to be decreasing in frequency. So I investigated further and did a bunch of numbers with the actual tornado dataset available from the Hurricane Prediction Center (SPC), which can be downloaded from the following website:
www.spc.noaa.gov/wcm/data/1950-2019_actual_tornadoes.csv

According to the article, the strength for a tornado is calculated using (i) an estimate of the wind speed (taken from the midpoint of the wind speed range of each type of EF, or 97 m/s in the case of Unbound EF5, (ii) tornado affected area (calculated by multiplying track length by width), and (iii) segmented track area estimates for each EF type (eg. for example, a tornado rated EF3 will actually consume 53% of its life as EF0, 27% as EF1, 13% as EF2, and only about 7% as average EF3). These fractional estimates for each EF tornado rating can be found in Fricker & Elsner²: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0131090

Section (ii) will be the focus of this article, specifically the historical path length and width data. At first glance through all of history, it is surprising to see so many tornadoes occurring with a path length equal to 0.1 mile and/or width equal to 10 meters. It’s almost as if this were some sort of default (or minimum) value to be specified if the actual path length (PL) or width (PW) is unknown or not measured. The data below shows the percentage of the time these values ​​occur over different time periods in the data set. Remember the authors’ results for the period 1994-2016.

Has anyone noticed a big problem?! Of particular interest, it appears that the discontinuity in PL occurred as soon as the F-scale was replaced by the EF-scale in January 2007. Digging deeper, below are the values ​​for PL & PW at levels. 25^order & 50^order percentiles for all tornadoes (EF0-EF5) over the two study spans of the authors.

The results clearly show that the pre-2007 path length and width are contaminated by a lot of very small path length = 0.1 mi and path width = 10 yd. Meteorologically, it is unpredictable whether the length of the track will increase by 2-3 times or the width of the tornado will increase by 50% in just 13 years. Climate change is not drastic! Unfortunately, this bias is also present in tornadoes with higher EF ratings (EF2-EF5).

The effect of faulty path length and width data should be obvious for a study spanning two periods. In particular, the cyclone energy dissipation data in the first part of the study (1994-2006) will be artificially low compared to the second part (after 2006) resulting in wrong uptrend. As a result of this finding, The conclusion from the authors that tornadoes are getting stronger and stronger should be completely disproved.

Another small flaw in the study is that the authors only plotted the annual, quartile, and 90 mean^order percent tornado power instead of plotting the total annual capacity. Using the former can lead to erroneous conclusions, especially if the number of tornadoes is increasing or decreasing over the time period. For example, let’s say a given year has 100 tornadoes in each EF category and the next year there are 200 tornadoes in each EF category. Graph the average strength or 90^order percentile values ​​will not show an increase from year to year, while plotting totals will provide a clearer picture.

Interestingly, assuming that the post-2007 data reasonably represent what actually happened, below is a graph of total annual tornado capacity for all tornadoes (EF0-EF5) in period 2007-2019 using the same methods described in Elsner et al. Although the results show a downward trend, keep in mind that this decrease is not statistically significant. Unfortunately, we’ll need to wait 10-20 years or more to see if a trend emerges. It is hoped that the NWS can take steps to ensure consistency in path length and width data (and EF ratings) in the future.

Presenter
^firstElsner, JB, T. Fricker and Z. Schroder, 2019: Increasingly Powerful Tornadoes in the United States. Geophysics. Res. Lett., 46, 392–398.

²Fricker T, Elsner JB (2015) Kinetic energy of tornadoes in the United States. PLoS ONE 10 (7): e0131090.

Disclaimer

I don’t get any big oil money and have never met the Koch Brothers.