I first examined this data 15 or more years ago (also see Weather). Discussion with a friend encouraged me to update reporting on the data. One has to go to the source data and spend time extracting it and putting it into a usable form. Here is the data as of the end of 2024. The Atlantic hurricane season ends in November so this is final data. The December tornado data may not be finalized until January or later.
Atlantic Cyclone (Hurricane) Data
Although there is data on North American hurricanes going back into the nineteenth century and earlier, it is not particularly useful because the accuracy of measurement had no consistent basis and many storms at sea where never detected. Over time, data has become more consistent and extensive.
Below is a chart showing hurricane numbers from 1850 to 2015. Because of the mentioned reporting issues, one cannot conclude that in present days, hurricanes are more frequent.
Figure 1. Atlantic Cyclones, 1850 to 2015. Source: Market Watch.
The primary source of data now is the National Hurricane Center (NHC) of the National Oceanic and Atmospheric Administration (NOAA). NHC’s data extends back to 1872 (see Monthly Weather Review – Annual Summaries of North Atlantic Storms, 1872-2011). Current data can be found at Index of /archive/text/TWSAT.
I took data for 1999-2002 from the former table but used the latter table for the rest. The methodology is is to open the folder of the year you want to look at. Open the last document on the list and you have the report for that year. All disturbances are called tropical cyclones. Of these, those that become significant are given names as tropical storms (TS). Of these, those that are measured at hurricane strength (H) are counted as a sub-group and of these, those that became major hurricanes (MH) are counted as a sub-sub-group. I extracted these numbers, entered them into an Excel spreadsheet and produced this chart:
Figure 2. Tropical cyclones from 1999 to 2024. Source: NOAA, NHC and thePOOG.
I don’t have the skill to add a regression line so one has to examine the data by inspection.
Suppose we use 1999 and 2000 as a benchmark, each with 8 hurricanes (they didn’t identify how many were major). Then count how many of the following 10 years had more than 8 hurricanes in total. There are 4 such years, 2001, 2004, 2005, and 2010.
Next, do the same with the most recent 10 years of which 3 had more than 8 hurricanes, 2017, 2020, and 2024. That might suggest that hurricane activity is not as strong as in in the early 2000s. The total tropical cyclone activity looks stronger in the most recent 10 years meaning that we are getting more tropical storms recently but fewer hurricanes.
In recent times, there has been a measure created which may be preferred for measuring hurricane strength rather than just wind speed. This is called Accumulated Cyclone Energy (ACE). This has only recently been used although back calculation to some extent may be possible. I haven’t bothered to find this data.
Tornadoes and Hail Storms
For tornado and storm data for the US we go to NOAA’s Storm Prediction Center (SPC). The methodology is as follows. Using the boxes at the top of the home page, select “National”, “All”, “Final”, “Tables”, “Annual”, and “Monthly Totals”. I find response time is slow so you might have to wait until the table of monthly data populates.
Next, at the bottom, click “Download Monthly Data”. Open the downloaded data into an Excel spreadsheet to create annual totals. Going back to the main SPC page, choose another year. It’s table should populate directly. I did experience a problem with the initial table not populating but by toggling between “Preliminary” and “Final” I seemed to fix that.
Although tornado data goes back to 1950 on this site, I chose to use the period 1999 to 2024 to align with the hurricane data. Below are the charts of the data. The 2023 tornado numbers are preliminary only and all the 2024 numbers are preliminary, Because of this, not a lot of attention can be placed on these numbers. There may be issues such as with initial double counting as observational data is compiled from many sources.
Here is the chart of tornado data followed by a chart with numbers of hail storms.
Figure 3. Numbers of US tornadoes by year. Source: SPC and thePOOG.
Figure 4. Numbers of US hail storms by year. Source: SPC and thePOOG.
I looked at the data for three years, 2010, 2015, and 2017 (no reason for the choices) and fond that the final tornado numbers were about 7% lower than preliminary numbers. The surprise was the number of hail storms for which the preliminary numbers were lower than the final numbers by a rather large 35% to 45%.
This information allows us to infer that the tornado numbers as shown in the chart above for 2023 and 2024 may be high by about 7%. The hail number for 2023 is final but if the hail number for 2024 is, say 40% above the preliminary number shown, the final number would be off the chart. Such a deviation becomes suspicious. We will have to see what the final number is, sometime next year.
Discussion
When one speaks of the severity of storms, one must be very specific what the criteria are for assessing “severity”. Economic cost is not a useful measure although it appeals to the media that want to sell stories. Over time, development has increased in vulnerable areas such as sea shores, flood planes, and steep hillsides. Since the value of our money is steadily eroded by inflation and demand for building materials has accelerated beyond inflation, it is very difficult to compare two storms of similar intensity but separated in time using damage cost comparison.
Sub-sequentially, I found this chart on John Rubino’s substack:
People generally confuse weather with climate change since they lack the cognitive ability to make assessments based on personal experience over time. With the help of the media, the current weather event is always more severe than similar events in the past.
One of the largest phenomena that impact our weather are ocean currents which have decadal and multi-decadal periodicity. For example, read: Pacific multi-decadal oscillation modulates the effect of Arctic oscillation and El Niño southern oscillation on the East Asian winter monsoon. Consequently, short time periods like the25 year period I chose in this article have no long-term meaning. Choosing two 10-year periods at the beginning and end of the data above shows a small decrease in storm frequency for Atlantic cyclones, US tornadoes, and US hail storms. The severity of hurricanes as measured above (red bars) appears not to have changed significantly. Tornadoes also are classifies on a 0-5 scale of strength or severity but I did not have this data for this. I therefore cannot comment on changes of severity for these storms.
In past research I had studied Pacific cyclones. These span a much larger ocean area and one has to collect data from other countries like Japan and other Southeast Asia sources as well as data from NOAA. This is a much larger project.
***
If you liked this post, consider subscribing – in right-hand sidebar – if you are not a subscriber already. Please forward it to anyone else you might think will benefit from it. We are at a critical point in time where we need to disseminate truth to those who have ears to hear, and be forming communities of like-minded individuals. If you want to discuss an idea or issue, contact me – see Contacts.
0 Comments