Extreme Weather: Tornadoes

2011 was one of the most horrific tornado years on record:

• 577 dead
• $22.5 billion in property damage
• 1,704 tornadoes

Did this result from climate change driven by human-caused CO2 emissions into the atmosphere? (known as AGW or anthropomorphic global warming)

Does it prove that the number of extreme tornado weather events is increasing?

Confusion abounds. The drum beat of the news media and AGW alarmists is that tornadoes are on the rise. It is said that loss of life, property damage and numbers of extreme tornadoes is increasing because of human-caused climate change.

The facts, as they stand today, do not support the AGW claims.

Violent tornado activity has decreased since 1950! 2011 is the exception, not the rule.

Total tornadoes by year 1950-2012 (Fig. 1 – NOAA 2012 State of the Climate Report)

NOAA’s Annual Tornado Report

In January 2013, NOAA’s National Climate Data Center (NCDC) put out it’s annual review of tornado activity for 2012. The above graph was prominently displayed at the top of the report. It shows tornado counts by year from 1950-2012.

The trend is undeniable. There are MORE recorded tornadoes now than in the 1950s and 1960s. It’s iron-clad proof of increasing numbers of tornado events, right? Wrong!

The key word here is “recorded”.  Account for that and everything changes.

Tornado Tallies

As one might suspect, better reporting accounts for some of the increase reported by NOAA. But… how much more?

A peer-reviewed study by the National Severe Storms Laboratory published in Weather and Forecasting answers that question:
“Storm Spotting and Public Awareness Since the First Tornado Forecasts of 1948“
– Charles A. Doswell III, et al., Weather and Forecasting, Vol 14, No. 4, pp. 544-557

This paper discusses the effect of a storm forecasting service started by the U.S. Weather Bureau in 1952. It’s called the Severe Local Storms Forecasting Unit (SELS).

According to the study, SELS is responsible for most of the reported increase in the number of tornadoes reported each year. Better reporting has also resulted in fewer lives lost.

The purpose of SELS is to reduce loss of life through greater public awareness and local networks of storm chasers who spot and report tornado size and direction of travel. SELS has been augmented with improved weather radar over time.

Total tornadoes and tornado days per year from 1915-1995 (Fig. 2)

This graph from Doswell III et al. shows the impact of the SELS program and related activities. A lot more tornadoes got spotted and reported after SELS started. Back in the 1940s there were between 100-200 tornadoes reported each year. By 1965 that number tripled to 600 and by 1995 that had zoomed up to 1,200 a year.

Supporting evidence that the greater numbers of tornadoes reported directly resulted from reporting improvements is that the tally of tornado days has remained static since 1955 while the total number of tornadoes reported doubled.

Tornado deaths per million population by year from 1880-1995 (Fig. 3)

Adjusting for population growth, Doswell III et al. show that the number of deaths caused by tornadoes each year has steadily decreased since 1930. They attribute the drop to greater public awareness of tornadoes and SELS.

Deaths per unit of damage decreased after SELS started (Fig. 4)

This important graph is harder to understand but shows something important. Since 1952, for a given amount of tornado-caused damage, the number of lives lost due to that damage has went down!

It would have been easier to understand if the authors had chosen to:

1. Reverse it to make it a Fatality/Damage Ratio instead
2. Display the horizontal timeline in years instead of  +/- the number of days since 1/1/1904

If they had done that then the ratio would decrease with a decrease in loss of life. That makes more sense. With a better horizontal scale, casual readers could then see the dates range from 1880 to 1995 instead of -23900 to +34500.

Doswell III et al. reveal:

• Deaths caused by tornadoes are decreasing over time, not increasing
• Total tornado counts have increased because of better life-saving reporting with SELS

Related Tornado Metrics

Two other meaningful types of data related to extreme tornadoes are:

1. Normalized tornado damage statistics
2. Timeline occurrence of extreme, very dangerous F4-F5 tornadoes

Normalized Damage Stats

A peer-reviewed paper addresses the topic of damage:
“Normalized tornado damage in the United States: 1950–2011“
– Kevin M. Simmons, et al., Environmental Hazards, 10/5/2012

Because of a growing human population and increasing amount of building activity, scientists must “normalize” those numbers first to make valid comparisons.

“Normalize” just means that they have to account for population and building changes over time. That is what Simmons et al. did.

Here is their result:

Tornado damage normalized to population, housing units and GDP from 1950-2011 (Fig. 5)

There are two noteworthy things found in Fig. 5 (Figure 2):

1. Relative tornado damage has generally declined since 1952
2. There is a great big uptick in the last year (2011 marked in yellow)

This graph clearly shows how unique 2011 was. There was a 58-year general decline in tornado activity before 2011 came out of nowhere. It has no statistical similarity to the steady rise in CO2 since 1950.

The graph ends in 2011. However, look back at the NOAA graph in Fig. 1 above. It shows that 2012 was, in terms of total numbers, the least active tornado year since 1989.

NOAA shows the decline in tornado activity returned again in 2012, right where it left off.

Extreme Tornadoes

For extreme weather events, what matters is only the number of extreme tornadoes over time. Extreme tornadoes are F4 or F5 on the Fajita damage scale. The weakest tornadoes of all, F0, account for the vast majority of all reported tornadoes.

The total number of extreme tornadoes (F4/F5) has been tallied at usatornadoes.com.

Number of extreme tornadoes (F4/F5) by year from 1950-2011 (Fig. 6)

This graph shows the 25 years before 1976 averaged more extreme tornadoes per year than the 37 years after. Again, 2011 sticks out as a very bad year at the end of a long, slight decline in extreme tornadoes since 1974. 2005, 2006 and 2009 had fewer extreme tornadoes than any other years since 1950. There were 3 years before 1975 that had more extreme tornadoes than 2011.

There were only four F4 tornadoes and no F5s in 2012, if Wikipedia can be believed.

This is consistent with the findings of both Simmons et al. and Doswell III et al.

Conclusions

This review shows that 2011 was an exceptionally terrible year for tornadoes in every respect.

It also shows that SELS reporting improvements have skewed the overall yearly tally of total tornadoes reported by NOAA in their State of the Climate Report for 2012.

The claims of AGW alarmists are not validated by Coswell III et al., Simmons et al. and/or usatornadoes.com

When data is normalized and extreme events pulled from the tallies, scientists tell us:

• Lives lost have been declining since 1930
• Property damage has been declining since 1952
• There were more extreme tornadoes (F4/F5) before 1976 than after
• The average number of extreme tornadoes since 1975 is in gradual decline
• 2005, 2006, and 2009 had the fewest number of extreme tornadoes of any years since 1950

2011 was unique. Its discontinuous leap to 20+ extreme tornadoes that year was followed by a precipitous drop right back down to only 4 in 2012.

What happened in 2011 is so vastly different from the rest of the tornado record that human-caused global warming, as we currently understand it, cannot be responsible for it.

Global warming is better aligned with decreases of extreme tornadoes than it is increases.