Gravity's Rainbow

botany, shoes, books, and justice

Figures Lie and Liars Figure: Heat Waves on the Rise


Dallas South Dakota 1936
Image via Wikipedia

If you’re new here, you might want to read the previous Figures Lie and Liars Figure posts, a public conversation on climate change between my skeptical aunt and myself.


Heat, ma’am! it was so dreadful here, that I found there was nothing left for it but to take off my flesh and sit in my bones.  ~Sydney Smith, Lady Holland’s Memoir


An EPA slideshow on climate change indicators says that

The frequency of heat waves in the United States decreased in the 1960s and 1970s, but has risen steadily since then. The percentage of the United States experiencing heat waves has also increased. The most severe heat waves in U.S. history remain those that occurred during the “Dust Bowl” in the 1930s, although average temperatures have increased since then.

My aunt’s response to this was

Horrific heat wave (dust bowl) of 30s is mentioned as if it was a one time short period, but then says heat waves decreased in 60s and 70s, so question is were there heat waves 30s, 40s, and 50s.  Now heat waves on rise again after 70s; perhaps 20-30 year cycles?

My aunt wants to know if heat waves happen in predictable cycles – if they do, maybe we’re just in an upswing of that cycle and we can’t use increasing heat wave frequencies to bolster the case for climate change.

Image via Wikipedia

Let’s start with the definition of a heat wave.  Everyone is familiar with heat waves – the dog days of summer – when for days or weeks at a time it’s so hot all you can do is lie naked in front of a fan and dream of shaded creeks, air conditioning, and ice cream.  They aren’t just uncomfortable – they are one of the most deadly kinds of natural disasters. A widespread heatwave just this past summer caused problems from Boston to Shenyang and was particularly deadly (and expensive) in Russia. Different governments and agencies define heatwaves slightly differently. I’ll primarily use the one adopted by the EPA when they built their heat wave index:

a four-day period with an average temperature that would only be expected to occur once every 10 years, based on the historical record.

That definition and all of the data and graphs below (unless otherwise noted) are from the document linked from the Learn more link on the EPA heat wave indicator slide my aunt takes issue with.

If we want to know if heat waves occur in cycles, we need to look at heat waves over time.  This graph shows heat waves in the continental 48 states over the last century:

Heat Wave Index for lower 48

"The index value for a given year could mean several different things. For example, an index value of 0.2 in any given year could mean that 20 percent of the recording stations experienced one heat wave; 10 percent of stations experienced two heat waves; or some other combina- tion of stations and episodes resulted in this value."

My aunt asks were there heat waves 30s, 40s, and 50s.  Clearly, there were. The 1930s have a very high heat wave frequency. That spike in the 30s obscures the increasing trend in heat wave frequency since the 70s. It does answer my aunt’s question about cycles, though: in the United States, over the last century, there is no clear cycle of heat waves.

Climate change is a global phenomenon and the US is a small fraction of the planet. So, if we consider heat waves worldwide, what do we see?

Warm nights and warm days

"Observed trends (days per decade) for 1951 to 2003 in the frequency of extreme temperatures, defined based on 1961 to 1990 values, as maps for the ... 90th percentile: (c) warm nights and (d) warm days. Trends were calculated only for grid boxes that had at least 40 years of data during this period and had data until at least 1999. Black lines enclose regions where trends are significant at the 5% level. Below each map are the global annual time series of anomalies (with respect to 1961 to 1990). The red line shows decadal variations. Trends are significant at the 5% level for all the global indices shown. Adapted from Alexander et al. (2006)."

We can’t directly compare the world graphs to the US annual heat wave index graph because they cover different time periods and because one shows a heat wave index and the other shows extreme temperatures directly.  The both show, however, that there is no apparent periodicity in heat waves at the time scale we’re looking at and that heat waves have become more common, especially since the 1970s.

Frequency of extremely warm days and nights in the USThe worldwide graph does something interesting – it separates out warm night and warm days. Extremely warm nights are increasing faster than extremely warm days. The additional information linked from the EPA slideshow shows that the same thing is happening in the US.  And while the 1930s heat wave index for the US shows that it was very, very bad – certainly worse than the last thirty years – the frequency of extremely warm nights is now higher than it was even during the dustbowl.

Why are hot nights such a big deal?  Think back to the last unbearably hot summer you experienced and how much you looked forward to evening or a thunderstorm for just a little break from the heat.  I don’t have air conditioning even though daytime temperatures in the summer are usually in 80s.  I’m rarely uncomfortable, though, because nights are in the 40s and 50s, even during the hottest part of the summer.  Every evening, we open the windows and the house cools down. It doesn’t start to get uncomfortably warm until midday.  If it didn’t cool down so much at night, I’d spend most of the summer sweaty, sleep deprived, and majorly cranky.  Cooler night time temperatures provide major relief for people and other animals during heat waves. More hot nights make heat waves extra deadly.

You may wonder, if we’ve been pumping more and more greenhouse gases into the atmosphere through the entire time period these graphs show, why did we see a decline in heat waves in the 1960s and 1970s? The 60s and 70s were actually quite cool, but this doesn’t mean that climate change – in the warming direction – isn’t occurring.  You see, greenhouse gases aren’t the only thing we’re dumping in the atmosphere.  We’re also putting lots and lots of sulfate aerosols in the atmosphere, which can have a cooling effect on the climate. While they cooled things down in the 60s and 70s, warming has overwhelmed their effects since then.

ASIDE: Putting more sulfate aerosols in the atmosphere is a geoengineering proposal that’s gotten quite a lot of press. It could, in theory, buy us more time to get our act together in dealing with climate change. However, quite a few people (myself included) object to using the entire earth as an experiment, and the long list of potential and probably side effects (drought, ozone depletion, uglier skies, warming of other layers of the atmosphere, changing cloud formation, affecting sunlight diffusion to influence plant growth, interfering with solar energy production, pollution as sulfur is deposited, and uneven effects) is daunting.

So, to recap: Heat waves are dangerous natural phenomena that are becoming more frequent worldwide because of human activities, as predicted by climate change models.

Related Posts Plugin for WordPress, Blogger...


  1. Fascinating. I’m wondering – those extreme values in the 30s – could the heat have been exacerbated by the loss of vegetation in the dust bowl? The article in wikipedia says that 100,000,000 acres (400,000 km2) were affected. A pretty big area. Could there have been a positive feedback effect?

    Hot nights are a big deal, as you so rightly pointed out. Those of us who live without air conditioning depend on those cool nights for relief.


    In other news, the relative warming of nights and days are also a reason that we know that temperature trends are due to greenhouse gasses, rather than other factors like the sun. 😀

Leave a Reply to Mary Cancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.