News Section
Stories from Climate Central's Science Journalists and Content Partners

Yes, Summers in the U.S. Really Are Getting Warmer

One of the surest signs that the planet is getting warmer is the fact that record high temperatures are outpacing record lows. As of early August, for example we were able to report that with the year a little more than half over, 2012 had already surpassed all of 2011 in terms of record highs in the U.S.

It’s unfair to compare just two years, of course, since that could just be a weird anomaly — but if you look on a much grander scale, the disparity of record highs in the U.S. vs. record lows has grown wider every decade since the 1970s.

NASA has come up with an animation, prepared in conjunction with the release of a paper by the Goddard Institute for Space Studies' James Hansen, that makes this point far more vividly than any static bar chart could. It simply plots summer temperatures in the Northern Hemisphere, starting in the 1950s, showing the number of days that were unusually hot (red), unusually cold (blue) or pretty much normal (gray). Back in the ‘50s, the plot is a nearly perfect normal distribution (more commonly known as a bell curve), with just as many days falling on the cold side as the warm side. It’s what you’d expect if temperatures were going up and down randomly.

But as the animation moves forward in time, the bell begins to slide to the right: unusually cold days diminish drastically, while unusually hot days become more and more common. By the time the animation ends, in 2011, there’s a whole lot more red than blue.

So if you think summers are getting hotter, you’re absolutely right. It’s just one more reminder that the globe really is warming, thanks in large part to the heat-trapping greenhouse gases we keep pumping into the atmosphere.

Comments

By dan in illinois
on September 4th, 2012

A plot like this is fairly useless without more information about the statistics behind it.  How much is one standard deviation in terms of degrees C?  If the distribution of temperatures is tight, one STD will be only a small fraction of 1 degree, while a wide distribution could be much more.

Reply to this comment

By Jason
on September 4th, 2012

Compelling video.  But where can the original NASA animation be found (not Climate Central’s version)? Is there a link? Any more info on the source data?

Reply to this comment

By Lindsay Harmon
on September 4th, 2012

@Jason,

Here is the link to the original NASA animation: http://svs.gsfc.nasa.gov/vis/a000000/a003900/a003975/

Sincerely,

Lindsay Harmon
Web Producer, Climate Central

Reply to this comment

By slett (hightstown)
on September 4th, 2012

You may want to look at the data in

“The relative increase of record high maximum temperatures compared to record low minimum temperatures in the U.S.”  Gerald A. Meehl, Claudia Tebaldi, Guy Walton, David Easterling, and Larry McDaniel , Geophysical Research Letters

Reply to this comment

By Quentin (02481)
on September 4th, 2012

Not useless at all - the standard dev degrees will vary according to the location, I assume.  This does not show that data and does show something very important.  There are no models
No theories no the above animation - it is just the facts.  We are warming.  Exactly as the scientists said we would.  For global warming to be a hoax you’d have to explain 1) why co2 was NOT responsible for the warming we see 2) what else WAS responsible for the warming we see
There are no answers to either question, let alone both.

Reply to this comment

By dan in illinois
on September 5th, 2012

Uh, Quentin, I don’t think you have a firm grasp of statistics.  There is a single number in degrees that corresponds to 1 STD.

Reply to this comment

By Larry (Emeryville,CA 94608)
on September 5th, 2012

Dan,
with a normal distribution, about 2/3 of the observations are within 1 Std Dev of the mean. looking at the graph, it would appear that the std dev is about 1 degree. about 90% of the observations are within 2 std deviations, and 99% within 3 std deviations.
what’s your point? are you going to claim that the trend is insignificant based on a statistical argument?

Reply to this comment

By dan in illinois
on September 5th, 2012

Larry,

How did you determine that 1 STD on the graph corresponds to 1 degree?  Do you think that’s 1 degree C or 1 degree F?  In fact, the horizontal axis has units of STD, not degrees.  1 STD could be 0.1 degree or it could be 10 degrees.  From the graph, nobody can say.  Your assertion that 1 degree = 1 STD doesn’t make any sense.

Reply to this comment

By Larry (Absecon, NJ 08201)
on September 6th, 2012

I think you guys didn’t read the text.  It says number of days above normal/below normal.  It graphs number of departures from normal/up/down.  Not degrees of departure from normal.

Reply to this comment

By dan in illinois
on September 6th, 2012

Uh, Larry, I did read the text.  If they’re just plotting “days above normal”, “days near normal”, and “days below normal”, then they would have a probability mass function (PMF) - that is, a distribution with just three discrete points on the graph.  It would be a bar chart.  Their chart clearly shows a continuous distribution of something.  To put it another way, if they’re just plotting those three discrete cases, how do they wind up with points on the graph that are at 2 STD from the center, 3 STD, etc.?  They clearly have a continuous variable that they’re plotting on the horizontal axis.

Reply to this comment

By Larry (Absecon, NJ 08201)
on September 6th, 2012

It’s been a long time since Stat 101 so I could be wrong but I do not think is is a continuous distribution.  I think it’s a discrete distribution of the number of days above, at, or below normal per year.

Reply to this comment

By Larry (Absecon, NJ 08201)
on September 6th, 2012

Oooops just looked again.  Looks like they used 10 year running average.

Reply to this comment

By Larry (ab)
on September 6th, 2012

Now I’m confused.  Is the graph of degrees of departure from normal, or number of days per some time period????

Reply to this comment

By Dan (Manhattan, ks, 66502)
on September 6th, 2012

Standard deviations are exactly what you want to use if you want to summarize changes across multiple sites. This is a pretty common practice, and essential if one doesn’t want to focus on one climate site and be accused of cherry picking. I am not a statistics professor, but I have 16 graduate-level credits in statistics. I can say, having done some analyzes of climate data (actually in review in peer-reviewed literature now), that 4+ sd anomalies in my neck of the woods (the central U.S.) are often close to 7C from the mean, and that’s a pretty big deal. Away from the coasts where the climate is more continental and variable, anomalies like this surely represent really huge deviations means…but that’s not even the point, as others have said, the point is that we are now seeing conditions, whether far from means or not, that were once exceptionally rare, if not unprecedented. Mean temperature for spring in Manhattan, KS this spring, for example, was 1.5C warmer than the second warmest year in a climate record going back to 1891!

Reply to this comment

By cliff (atlanta)
on July 15th, 2013

2013 is the summer of mean reversion. Coldest summer since the 70s at least? When was there a summer this cool?  Anyway, the droughts of the late 1990s and mid 2000s skewed temperatures hot, but now that rain is back, things are 10 degrees cooler. Maybe all this global warming was just reacting to droughts?

Reply to this comment

Name (required):
Email (required):
City/State/Zip:
Enter the word "climate" in the box below:

[+] View our comment guidelines.

Please note: Comment moderation is enabled. Your comment will not appear until reviewed by Climate Central staff. Thank you for your patience.