Damage in Santiago, Chile. Photo by Reuters/Marco Fredes
After the massive earthquake this past weekend in Chile, MSNBC published a sensationalistic piece entitled, “Is nature out of control?” The Wall Street Journal asked if three massive earthquakes around the world in two months are related and a cause for alarm. The mainstream media, always searching for sensationalistic or fear mongering news, has latched onto the question; are we seeing more earthquakes than normal?
Well, not really.
To better understand why, let’s take a look at how many earthquakes occur each year on average. The USGS has a fascinating page of earthquake facts and statistics, with the following table:
| Magnitude | Average Annually |
|---|---|
| 8 and higher | 1 ¹ |
| 7 – 7.9 | 17 ² |
| 6 – 6.9 | 134 ² |
| 5 – 5.9 | 1319 ² |
| 4 – 4.9 | 13,000 (estimated) |
| 3 – 3.9 | 130,000 (estimated) |
| 2 – 2.9 | 1,300,000 (estimated) |
¹ Based on observations since 1900.
² Based on observations since 1990.
For our analysis, let’s take earthquakes based in the magnitude 6.0 – 6.9 range. Why am I picking earthquakes in the M6 range? It’s arbitrary. You can repeat this process for earthquakes of any range. Based on data recorded since 1990, we’d expect to see an earthquake within this magnitude range occur every 2.7 days or so.
So here we are, on March 1st, 2010, the 60th day of the year. How many earthquakes in the M6.0 – M6.9 range have we had this year? According to this handy search tool from the USGS, there have been 25 earthquakes of M6.0 – M6.9 in 2010.
That works out to roughly one earthquake in the magnitude 6 range every 2.4 days. That doesn’t seem totally unreasonable or a reason for alarm, but we should do some further work to put it in context.
We can plot up the number of earthquakes per year and come up with a standard deviation, assuming a normal distribution of earthquakes in any given magnitude range.
Total results: 21
Mean (average): 2.67143
Standard deviation: 0.41732
So, the number of magnitude 6 earthquakes that we’ve had in 2010 falls within one standard deviation of the mean. If we were to plot up a graph, it’d look like this. The error bars represent one standard deviation.
Awesome! Well, what about those ranges of values that fall outside of one standard deviation from the mean? For those that don’t understand how statistics works, check out the following bell curve from Wikipedia.
This shows roughly the percentage of values that you’d expect to fall within a specific standard deviation away from the mean value.
Dark blue is less than one standard deviation from the mean. For the normal distribution, this accounts for about 68% of the set (dark blue), while two standard deviations from the mean (medium and dark blue) account for about 95%, and three standard deviations (light, medium, and dark blue) account for about 99.7%.
So, if we modify our graph to show an error bar of 2 standard deviations, you’ll notice that every result since 1990 fits inside this model! Statistically speaking, you would expect to find 95% of all results falling within two standard deviations of your average. Simply put, there is absolutely nothing strange happening.
In fact, thanks to this normal curve you can basically predict, with a 99.7% chance of success (three standard deviations), that an earthquake of equal to or greater than M6.0 will occur somewhere around the world within the next 3.5 days. Update: Proven correct! A M6.4 earthquake occurred in Taiwan on March 3rd.
Alright, so what’s with all the coverage on earthquakes? It sure seems like a lot is happening, right? We can attribute this to observer bias. The massive devastation in Haiti warranted a large amount of news coverage. Because this is so fresh in everyone’s mind, people are more likely to notice any news or information related to earthquakes anywhere in the world. An earthquake of M6.0 or greater, usually garners international attention.
It’s the same principle that happens whenever you acquire some new toy, gadget, or piece of clothing. Suddenly, you notice that particular item around all the time. It’s like everyone has it.
So, bottom line, the Earth isn’t becoming more active, more dangerous, or even “out of control.” Despite the fear mongering and what esteemed mainstream media networks would have you believe, the simple reality is that the numbers prove things are happening at an expected rate. Keep that in mind the next time a large earthquake happens and everyone is wondering why the Earth seems so active!
Update (April 21, 2010): Chris Rowan at Highly Allochthonous has a great post on yearly earthquake averages and variability with larger magnitudes.
In the last 28 years, there have been on average around 13 such ‘significant’ earthquakes a year, with a magnitude 8 occuring about every year and a half. This average rate is marked by the grey line on the plot: if we extraplolate the six major earthquakes recorded in the first four months, 2010 is on course to experience 18 major earthquakes, a little above average but well within the variability shown by the whole dataset (and it’s actually closer to the centennial average of 16 major quakes a year reported by the USGS above).
27 Comments
Thanks for that, i was wondering
Very nice, thanks for doing your homework.
Thanks Ken. It’s something I’m definitely interested in – I was a geologist is a past career. Earthquakes still fascinate me and I absolutely cringe when I see these hyperbolic news reports!
Don’t you wish more reports were based on facts like this one? Don’t you wish more people understood what all them fancy numbers and graphs meant? While I’m at it; can I have a pony?
Great analysis!
Thanks Matt!
Totally agree. I wish more people would realize this. Statistics is pretty easy to grasp (and one of my favorite types of mathematics). Lifehacker had a pretty good article yesterday about this:
http://lifehacker.com/5485624/look-at-data-like-a-statistician-minus-the-ph-d
I really appreciate this, all laid out in facts and graphs. It does feel like there has been more going on than usual, but I questioned it because I also have noticed how much the media loves to build everything up to a frenzy.
Thank you
Hah, someone on the Geology News blog was freaking out about how there’s still been a lot of large earthquakes, so I re-ran the numbers as of today.
Today is March 12th, the 71st day of the year.
According to the USGS, there have no been 41 M6.0 – M6.9 earthquakes so far in 2010 (Whoa! One would assume that a lot of those are aftershocks from the great Chilean earthquake).
That means we’ve had an earthquake in the M6 range every 1.73 days. Alright, that number is getting down there. Is it a cause for alarm?
No. Why?
The standard deviation is 0.417 days with a mean of 2.671 days between earthquakes of this size (if you’re curious how I got those numbers, reread the article). Since we can expect to find 99.7% of all results within 3 standard deviations of the mean, let’s figure that out. 0.417 x 3 days = +\- 1.251 days.
2.671 days – 1.251 days = 1.42 days.
If you start seeing a M6 earthquake more than every 1.42 days, then come talk to me.
There’s nothing to worry about.
Here’s the break down for other magnitudes in 2010. If you want to search yourself, you can use this tool:
http://earthquake.usgs.gov/earthquakes/eqarchives/epic/epic_global.php
2010 earthquakes:
M7.0 – M7.9: 3 earthquakes so far in 2010 (yearly average: 17)
M8.0 – M8.9: 1 earthquake so far in 2010 (yearly average: ~1)
If you want to account for M7 earthquakes:
Average: 365 days / 17 eq’s per year = 21.47 days per EQ.
2010: 71 days / 3 eq’s in 2010 = 23.67 days per EQ.
Magnitude 8 earthquakes? Average is about 1 a year. We had one.
I wrote on this subject a few weeks ago. The elephant in the room that you are not addressing, the reason people care whether earthquakes are increasing is that they know, albeit vaguely, that there’s something in the Bible about earthquakes in the last days. But as I said in my column, Jesus didn’t say earthquakes would increase in the last days. He simply said that earthquakes would be one part of the composite sign of the last days.
http://www.examiner.com/x-17373-Phoenix-Signs-of-the-Times-Examiner~y2010m3d16-Are-earthquakes-increasing
Just being lazy here, plus I never was good at math or statistics:
So how are we doing now? The 7.2 Baja on April 4, and then 7.8 off Sumatra today, April6.
Bottom line: I will trust you will run a “Time to Freak Out!” post if the shaking level goes off the charts and deviates in a non-standard way.
Who likes deviates for the most part anyway.
This is awesome! This is just the thing I was looking for and you managed to make it easy to digest as well, much appreciated.
Just wanted to let you know that I very much appreciated your post it was very helpful!!
Thank you for mapping this out! I was also wondering about the stats. I actually don’t watch the news but everyone is talking about them so it does seem to be more than usual.
Thanks for this thoughtful analysis. It not only helped explain earthquake prevalence but I think I finally understand standard deviation as well now.
Hi, thanks for this
Found it because there’s just been another one in China today and it felt like there’s been a lot so far this year. Thanks for the statistics.
There is another aspect your analysis didn’t address, which is the location of the epicenter. While the quantity and magnitude of earthquakes hasn’t changed, what about the impact on civilization centers?
That’s an interesting point, but I don’t think it’s indicative of anything. That is to say, I don’t think one can necessarily prove that earthquakes are targeting population centers. It’s the luck of the draw where an earthquake happens and a local population’s ability to prepare for and cope with an earthquake.
Good article, well written and considered.
I’d like to draw your attention to a USGS infographic, which has since been pulled offline, perhaps for the sensationalist observations that could be drawn from this visual.
<a href ="http://www.bradfredricks.com/earthquakes%20usgs%20graph08.jpg" title="USGS Graph of Deadly Earth Quakes"
In the image, you’ll note a MASSIVE uptick in deadly and destructive earthquakes. USGS wisely pulled this offline, clearly its a disturbing trend.
I’d like to hear your thoughts on this graph, vs the data you’ve pulled. Could you revisit your data, looking at the same information and time frames taken into account in this USGS infographic? Essentially, third party verification.
Thanks,
Brad
Cool Dave. Makes a lot of sense. I do have one Thought though.
We, man, are now making earthquakes. At least that is an article in April edition of P.S. is saying. Then we have that nasty fracturing for gas going on, then test drilling into the mantel.
Ewwwww, are we creating our own disaster movie?
Yeah, this is something we’ve known about for decades. Specifically, geothermal power production is known to produce small earthquakes due to injected water lubricating the faults. Most of these quakes are between M2 and M3.
Some have argued that loading caused by the creation of a reservoir (e.g., China’s Three Gorges Dam) might induce earthquakes as well, since the weight of water can significantly alter the distribution of stress on local faults.
Rerun your analysis. Now it most certainly is outside the ordinary. At 2.17 standard deviations. Probably just caught it on a statistical fluctuation though.
try the analysis again….try it with 5.0 – 5.9 too…current rate for 5.0 – 5.9 is to end 2010 with 2400+ vs a 1320 average. Hmm……everything is perfectLIE normal.
6.0-6.9 on the increase too…also recent China quake 7.1 (Per China) or 6.9?
Actually, there has been an increase in major seismic activity from 1973 to 2009 (the years for which there is online information at USGS.gov)
I was curious, so I took only those quakes of 7.0 magnitude or greater for each year and compared the seismic action on a linear scale (rather than the logarithmic Richter scale). Obviously, there are years of high seismic activity and years of low activity; but, when I ran an OLS regression the slope was +0.97 per year over 37 years. This is the equivalent of one additional magnitude 7.0 earthquake EACH YEAR! Keeping in mind that the quake that recently leveled Haiti was a 7.0 magnitude quake, there are 37 more 7.0 earthquakes per year in 2009 than there were in 1973. That’s significantly more shaking now than in 1973.
Anyone can do the same analysis that I did and get the same result.
No, actually there hasn’t been a significant increase in major seismic activity.
Please read: http://scienceblogs.com/highlyallochthonous/2010/04/the_seismic_non-pocalypse.php
I can provide the Excel spreadsheet I used; but I don’t know how to upload an Excel Graph as an image.
I agree that there has been no “extra” siesmic activity, but I do believe the effects for this years earthquakes have been different than in the past.
My theory is the earth`s mass is trying to move closer to the center of the earth to increase the rotation of the earth , because of the thousands of years of mankind digging from the ground, building ontop of the ground. Thousands of years of kicking dust in the air, we have effectively increased the diameter of the earth making the earth`s rotation to slow down. When our solar system crosses the equator of the milky way the earth WILL be in the same position it was the last time and every time we pass the galactic equator. It`s exactly like the second hand on a clock when it strikes midnight all clock hands are in the same position each time.
Proof of this was evident in the Chile earthquake, the earth day got shorter by 1 millionth of a sec, that may seem miniscule but if you work that out against the circumference of the earth, it isn`t so miniscule, especially when the indonesian earthquake shortened our day by more than 1 millionth of a sec, i dont know the exact figure.
Our galaxy is trying to make us speed up so the earth is in exactly the right spot at the right time, and the earth is reacting to it by moving its mass closer to the centre.
There is a reason the skin colour of our earth inhabitants,dark on on side, white on the other.
How close to Africa are your gonna be during the main event? Depends what skin colour you want to come out with.
…I don’t even know where to start.
If you use the USGS tool and indicate a magnitude 2 to 10. Can you explain why from January 1 to July 17…that in 2005 there were 16,944 quakes; 2006 15,509 quakes; 2007 15,237 quakes; 2008 17,211 quakes ….yet in 2009 they are only indicating 8,090 and for 2010 they only indicate 9,676 quakes. How can the activity basically get cut in 1/2 for the past two years?