This is a revisit due to something Albert spotted out on the Internet;
GeoLurking / April 22, 2012
Part I: The Moon
Time and time again, someone bops along with the idea that the Moon or the Sun causes an increase in seismicity. They climb up on their soapbox and thump their chest denouncing the world (that would be the rest of us) as being blind to the obvious correlation. That we will all suffer some sort of ruination if we don’t heed their warnings or suffer some calamity akin to a slow and brutal death.
Hey, sounds like fun. Let’s play.
Here is a plot of all earthquakes greater than Magnitude 4.5 as listed on the USGS website from 1973 to 2010.
Wow, that looks a bit compelling. How about the power distribution across that same data?
Well… that seals the deal. Right?
Not so fast.
First, I would like to point out that there is some research that points to a lunar influence in the activity of certain already seismically active regions, but that this research is founded on actual science. The effect is ephemeral and buried in noise. This is not intended to debunk that research, only to illustrate just how misleading some of the source data is, and how easy it is to jump to conclusions.
Now here is the nugget-o-truth that most people tend to miss:
The longer that the Moon spends at a specific location and phase, the more likely it is that quakes will occur while it is at that location. There are a couple of things that need to be taken into account. The Moon completes an orbit around the Earth about every 27.321 days. All orbits have a Perapsis (closest point on orbit) and an Apoapsis (furthest point). At perapsis the Moon is at it’s highest rate of speed at about 1.076 km/s. At apoapsis, it is moving at 0.964 km/s. Obviously, this speed is not constant. The period of the Lunar orbit is 27.321582 days. The Moon goes through a full phase cycle in about 29.53 days3. That’s almost the same period… but it’s not. The reason is that the Earth moves as well, in its orbit around the Sun. Say it is New Moon, with the Moon in-between the Earth and the Sun, so that we only see its night side. 27 days later, the Moon is back in the same position. But because the Earth has moved, the Sun appear displaced – by about 15 degrees. The Mon has to cover an extra distance before it is again in-between the Earth and the Sun: it takes a bit over 2 days. The first period gives us the (lunar) tides, and the varying distance to the Moon. The second period gives us the Moon phases. And each lunar phase does not last an equal amount of time. The day/night terminator moves across the Moon’s surface at about 10 miles per hour. When the Moon is at first or third quarter (half illuminated), the terminator is mid-way across the Moon (as seen by us) and you can see it move. When the Moon is new or full, the terminator is on the edge of the Moon, and it is moving directly towards us or away from us. And that you can’t see: the phase doesn’t change until the terminator has moved far enough that it is beginning to move sideways – as seen by us. So New Moon and Full Moon seem to last longer than a half-lit Moon.
Couple this with the dynamics of an elliptical orbit, and you get this odd characteristic.
This is the dwell time of the Moon on two separate phase cycles. Notice that the curves, though similar, do not match. This is due to the ‘not quite the same’ durations of the phase cycle and the orbital period, so that the effect of the elliptical orbit is moving around the lunar cycle. Also notice that the amount of time spent at the New and Full phases is longer than at mid phase.
Let’s take a look at several cycles in order to see if there is a pattern.
Sure enough… that orange is the plot of several phase cycles. The blue is an average of what is seen at that phase over those same cycles. (the average of the orange curve). We can go a step further and run this through a curve fitting program in order to see if there is a function that matches.
That’s pretty good… but note the end points, even though the curve is a good fit, it leaves enough uncertainty on the ends to make it mostly useless. I provided the plot mainly since I pissed away about two and a half hours in Eureqa’s “Formulize” in order to find it. <del datetime=”2019-04-20T04:18:31+00:00″>(it’s a really great program though)</del> I no longer think this. Their shoddy liscencing protocol is not something I wish upon anyone. Plus, I do not wish to spend the next few years arguing with the company that bought them out, so I got rid of the software. The “KISS” principle is a better method. “Keep It Simple Stupid” and you will find simpler less convoluted formulas are easier to make sense of.
Taking the idea of using the mean of the curve to calculate a correction factor, and using the 1000 bin average from the previous plot (the one with the orange and blue), we can apply that to the quake count curve.
Err… where did the trends go? Okay, maybe the power curve will still show the significant signal.
Hmm… not looking so good.
There is still an artifact in there… at least it seems to me like there is an artifact in there… but it’s small. So small that the last thing I would do would be to stand on a soap box preaching at people about it.
Part II: The Sun and the Moon
I realize that some people are adamant about the seismic connection with the Sun and the Moon. I also realize that I have pointed out a few issues with making this connection. One might argue that I was being very selective in presenting the data… okay, fair enough.
Here are some more plots that may, or may not, show a connection. You be the judge.
Nothing there that really jumps out at ya eh? Okay, a few more:
Do note that the apparent dwell time of the Sun at mid Winter and Mid Summer really stands out in that last plot. By the way, see those horizontal bands? Those are the latitudes of seismically active areas.
Again, the bands equate to known active areas… this time in longitude.
You may think me an ass for not believing in the Sun-Moon-Earth connection. That’s your prerogative. But unlike some, I actually went out and looked for myself. I’m not one to buy a pig in a poke. Personally, I don’t see it in the data. If your numeric skills are better, knock yourself out. I could stand to learn a thing or two while reading it. But if it’s BS, I’m not gonna buy it.
One thing that you should really take away from this, is the realization that as humans, we see patterns in anything. What we see is not always something significant. Bonafide researches spend countless hours doing something similar but more advanced, just for the purpose of proving to themselves that what they see is actually real… or not. Proving something wrong is just as important as finding something new.
As Richard Feynman would see it, proving a wrong lets you know that you still need to seek a better answer.