Today the ever dependable astronomer Phil Plait comes to the rescue:
A BBC report report this morning says that "experts" — no names, affiliations, or quotations were given — are concerned that the full Moon may trigger an eruption of the Philippine volcano Mount Mayon. It’s a very active volcano and ripe for an event. It could blow any time now. So is there cause for concern given the full Moon tonight?
I’m gona go with "no". The gravity of the Moon does affect the Earth, of course, mostly through tides. As it happens, tides are strongest when the Moon is full (and when it’s new as well, which the article doesn’t mention), so there is at least some reason to investigate this. And the BBC report says that the full Moon "coincided with at least three of Mayon’s 47 eruptions, including the two most recent ones in 2000 and 2001".
But let’s look at this critically, shall we? First of all, what does "coincide" mean?
First, Mayon is a very active volcano. It has quakes, minor explosions, lahars (mud flows) and such all the time. Certainly some will coincide with the full and new Moon. Let’s be generous and say that the time period around the full Moon is 2 days: a day before and a day after. The Moon goes through a complete cycle in roughly 29 days, so it’s full for 2/29 = 1/15th of the time. If you then look at 47 eruptions, then you expect to see 47/15 = 3 eruptions near the full Moon. And hey, that’s exactly what the report says!
So, statistically speaking, the Moon has nothing to do with eruptions. If it did, you’d expect to see a bump in the number of events near the full Moon. But the number of eruptions near the full Moon is what you’d expect from random chance. In other words, on average it doesn’t matter if the Moon is full, new, first quarter, or whatever. Now to be fair, the article doesn’t say how big a time period they used around the full Moon. Maybe they only used one day, not two. Even then, the correlation would be weak, because 47 eruptions isn’t a big enough sample to choose from. It’s small number statistics, like flipping a coin three times and having it come up heads each time. It’s rare, but it does happen on average one out of every eight times. You need bigger samples to get good statistics.
Now, there is some evidence that the Moon can cause earthquakes, and maybe even near volcanoes. But even then, if this were true in the case of Mount Mayon you’d expect more eruptions near the full Moon. It’s not seen, so again I think the correlation here is very weak.
So I am not totally discounting a connection between the Moon and this volcano, but I am saying that at best such a link is very weak, and probably not worth worrying about. The folks who live on the banks of Mayon have enough to worry about already!
I wonder: if it does erupt tonight, what will those "experts" say? But more interestingly, what if it waits three days?
Indeed. When these reporters tell us--as had been in the case in the news clip I saw last night--only the hits but not the misses (i.e., the times when Mayon erupted during the other phases of the full moon) those watching can easily be gulled into believing that there is in fact a strong connection (correlation) between full moons and Mayon eruptions. But as the BBC report specifies, only 3 out of 47 have been during full moons. And as Plait so clearly explains and computes, chance alone is enough to explain such coincidences. And even if Mayon were to erupt today, that still wouldn't prove full moons are in fact inextricably linked with Mayon eruptions.
Cum hoc ergo propter hoc.