Mt. Fuji in the autumn.
skyseeker / Wikimedia Commons Sometimes, it's the sales pitch that gets you rather than the actual car. That seems to be the case with the latest rash of media coverage over the "critical state" at Japan's Fuji. You read the news coverage and you'd think that Fuji will erupt any second now, all thanks to the 2011 Tohoku earthquake that struck off the coast of Japan. Now, I wouldn't blame you if you got that message -- it is exactly what press releases and quotes from the authors make it seem is the case. Dr. Frolent Brenguier, the lead author on a new study that appeared in Science, was quoted as saying "All we can say is that Mount Fuji is now in a state of pressure, which means it displays a high potential for eruption. The risk is clearly higher." Seems pretty straight forward, doesn't it? Their research must clearly show that Fuji is now in a state ready to erupt and we know that from some sort of pressure measurement. Now, it is hard for me to blame the media for not going back and carefully reading the Science article to see if their data supports such grandiose claims. You have to believe that if a paper is published in Science, then it is supported by verifiable data -- and for the most part, they are. Like any reputable journal, Science is rigorously peer-reviewed before any article is published. Now, big name journals like Nature and Science do attract media attention. Not only do they want what they see as quality scientific research, but they also want it to be flashy. So, you might have done the best study ever on the eruptive history of Mt. X, but Nature and Science wouldn't touch it unless you can make it flashy: Is Mt. X a "supervolcano"? Did it change global climate? Will it destroy us all in the future? In a sense, Nature and Science are the Hollywood of science publications -- they want the big tentpole papers and the everyone wants to star in one of those. That is where the danger lies: if you do get published in Science or Nature, you want to get media coverage (because that surely helps your career). How far can you push the interpretations, possibly even from outside the paper itself, to get the attention you desire? (Update: please see my note at the bottom) Back to the Brenguier and others (2014) study on Mt. Fuji. They examined how the state of pressure in the crust across Japan changed after the massive M9 Tohoku earthquake in 2011. That earthquake released a massive amount of energy, and although it relieved stress near its epicenter, it likely caused stress in the crust to increase in other places as that energy was displaced. By examining how quickly seismic waves move through the crust (which is partially controlled by the state of stress in the crust), they could see where new stress has accumulated. They argue that places with the largest velocity reduction after the Tohoku earthquake are the places where the crust is feeling low effective pressure. This low effective pressure is caused by pressurized fluids, like magma or hydrothermal fluids (i.e., water), in the crust pushing outward on the rocks.
Figure 2 from Brenguier and others (2014) showing the change in seismic velocity across Japan after the M9 2011 Tohoku earthquake.
Brenguier and others (2014), Science. Not surprisingly, the places that saw the largest velocity reduction were places underneath all the active volcanoes across Japan (see right). In contrast, the smallest reduction occurred in places with rigid rocks, like granite. This change in seismic velocity is tiny -- even in the areas with the largest change, it was only by ~0.12%. Now, this is where it gets tricky. They state: "The seismic velocity susceptibility to stress can be used as a proxy to the level of pressurization of the hydrothermal and/or magmatic fluids in volcanic areas." This means that anywhere that either hydrothermal or magmatic fluids are present can experience the large drop in seismic velocities. So, you can measure changes in seismic velocity to understand changes in pressurization of the crust -- such as when new magma is intruding or hydrothermal fluids are moving through the crust. In my mind, that is their key conclusion. It does not mean that the Tohoku earthquake caused the pressurization of the area as such. Rather, that changes in seismic velocity after the earthquake can tell us something about the state of pressurization in the crust. They do go on to say that an earthquake occurred 4 days after the Tohoku temblor, and it happened to be near Fuji (which hasn't erupted since 1707, making people worry it's "overdue" -- it isn't), but this correlation is not a piece of their evidence for their conclusion, but rather their way of trying to say the crust was prone to new earthquakes already and Tohoku triggered it. This is a bit of a stretch without further research to support this triggering. The one thing they never say in the paper is that Fuji is more likely to erupt thanks to the Tohoku earthquake. Never. Not once. So, why is this the message that we're being fed in the news? Well, it's thanks to the authors deciding that a conclusion that is outside their paper is the one that most media-ready. Would the media be all over a study that made the bold claim that changes in seismic velocities can tell us a little something about the state of pressure in the crust? I would venture to say no. Now, if you then say the change after the earthquake puts a big volcano in Japan -- a national icon -- into a "critical state" that could mean an eruption will occur soon? Stop the presses! Yet, this isn't the conclusion of the actual Science article at all. I have no way of knowing, but this external "conclusion" about Fuji could have been originally included but was removed in the process of peer-review. I mean, we've seen this idea before -- that a certain increase in pressure mean Fuji will erupt -- but it has never really been shown to be verifiable. We're actually stuck in a chicken-and-egg loop here: Did the earthquake tell us that pressure is high enough for an eruption (that was going to happen anyway), or did the earthquake add more pressure and make an eruption more likely? Fuji is a dangerous (yet wonderful) volcano, as is any volcano near large population centers, so understanding its behavior and planning for an eruption is important. This is not to say that the science in the Brenguier and others (2014) article isn't good science. From what I can tell, it is. However, there is a fine line in my mind between promoting your work and going all P.T. Barnum on everyone. Maybe the quotes were taken out of context (although it seems unlikely). Without understanding what actually triggers an eruption at Fuji (or any volcano for that matter) and without knowing whether the pressure in the crust in these volcanoes is due to magma or hydrothermal fluids, it is definitely a stretch to say that "the risk is clearly higher." However, it does make much splashier press to lead with "Fuji could erupt" over "seismic waves changed velocity." Author's note: Nick Wigginton is right to point out that Science itself did not promote the Fuji angle. However, Science is being used as a platform by the author's to promote this idea even if it isn't in their paper. At what point can a journal reign in the press releases from the authors or authors' institution if the message they are sending isn't in line with the published paper? The oddest thing about the paper is that it starts with the supposition that changes in pressure can trigger eruptions, yet then never directly ties that to a potential Fuji eruption. I think it's this angle, again, that is the "flashiness" that Science and Nature seek - a paper on seismic wave behavior is linked to volcanic behavior, then promoted by the authors in the media as if there is an obvious and direct connection, thanks to the press release from the author's home institution and author interviews.