Man-made earthquakes and science communication

Two papers on man-made earthquakes have been published last week and both have had received media coverage. Especially now, few days after the L’Aquila trial, the public is interested in any earthquake story and so the new findings that severe earthquakes happened due to human action caused some attention. Additionally, it caused me a headache and triggered not an earthquake, but a feeling of anger. Let’s talk about some good examples for bad science communication. The Lorca Earthquake of 11 May 2011 resulted not only in 9 fatalities, but also led to severe damages. The earthquake reached a magnitude of Mw5.1, which is rather low, and happended at a depth of ~ 3 km, which is rather shallow. During the last decades, intense groundwater extraction took place in the Lorca region. This resulted in a groundwater level that is ~250 m lower than in the 1960s. González et al. (2012) modelled the stress change due to the crustal unloading and found that the main slip during the earthquake occurred where the stress change was significant. This led them to conclude “…that the distribution of shallow slip during the Lorca earthquake could be controlled by crustal unloading stresses at the upper frictional transition of the seismogenic layer, induced by groundwater extraction. Our results imply that anthropogenic activities could influence how and when earthquakes occur.” Well, other studies didn’t find anything peculiar related to the Lorca event (López-Comino et al., 2012; Martínez-Díaz et al., 2012), and I would be extremely careful in announcing that human activity could determine whenan earthquake occurs. I think we can trigger tectonic earthquakes, but we need a fault which is already at the end of its seismic cycle. Then, an earthquake would have happened anyhow within a short time. However, Gonzáles and his co-authors used “could” two times in the last two sentences of their abstract, so don’t blame them.

Impact factor 5.1.

Is it all about impact factors? Not always, but sometimes.

What exactly comes into my mind when I read that spectacular headings in journals like Nature or Science is the bon mot: “It was published in Nature, however, it might be true.” What journalists now made of this story is headlines like the following ones:

[But also quite fair ones like this:

Do you see the difference? The original research article might even include some audacious statements, but the media turned possible interpretations into true statements. So, how to tell the public that things are different when even science journalists get it wrong? The paper of Gonzáles et al. (2012) came accompanied by a letter from Avouac (2012). Here, the author discusses the issue of human-induced shaking and ends up with this conclusion “The consequences are far reaching: if ever the effect of human-induced stress perturbations on seismicity is fully understood, and provided it is proven to be a deterministic process, we might dream of one day being able to tame natural faults with geo-engineering.” I think statements like this aren’t helpful. There is really no point in adding this sentence to an article, except for causing some attention in the media. Controlling earthquakes by geo-engineering? Come on, don’t underestimate the power of nature.

The Klose (2012) article and German media

So, while the German news magazine SpiegelOnline did not bad in the Lorca case, few days later Axel Bojanowski, responsible for earth science articles, fell exactly into that tone of alarmism that the other media were already used to. He wrote about a paper of Klose (2012) dealing with the evidence of the causality of human-made mass shifts on earthquakes and obviously, he didn’t even read the abstract. But let’s start with the paper. Klose investigated small- to large-sized earthquakes and their relation to nearby large-scaled mass shifts like

  • groundwater extraction
  • groundwater injection
  • mining
  • hydrocarbon production
  • reservoir flooding etc.

Sounds like a reasonable idea to check if artificial stress changes trigger earthquakes, and we’re speaking of mass shifts of up to almost one Tt (1 teraton = 1,000,000,000,000,000 kg). He did statistical analyses and found that “relationships exist with statistical significance between (a) seismic moment magnitudes M of observed earthquakes, (b) lateral distances of the earthquake hypocenters to the geoengineering “operation points” and (c) mass removals or accumulations on the Earth’s crust“. Klose wrote that several EQs were induced artificially, others were triggered. Induced seismicity, he writes, might be observed in mining areas, despite I can not believe Australia’s 1989 Mw5.6 Newcastle earthquake would not have happened without coal mining. However, we know this from several areas, so no big news, but some additional knowledge, why not. Triggering was observed in other cases, like China’s 2008 M7.9 Wenchuan earthquake, since there the rupture nucleated below a recently built reservoir. Here, it’s important to mention that this is about a major quake, magnitude 7.9! We are not talking about human induced things like a mine collapse or settling due to reservoir load. We are talking about a 100 km long tectonic fault, which caused an immense earthquake. Human activity might maybe probably have had an influence on where the rupture nucleated (maybe), but not on the fact that an earthquake would have happened anyhow because the fault was due. Now here comes the SpiegelOnline article from 25 October. It’s headline is: Manmade earthquakes: The dissimulated causes of catastrophes (Menschengemachte Erdbeben: Die verheimlichten Ursachen der Katastrophen). Not only does Mr Bojanowski try to set up a kind of conspiracy story with bad mining companies trying to hide what they’ve done to the people, but he’s also completely ignoring the difference between triggering and inducing. Instead, he simply says the following earthquakes are caused by humans:

  • China 2008, Mw 7.9
  • Gazli, Uzbekistan 1984, M7.2
  • Hebgen, USA 1959, M 7,1
  • Coalinga, USA 1983, M6.5
  • Koyna, India 1962, M6.3

Let this settle a little. We caused several M7+ earthquakes.

Despite I doubt some of the correlations Klose (2012) came up with, he exactly discriminated between triggering and inducing events. Furthermore, Bojanowski is worrying about the San Andreas Fault, because it crosses the Salton Sea in California. 400,000,000 tons of water were pumped there already, according to Bojanowski 60 times the weight of the Cheops Pyramide (sic!). Klose wrote in his paper: “Pure strike-slip faults seem to rupture randomly and independently from the magnitude of the mass changes.” Unfortunately, this is not the first time to happen at SpiegelOnline, and it happens so often when it comes to natural hazards. After the Christchurch EQ, they interpreted the fringes of a DInSAR image as the underground being folded in a wave-like pattern. And Spiegel counts as one of the most reliebale news sources in Germany – imagine what the Yellow Press writes…

In terms of science communication, there is still a long way to go.

References:

Where is it?

Click on map to expand
Click to expand

Who was it?

Christoph Grützner
works at the Neotectonics and Natural Hazards Group, RWTH Aachen University, Germany. He likes the Mediterranean and uses geophysics to search for ancient earthquakes.