First the link to this week’s complete list as HTML and as PDF.
The commentary by Beroza is wrongly listed under geology, it ought to be placed prominently in the social sciences, epidemiology and probably climatology. “A more subtle effect is cherry picking, which amounts to defining retrospectively the behaviour that is considered anomalous… They find, for example, that, when anomalous clusters are defined retrospectively, 30% of realizations of a Poissonian earthquake catalog will contain clusters that should occur less than 1% of the time.” After having refined all your parameters on one set of data you ought only to report the results on another and independent set, difficult to do where data are rare and can’t be generated at will as in archaeology.
An enormous amount of effort has been expanded over many years on an effect without first making sure it’s really there. Why do Fritz et al. remind me of anthropogenic global warming?
Applying the rule “if you can’t see it, it aint there” to Carrière et al.’s figure 3 I’d say that even if the “prediction” were valid it is at least not meaningful in practice.
What more can I add to Jefferson except the two words “publication bias”. Apparently countries can spare billions to stockpile a useless remedy but lack a few hundred for reviewing data. And again: No replication – no science. See also theconversation.edu.au/
I’m glad that France takes a more sensible approach to safety than Dr. rer. nat. A. Merkel, who saw fit to replace clean German nuclear plants with old and polluting oil-fired Austrian ones. That said I wonder where Mr D. Butler has been hibernating the last five decades. While bunkered emergency systems doubtlessly are a worth-while quantitative improvement “keeping systems running in the crucial hours after an accident” is hardly a novel idea but the most basic rule taught in Aachen decades ago. Also in Fukishima the lines of ‘defense in depth’ did not collapse but against better judgement at the time were not installed in the first place. The basic design rule of diversity translates to “redundant safety systems must not all fail from one common cause”. In spite of being situated in a known tsunami area not a single generator was protected against flooding.