That's very interesting. If the inner core rotates 3 degrees a year and the earthquake data covers 20 years, that would be a 60 degree shift from start to end o

That's very interesting. If the inner core rotates 3 degrees a year and the earthquake data covers 20 years, that would be a 60 degree shift from start to end of the data set. I looked over the paper again, and they don't mention this at all. Edit: I emailed the author of the paper (Xiaodong Song). He says that more recent work suggests 0.3-0.5 deg/year for core rotation. (This would make the rotation much less important.) http://www.geology.illinois.edu/people/xsong/Sites/papers/zhangetal05_science.pdf http://www.geology.illinois.edu/people/xsong/Sites/papers/zh... http://www.geology.illinois.edu/people/xsong/Sites/papers/lindner_etal10_jgr.pdf http://www.geology.illinois.edu/people/xsong/Sites/papers/li...