Inner core rotation from event-pair analysis

The last decade has witnessed an animated debate on whether the inner core rotation is a fact or an artifact. Here we examine the temporal change of inner core waves using a technique that compares differential travel times at the same station but between two events. The method does not require precise knowledge of earthquake locations and earth models. The pairing of the events creates a large data set for the application of statistical tools. Using measurements from 87 events in the South Sandwich Islands recorded at College, Alaska station, we conclude the temporal change is robust. The estimates of the temporal change range from about 0.07 to 0.10 s/decade over the past 50 yr. If we used only pairs with small inter-event distances, which reduce the influence of mantle heterogeneity, the rates range from 0.084 to 0.098 s/decade, nearly identical to the rate inferred by Zhang et al. [Zhang, J., Song, X.D., Li, Y.C., Richards, P.G., Sun, X.L., Waldhauser, F., Inner core differential motion confirmed by earthquake waveform doublets, Science 309 (5739) (2005) 1357-1360.] from waveform doublets. The rate of the DF change seems to change with time, which may be explained by lateral variation of the inner core structure or the change in rotation rate on decadal time scale.