Extended Data Fig. 3: Overlapping PKiKP coda observations.
From: Enhanced inner core fine-scale heterogeneity towards Earth’s centre
a, Locations of arrays AKASG (red triangle) and CMAR (blue triangle). The red star is a Category 1 earthquake recorded at AKASG and the blue stars are three Category 2 earthquakes recorded at CMAR. The solid lines are the ray paths from earthquakes to arrays and the dots are the corresponding bounce points on the ICB. These different earthquake-array combinations sample the same inner core region but different crust and mantle structure. Array response functions at 1 Hz with a slowness of 0 s/° are shown for b, AKASG and c, CMAR. d, Single (red) and stacked (blue) denoised PKiKP envelopes from AKASG and CMAR, respectively. e, Example delay-and-sum beam at the YKA seismic array in Canada. The earthquake (Mw 6.1, 2013/04/21 03:22:16 utc, 29.93°N, 138.89.62°E, 422 km depth, USGS-NEIC) occurred near Japan about 71°. Grey windows mark the signals used in f and g. f, Slowness vector analysis of the direct PKiKP phase (first grey window on e). Blue dots mark the theoretical slowness from AK135. White crosses indicate vertical incidence. Pink crosses represent bootstrap solutions determined by repeating the slowness analysis for 50 randomly resampled sets of array elements. g, Similar to f but for a time window deep in the coda (second grey window on e). Note that the coda energy has a lower ray parameter than the direct PKiKP wave but with a similar back azimuth, implying that it is arriving from great circle heterogeneity deep within the inner core beneath the PKiKP bounce point.
