Temporal Change Of Earth’s Inner Core Surface,Seismic Structure Near The CMB And North Korea’s Nuclear Test

Seismogram is the main tool in seismologic studies,and contains contributions by the seimic structure of the Earth and seismic source,both of which are the most significant research directions in seimological community.In general,it is hard to isolate these two contributions,so we usually use the relative method.In this PhD thesis,we present the following studies for seimic structures of the Earth and seismic source based on this method:(1)we study the physical mechanism for seismic observations of temporal change of PKiKP and PKIKP and their coda waves,and the temporal change of Earth’s inner core surface based on those observations;(2)we determine seismic structure near the core-mantle boundary(CMB)beneath Southeast Asia based on travel time analysis of S,ScS,P and ScP,and migration of PKP precursors;(3)using the previous test and small seismic events near the nuclear test site as references,we are able to determine high-precision geographic location of North Korea’s 2017 test,constrain its burial depth and source characteristics,and estimate its yield.Seismic observations of temporal changes of PKiKP and PKIKP travel times and waveforms have been reported by many studies during this two decades.Inner core differential rotation and temporal change of inner core surface are reported to provide explanations for these observations.However,the time frame and sampling regions for high-quality seismic observations are short and limited,respectively,and far from representative.A long-time and larger spatial observations are needed to resolve the issue that which of inner core differential rotation and temporal change of inner core surface can provide a reasonable and consistent explanation for the observed PKiKP and PKIKP temporal change.Firstly,we detect nine high-quality repeating earthquakes occurring in the South Sandwich Islands between 1993 and 2013.Based on PKiKP and PKIKP travel time analysis,it is too much a coincidence for inner core differential rotation to explain the long-time seismic observations,while temporal change of inner core surface provides a simpler one.In the early period(before 2003),inner core surface is enlarged by 0.53～1.78 km beneath the western coast of Gabon,Republic of Congo and southwest Tanzania,while the regions beneath Zimbabwe or/and Kenya,and beneath west Angola or/and north Central African Republic experience little change.In the later period(after 1998),the regions at the inner core boundary beneath the western coast of Gabon,Republic of Congo and southwest Tanzania either shrink or remain unchanged,and the temporal change migrates to the inner core surface beneath Zimbabwe or/and Kenya,and beneath west Angola or/and north Central African Republic,with decrease of inner core surface by up to 5.59 km between 1998 and 2009 beneath Zimbabwe or Kenya,and by 1.73 km beneath west Angola or north Central African Republic between 1998 and 2013.These seismic results indicate that the temporal change of the inner core surface occurs in localized regions and is episodic,rapidly migrating and alternately enlarged and shrunk in a time scale of years.We then conduct an extensive search of repeating earthquakes from 107127 events occurring globally between January 1 1990 and July 31 2015,and finally obtain high-quality seismic observations for one doublet and one cluster in the South Sandwich Islands and four clusters in the Middle America subduction zone selected from the global repeating earthquake dataset.We present three lines of seismic evidence showing that inner core differential rotation cannot provide a reasonable or consistent interpretation for the observed temporal change of seismic data whereas temporal change of inner core surface can.Firstly,inner core differential rotation provides an implausible explanation for a disappearing inner core scatterer located at the inner core boundary beneath northern Brazil.Secondly,inner core differential rotation yields a contradictory interpretation for the observed temporal change of PKIKP waves and their coda among a cluster,with one set of the data requiring inner core differential rotation and the other non-rotation.And thirdly,inner core differential rotation cannot provide a reasonable explanation for the temporal change of PKiKP waves in a short time scale.On the other hand,temporal change of inner core surface provides a consistent interpretation for all the observed temporal changes of PKIKP and PKiKP and their coda waves.We conclude that the observed temporal changes of inner core phases are caused by temporal changes of inner core surface.The temporal changes of inner core surface are found to occur in some localized regions and up to a time scale of days or months,a phenomenon that should provide important clues to our understanding of core dynamics.We constrain the average shear-velocity structure near the CMB beneath Southeast Asia based on travel time analysis of S,ScS,P and ScP phases.Using the observed PKP precursory energy,we map seismic scatterering in the region.The inferred average velocity perturbations in the lowermost 200 km of the mantle range from about-6%to 6%and exhibit a complex geographic distribution of alternate low-and high-velocity patches adjacent to each other,surrounded by a high-velocity anomaly in the south.The seismic scatterers exhibit a crescent shape distributed from the South China Sea to the Maluku Islands with large velocity perturbations.The geographic distribution of seismic scatterers coincides with the westernmost low-velocity patches,suggesting that the seismic scattering is caused by ultra-low velocity zones(ULVZs)in the region.We suggest that the seismic structure in the region likely results from a complex interaction between a downwelling and a low-velocity region near the CMB.The downwelling(the high-velocity patches)displaces the low-velocity region into many low-velocity patches and pushes the ULVZs to the edge of the low-velocity region.Taking advantage of our high-quality seismic observations of both body waves and surface waves with good azimuth coverage related to North Korea’s 2017 nuclear test,satellite imagery and the unprecedented opportunity of using small seismic events near test site as references,we are able to determine high-precision geographic location of the 2017 test,constrain its burial depth and source characteristics,and estimate its yield.The test is determined to be located at(41 ° 17’53.52"N,129°4’27.12"E)with a geographic precision of 100 m.The five tests since 2009 are located several hundred meters apart,beneath the same mountain Mantap.The burial depth of the 2017 test is estimated to be 760 m based on the absolute depth of an earthquake and relative depth between the test and the earthquake,which is consistent with that inferred from surface elevation difference between the associated tunnel entrance and the identified test location.We present a complete explosion source model for an underground nuclear test with an isotropic source accompanied by a double-couple attributed to tectonic release and an oblique compensated linear vector dipole(CLVD)attributed to source medium damage.We propose that the oblique CLVD accompanies the 2017 test,and tectonic release inferred from historic underground explosions may just be the appearance of an oblique CLVD.Surface waves excited by the isotropic source of the 2017 test are strongly contaminated by those related to the secondary source processes,so we regard the estimated yield of the 2017 test based on mb more convincing than that inferred from Ms.The yield of the 2017 test is estimated to be 108±48 kt based on its Lg-wave magnitude and inferred burial depth.