Topic Additional seismogram examples at distances beyond 100. Example 1: Earthquake in the Chile-Bolivia border region (intermediate source depth)

Similar documents
Datasheet DS USGS NEIC-data: OT 11:56: N 23.55E h = 10km mb = 5.8

Estimating the epicenters of local and regional seismic sources, using the circle and chord method (Tutorial with exercise by hand and movies)

Global Broadband Arrays a View from NORSAR

U.S. Department of the Interior. Geological Survey. Experiments with SRO and GRF-array data. George L. Choy. Open-File Report

Magnitude determination using duration of high frequency energy radiation for the 2011 Off the Pacific Coast of Tohoku Earthquake

Coda Waveform Correlations

Short Notes Characterization of a Continuous, Very Narrowband Seismic Signal near 2.08 Hz

Array-seismology - Lecture 1

Multicomponent seismic polarization analysis

LOS 1 LASER OPTICS SET

Module 2 WAVE PROPAGATION (Lectures 7 to 9)

Spatial variations in field data

=, (1) Summary. Theory. Introduction

A k-mean characteristic function to improve STA/LTA detection

10.2 Images Formed by Lenses SUMMARY. Refraction in Lenses. Section 10.1 Questions

A Rayleigh wave back-projection method applied to the 2011 Tohoku earthquake

TOWARD A RAYLEIGH WAVE ATTENUATION MODEL FOR EURASIA AND CALIBRATING A NEW M S FORMULA

Identification of High Frequency pulse from Earthquake asperities along Chilean subduction zone using strong motion

A TECHNIQUE FOR AUTOMATIC DETECTION OF ONSET TIME OF P- AND S-PHASES IN STRONG MOTION RECORDS

Retrieving Focal Mechanism of Earthquakes Using the CAP Method

Environmental Acoustics and Intensity Vector Acoustics with Emphasis on Shallow Water Effects and the Sea Surface

CONTRIBUTION OF THE IMS GLOBAL NETWORK OF HYDROACOUSTIC STATIONS FOR MONITORING THE CTBT PAULINA BITTNER, EZEKIEL JONATHAN, MARCELA VILLARROEL

New Metrics Developed for a Complex Cepstrum Depth Program

Detection and Identification of Small Regional Seismic Events

Rec. ITU-R P RECOMMENDATION ITU-R P PROPAGATION BY DIFFRACTION. (Question ITU-R 202/3)

Physics. Light Waves & Physical Optics

Attenuation in the uppermost inner core from PKP data observed at Romanian stations

Presence/absence of temporal change of inner core travel times. observed for six new seismic ray paths: further evidence for inner core

Unit 1.5 Waves. The number waves per second. 1 Hz is 1waves per second. If there are 40 waves in 10 seconds then the frequency is 4 Hz.

Fine Recording in Time Direction for Seismological Observation

A multi-window algorithm for real-time automatic detection and picking of P-phases of microseismic events

On the inner outer core density contrast from PKiKP/PcP amplitude ratios and uncertainties caused by seismic noise

This presentation was prepared as part of Sensor Geophysical Ltd. s 2010 Technology Forum presented at the Telus Convention Center on April 15, 2010.

Site Response from Incident Pnl Waves

RECOMMENDATION ITU-R P Prediction of sky-wave field strength at frequencies between about 150 and khz

Estimation of the Earth s Impulse Response: Deconvolution and Beyond. Gary Pavlis Indiana University Rick Aster New Mexico Tech

Study of Low-frequency Seismic Events Sources in the Mines of the Verkhnekamskoye Potash Deposit

29th Monitoring Research Review: Ground-Based Nuclear Explosion Monitoring Technologies REGIONAL EVENT IDENTIFICATION RESEARCH IN ASIA

2008 Monitoring Research Review: Ground-Based Nuclear Explosion Monitoring Technologies

24th Seismic Research Review Nuclear Explosion Monitoring: Innovation and Integration

Comparison of regional seismic phases interpretation in REB and KazNDC bulletins. Zlata I. Sinyova, Natalya N. Mikhailova

A robust x-t domain deghosting method for various source/receiver configurations Yilmaz, O., and Baysal, E., Paradigm Geophysical

RADIO WAVE PROPAGATION

From Last Time Wave Properties. Description of a Wave. Water waves? Water waves occur on the surface. They are a kind of transverse wave.

Performance of the GSN station SSE-IC,

MULTI-CHANNEL SAR EXPERIMENTS FROM THE SPACE AND FROM GROUND: POTENTIAL EVOLUTION OF PRESENT GENERATION SPACEBORNE SAR

28th Seismic Research Review: Ground-Based Nuclear Explosion Monitoring Technologies SEL0: A FAST PROTOTYPE BULLETIN PRODUCTION PIPELINE AT THE CTBTO

EWGAE 2010 Vienna, 8th to 10th September

A novel tunable diode laser using volume holographic gratings

A COMPARISON OF TIME- AND FREQUENCY-DOMAIN AMPLITUDE MEASUREMENTS. Hans E. Hartse. Los Alamos National Laboratory

By choosing to view this document, you agree to all provisions of the copyright laws protecting it.

Seismo-Acoustic Studies in the European Arctic

INFLUENCE OF STATIC DISPLACEMENT ON PEAK GROUND VELOCITY AT SITES THAT EXPERIENCED FORWARD-RUPTURE DIRECTIVITY

Bulletin of the Seismological Society of America, Vol. 73, No. 1. pp , February 1983

Performance Factors. Technical Assistance. Fundamental Optics

Full Polarimetric THz Imaging System in Comparison with Infrared Thermography

28th Seismic Research Review: Ground-Based Nuclear Explosion Monitoring Technologies

29th Monitoring Research Review: Ground-Based Nuclear Explosion Monitoring Technologies

TRAIN INDUCED SEISMIC NOISE OF ACCELERATING AND DECELERATING TRAIN SETS

North Pacific Acoustic Laboratory (NPAL) Towed Array Measurements

REGIONAL WAVEFIELD ANALYSIS USING THREE-COMPONENT SEISMIC ARRAY DATA

Radio Direction Finding System, a new perspective for global crust diagnosis

Here the goal is to find the location of the ore body, and then evaluate its size and depth.

Microwave Optics. Department of Physics & Astronomy Texas Christian University, Fort Worth, TX. January 16, 2014

Experiment 1: Fraunhofer Diffraction of Light by a Single Slit

CS 445 HW#2 Solutions

FRAUNHOFER AND FRESNEL DIFFRACTION IN ONE DIMENSION

Establishment of New Low-Cost and High-Resolution Real-Time Continuous Strong Motion Observation Network by CEORKA

SURFACE WAVE SIMULATION AND PROCESSING WITH MATSEIS

Low wavenumber reflectors

The ICB separates the liquid outer core from the solid inner core

PHYSICS 289 Experiment 8 Fall Geometric Optics II Thin Lenses

South Africa CO2 Seismic Program

Radial trace filtering revisited: current practice and enhancements

29th Monitoring Research Review: Ground-Based Nuclear Explosion Monitoring Technologies

Applications of Acoustic-to-Seismic Coupling for Landmine Detection

Constructing response curves: Introduction to the BODE-diagram

Polarization Filter by Eigenimages and Adaptive Subtraction to Attenuate Surface-Wave Noise

stacking broadside collinear

(A) 2f (B) 2 f (C) f ( D) 2 (E) 2

Part 1: Standing Waves - Measuring Wavelengths

Rec. ITU-R F RECOMMENDATION ITU-R F *

HIGH-ORDER STATISTICS APPROACH: AUTOMATIC DETERMINATION OF SIGN AND ARRIVAL TIME OF ACOUSTIC EMISSION SIGNALS

GRENOUILLE.

Experiment 19. Microwave Optics 1

Diffraction. Interference with more than 2 beams. Diffraction gratings. Diffraction by an aperture. Diffraction of a laser beam

MODULE P6: THE WAVE MODEL OF RADIATION OVERVIEW

Infrasonic Observations of the Hekla Eruption of February 26, 2000

Travelling Wave, Broadband, and Frequency Independent Antennas. EE-4382/ Antenna Engineering

From Last Time Wave Properties. Description of a Wave. Question. Examples. More types of waves. Seismic waves

PRIME FOCUS FEEDS FOR THE COMPACT RANGE

EXPLOITING AMBIENT NOISE FOR SOURCE CHARACTERIZATION OF REGIONAL SEISMIC EVENTS

Aberrations of a lens

UNIT Write short notes on travelling wave antenna? Ans: Travelling Wave Antenna

Chapter 9. Seismic Arrays (Version December 2011; DOI: /GFZ.NMSOP-2_ch9)

Unexplained Resonances in the Gravitation Field of the Earth

Global Maps with Contoured Ionosphere Properties Some F-Layer Anomalies Revealed By Marcel H. De Canck, ON5AU. E Layer Critical Frequencies Maps

The 16 August 1997 Novaya Zemlya Seismic Event As Viewed From GSN Stations KEV and KBS

OPERATION OF THE HITACHI S-450 SCANNING ELECTRON MICROSCOPE. by Doug Bray Department of Biological Sciences University of Lethbridge

Chapter 29/30. Wave Fronts and Rays. Refraction of Sound. Dispersion in a Prism. Index of Refraction. Refraction and Lenses

Transcription:

Topic Additional seismogram examples at distances beyond 100 Authors Klaus Klinge, Federal Institute for Geosciences and Natural Resources, Seismological Central Observatory, Gräfenberg (SZGRF), Mozartstrasse 57, 91052 Erlangen, Germany, Fax: +49 9131 8104 099, E-mail: klinge@szgrf.bgr.de Siegfried Wendt, Geophysical Observatory Collm, University of Leipzig, D-04779 Wermsdorf, Germany, E-mail: wendt@rz.uni-leipzig.de Peter Bormann (formerly GeoForschungsZentrum Potsdam, Telegrafenberg, 14473 Potsdam, Germany); E-mail: pb65@gmx.net Version October, 2002 Note: Most of the examples given below show either records of the German Regional Seismic Network (GRSN; aperture about 500 x 800 km) or of the Gräfenberg broadband array (GRF; aperture 45 x 110 km; see Figs. 8.14 and 9.4 in the manual Chapters 8 and 9, respectively). The following abbreviations have been used: OT origin time in UT (universal time), D epicentral distance in degrees, BAZ backazimuth in degrees, h focal depth in kilometer. Example 1: Earthquake in the Chile-Bolivia border region (intermediate source depth) USGS NEIC-data: 2002-03-08 OT 04:56:21.7 21.6S 68.2W h = 123 km Mw = 6.5; The epicentral distances of the GRSN stations range between 97 < D < 103. Figures 1 a and b show enlarged sections of the recordings made of this intermediate depth earthquake at stations of the GRSN around the beginning of the shadow zone of the Earth's core (for the full records see Fig. 11.57 left). Figure 1a depicts the first 24 minutes of the long-period recordings (SRO-LP simulation) of the horizontal R and T components whereas Figure 1 b shows the related Z-component recordings of the first 8 min (lower part) and 17 min (upper part), respectively. The following diagnostic features can be recognized: Pdif is by far the smallest body-wave arrival, even in the vertical component, however, its SNR is still large enough for this medium-size earthquake in LP recordings; SKS (here also with its depth phase ssks) forms the first strong shear-wave arrival, necessarily in the radial (R) horizontal component, whereas S (also with its depth phase ss) is a comparably strong later shear-wave arrival, here (however not always!) with large amplitudes in the transverse (T)-component records; SP (and its depth phase) is also (necessarily!) strongest in the R component whereas SS is frequently (not always! see Figure 2) strongest in the long-period T component; SKS and SP have also well developed onsets in Z-component LP records with amplitudes comparably strong or even larger than for Pdif and PP; Pdif and PP may also have well developed depth phases (here spdif and spp, respectively) in Z-component LP records; the missing of the depth phases ppdif and ppp in Figure 1b is due to the different P- and S-wave radiation pattern from a shear source (see Figs. 3.25 and 3.26) and the specific rupture orientation with respect to the seismic station in the considered case; this is not a general feature, rather pp, ppp etc. are often stronger; PP has, when compared with Pdif, a larger R component because of its larger incidence angle; and SP a larger Z component than PS. 1

Figure 1a SRO-LP filtered BB records of GRSN stations. Upper traces: R component; lower traces: T component. Note that the phases SKS and SP, as well as their depth phases, have necessarily their largest horizontal amplitudes in the R component, as the longitudinal waves Pdif and PP. In contrast, S and SS may have their largest amplitudes in the T component, depending on the primary ratio of SV/SH energy radiated by the earthquake source. 2

Figure 1b Enlarged vertical (Z)-component SRO-LP filtered BB records of GRSN stations shown in Figure 1a. Upper traces: Body-wave phases up to SP; lower traces: cut-out of the first 8 minutes of the record with longitudinal phases only. Note the well developed depth phases spdif and spp. The theoretically expected arrival times according to the IASP91 model (Kennett and Engdahl, 1991) have been inserted in both Figure 1a and b. 3

Example 2: source depth) Earthquake in the Mariana Islands region (intermediate USGS NEIC-data: 2002-04-26 OT 16:06:06.8 13.1N 144.6E h = 85 km Mw = 7.1; The epicentral distances of the GRSN stations range between 101 < D < 108. The waves from this source at nearly the same distance as for the Chile-Bolivia earthquake approach the GRSN stations from nearly the opposite backazimuth (see Fig. 11.57), however, the general wave types and waveform features in different record components are nearly the same as Figure 1, with one exception: SS is best developed in the R component and there is no S wave visible in the T component, i.e., the shear-wave energy generated by this source was almost exclusively of SV type. Figure 2a SRO-LP filtered BB records of GRSN stations. Upper traces: R component; lower traces: T component. Note that in these records PS has about four times larger amplitudes than SKS and all phases are not visible or rather weak in the T component. 4

Figure 2b Enlarged cut-out of the SRO-LP filtered BB records of GRSN stations shown in Figure 2a. Upper traces: Z component; lower traces: R component. Note the well developed depth phases spdif and spp. The theoretically expected arrival times according to the IASP91 model (Kennett and Engdahl, 1991) have been inserted in both Figure 2a and b. Note that the amplitude ratio Z/R is larger than for Pdif than for PP because of the smaller (steeper) incidence angle of Pdif. 5

Example 3: Earthquake in the Banda Sea region (intermediate source depth) with phases PKPdif, PKKPab and PKKPbc USGS NEIC-data: 2000-03-03 OT 22:09:13.5 7.313S 128.642E h = 148 km mb = 6.4; (D =113 and BAZ =73.9 from GRF, h = 140 km) Shown are various short-period filtered seismograms (WWSSN-SP simulation) recorded at GRSN-, GRF- and GEOFON-stations. All traces are sorted according to increasing epicentral distance within the range D = 110.8 (RUE) and 116.2 (WLF). Phases Pdif, PKiKP, PP, SP, PKKPbc and PKKPab are shown. Additionally, ray-paths and travel-time curves are presented. Figure 3a Pdif (old name Pdiff) is the first arrival, however, as a diffracted wave, rather small, particularly in short-period records. Only a few GRSN stations have recorded it from this event. For more clear long-period records of Pdif see Figures 4b, 6c and 7b in the next examples. 6

Figure 3b Phases PKiKP, PP and (ppp) are shown on this seismogram. PKiKP arrives about 4 min after Pdif shown in Figure 3a. Note the strong variation at PKiKP-amplitudes within the network by a factor of about 10. The onset-time of PP can not be determined exactly. The reflection point of PP is below the complex crustal structure of Tibet. 7

Figure 3c Ray paths for Pdif, PKiKP, PKKPbc and PKKPab for the considered event at a focal depth of 148 km and with an epicentral distance to station CLL of D = 111. Figure 3d Travel-time curves for PKKPbc and PKKPab for a focal depth of 148 km. 8

Figure 3e Vertical component short-period seismograms starting about 10 min after the PKiKP arrival in Figure 3b. Marked are the onset times of the phases SP, PKKPbc and PKKPab. For the branching of PKKP between about 90 and 125 see Figure 9 in EX 11.3. These late secondary phases appear in short-period records only. Their ray paths are shown in Figure 3c and their travel-time curves in Figure 3d above. Example 4: Earthquake in Papua New Guinea USGS NEIC-data: 1998-07-17 OT 08:49:15 3.08S 141.76E h = 33G Ms = 7.0; (D = 117.5 and BAZ = 58.8 from GRA1) This earthquake occurred near the coast of Papua New Guinea. Tsunami waves with a height of 10 m flooded the coast and killed about 3000 people. 9

Figure 4a Long-period filtered three-component seismogram (SRO-LP simulation) recorded at station GRA1 (D = 117.5 ). The horizontal components (RT) are rotated with R into the epicentral direction. Phases Pdif, PP and a strong SP are visible on the vertical component. While the phases SKS, SKKS and PS, polarized in the vertical propagation plane, are strong on the radial (R) component Sdif, SS and SSS are strong on the transverse (T) component. Figure 4b Long-period filtered vertical-component seismograms (WWSSN-LP simulation) recorded at 11 GRSN- and GRF-stations. Traces are sorted according to increasing distance (D = 115.4 to BRG and 119.9 to BFO). Long-period onsets of Pdif and PP are very clear as well as an unidentified phase X ahead of PP. The time differences between X and PP differ from station to station. Probably X results from interfering waves PKPdf and PKiKP. An answer is given in Figure 4c. 10

Three minutes before the Pdif onset from the earthquake in Papua New Guinea, a small P- wave onset was recorded in the seismograms of GRSN stations from an earthquake in Costa Rica. Its source parameters were: USGS NEIC-data: 1998-07-17 OT 08:49:01 8.59N 83.07W h = 33G mb = 5.9; (D = 86.3 and BAZ = 278.8 from or GRA1) Figure 4c Long-period filtered vertical-component seismograms (WWSSN-LP simulation) recorded at 11 GRF-array stations within the distance range D = 86.3 to 86.7. Long-period onsets of the P wave from the Costa Rica event appear first, followed by Pdif, PKP and PP from the earthquake in Papua New Guinea. The phase X from Figure 4b could be identified from the GRF-array records as being PKPdf from the Papua New Guinea earthquake with a slowness of about 2.0 s/. 11

Figure 4d Short-period filtered seismograms (WWSSN-SP simulation) recorded from the Costa Rica earthquake at 11 GRSN-stations. Traces are sorted according to increasing epicentral distance (D = 83.1 to GSH and 88.0 to GEC2). The station-network allows to separate both events by way of slowness and azimuth determination. The estimated bodywave magnitudes mb vary for the GRSN-stations between values below 5.0 (BUG, GSH) and 5.4 for station WET. Example 5: Earthquake in the region of New Britain, P.N.G. USGS NEIC-data: 1999-05-10 20:33:02.1 5.173S 150.915E h = 137D mb = 6.5; (D = 124 and BAZ = 51 from GRFO) This magnitude 6.5 earthquake, recorded at GRFO and at the GRSN-network, shows the phases Pdif, PKP, PP, Sdif, PS, SS, depth phases and because of a focal depth of 137 km weak surface waves. A rare example is the well recorded phase P'P'P'P' = PKPPKPPKPPKP in Figure 3e below. The ray paths for the identified phases of this event are shown in Fig. 11.60. For animation of ray propagation and seismogram formation see CD-ROM attached to Volume 2 and complementary explanations given in IS 11.3. 12

Figure 5a Long-period filtered three-component seismogram (WWSSN-LP simulation) recorded at station GRFO (D = 124 ). The length of the record is about 1 hour. Phases Pdif, PKP, strong PP, Sdif, strong PS, strong SS and according to the large focal depth of the earthquake (h = 137 km) relatively weak surface waves are recorded. Figure 5b Shown are the first 10 minutes from the seismogram in Figure 5a. The higher time resolution separates well the phases Pdif, PKPdf, ppkpdf, an unidentified phase X (25s ahead of PP), PP and the depth phases ppp and spp. 13

Figure 5c Long-period Z-component records (WWSSN-LP simulation) at 14 GRSN-, GEOFON- and GERESS-stations within the distance range D = 120.1 (RGN) to 126.6 (WLF). This record section was used to perform the vespagram analysis ( see Figure5d). Figure 5d Vespagram analysis for vertical-components was used for slowness determination and phase identification. Phases were identified according to slowness values and travel times. The seismogram analysis program SHM allows to choose slowness steps (here in increments of 0.5 s/ ). The analysis yields slowness values of 4.5s/ for Pdif, 2.0s/ for PKPdf and ppkpdf, 7.0s/ for PP and for the phase X a value that would correspond to Pdif. The ray path could not be identified. 14

Figure 5e The network records show the rare phase 4P' (PKPPKPPKPPKP) with slowness S = 2.5 s/.it arrives 79 minutes after the origin time from the opposite azimuth direction (BAZ = 239 ). Example 6: Earthquake in the region of Salomon Islands USGS NEIC-data: 1996-04-29 OT 14:40:41.2 6.516S 155.037E h = 44G mb = 6.3 Ms = 7.5; (D = 127.3 and BAZ = 47.5 from GRA1) The example below presents clear Pdif onsets on long-period filtered seismograms as well as an excellent phase 4P' = P'P'P'P' = PKPPKPPKPPKP arriving about 63 min after the first onset. All records were made at the GRF-array (see Fig. 11.3). 15

Figure 6a Broadband Z-component records at the GRF array within the distance range D = 127.2 to 127.8 which show, besides the rather long-period Pdif, also PKPdf and PP. Figure 6b The same record as Figure 6a, however, a short-period filter was applied (WWSSN-SP simulation). Pdif is no longer visible above the noise level. 16

Figure 6c The same records as in Figures 6a and b, however, after applying a long-period filter (SRO-LP simulation). A distinct Pdif appears on the record. Figure 6d Short-period filtered (WWSSN-SP simulation) seismogram. The rare phase 4P' = P'P'P'P' (PKPPKPPKPPKP) was clearly recorded at the GRF-array coming from opposite direction (BAZ = 226 and S = 2.56 s/ ) about 63 min after the first P onset. Trace No.12 (the sum of all traces = BEAM) shows the best signal-noise-ratio. 17

Example 7: Earthquake in Santa Cruz Island USGS NEIC-data: 1998-07-16 OT 11:56:36 11.1S 165.9E h = 110G; (D = 136.2 and BAZ = 37.3 from GRA1). Figure 7a Vertical-component velocity BB seismograms recorded at the GRF-array (distance range between 136.1 and 136.8 ). Note the very weak Pdif. The depth phases of PKPdf and PP correspond to a depth of 110 km. 18

Figure 7b The same record as in Figure 7a, however, after long-period filtering (WWSSN- LP simulation). Note the pronounced onsets Pdif and ppdif. Also remarkable is the longperiod and slowly emerging onset 24 sec before PKP. Example 8: Earthquake in the region of Fiji Islands USGS-QED-data: 1994-09-30 OT 19:30:06.9 21.102S 179.204W h = 500km(G), mb = 5.1; The example shows pronounced core phases in the distance range 148-152. Short-period P waves reappear in this distance range but with discontinuous travel-time branches. In the region of the caustic, i.e. around 144, the whole energy of direct longitudinal core phases is concentrated thus forming a strong PKP onset. Its amplitudes are comparable with that of P waves at much shorter distances (D around 50 ). At distances beyond the caustic point PKP onsets are separated into individual PKP branches. The energy distribution changes with increasing distance. PKPbc (or PKP1) is the dominant branch just beyond the caustic, up to about 153. In records of weaker events, PKPbc is often the first visible onset since PKPdf (or PKIKP), preceding PKPbc, is too weak to be observed. With increasing distance near 160 PKPbc vanishes from the record and PKPab (or PKP2) dominates the seismogram. For more details see sub-section 11.5.2.4. 19

Figure 8a Short-period filtered three-component seismogram (WWSSN-SP simulation) recorded at the broadband station CLL at an epicentral distance of D = 148.4 (BAZ = 22 ). Phases PKPdf, PKPbc and PKPab are easy to analyse in this distance range. Travel-time curves for the different branches of these core phases allow to determine the epicentral distance from records of a single station only with an error less than 2. Note the relatively very small amplitudes on the horizontal components because of the very small (steep) incidence angle at this large distance. Figure 8b Vertical-component record of station CLL, displayed with higher magnification. 20

Figure 8c Short-period filtered seismograms (WWSSN-SP simulation) recorded at 9 GRSNstations. All traces are time-shifted and aligned for PKPdf and sorted according to increasing distance. The epicentral distance for CLL is 148.4 and for the most distant station BFO 152.2. Example 9: Deep earthquake in the region of Fiji Islands USGS NEIC-data: 2000-12-18 OT 01:19:21.6 21.18S 179.12W h = 628 km mb = 6.4; The epicentral distances of the GRSN stations range between 146 < D < 153. Shown are enlarged cut-outs of the the PKP-wave group from the seismograms presented in Fig. 11.64. The PKP-wave group consists between 146 < D < 156 (160 ) of three distinct wave arrivals PKPab, bc and df. Fig. 11.59 shows the respective travel-time curves. The figures below demonstrate the effect of the record filter on the discrimination of these closely spaced wave arrivals and the possibility to pick their onset times. The longer the center period of applied narrowband band-pass filters, the worse is the discrimination and onset-time picking for these three core-phase arrivals. Therefore, PKP phases are best picked in SP records, however, the later phases PP and PPP from so distant earthquakes are usually recognizable only in LP or BB records. 21

Figure 9a Z component long-period (SRO-LP simulated) records of longitudinal wave groups from a deep (h = 628 km) Fiji Islands earthquake at stations of the GRSN. Note the strong onsets of the depth phases spkpdf and spp prior to onsets PP and PPP, respectively, which follow closely. 22

Figure 9b Enlarged cut-outs of the PKP-wave group as recorded in WWSSN-SP (upper traces) and SRO-LP filtered seismograms of the GRSN stations. Note that for all three core phases the theoretically expected onset times according to the IASP91 travel-time model are about 2 s too early. 23

Figure 9c Enlarged cut-outs of the PKP-wave group of the deep Fiji Islands earthquake of 18 December 2000 recorded at stations of the GRSN. Shown are, from top to bottom, the bandpass filtered BB records of two octaves bandwidth with center periods around 1s, 2s, 4s and 8 s. For center periods larger than 3s the separation of the different PKP-wave arrivals and recognition of their onset times becomes less and less clear. Note that according to the IASP91 travel-time model the three core phases should arrive about three seconds earlier. 24