GEOPHYSICAL RESEARCH LETTERS, VOL. 22, NO. 16, PAGES 2237-2240, AUGUST 15, 1995 A very broadband study of the 1994 deep Bolivia earthquake sequence JunWu, Terry Wallace, Susan Beck SASOand Department of Geosciences, University of Arizona, Tucson Abstract. We have examinedthe source process of the June9, 1994 Bolivia earthquake in the period range of body waves to free oscillations. A series of moment tensor inversions were performed on various teleseismic wave types andfrequency bands to infer the source parameters andto map the temporal character of the source process. Time independent moment tensor inversion of the bodywaves (P andSH) yields a source time functionof 46 seconds duration with four major "pulses" of moment release, with moment release dominated by the third pulse. The focal mechanism is dominantly dip-slip onthesub-horizontal nodal plane (strike 2870 , dip 8 oand slip -740). The best bodywave fit requires a non-double-couple component of approximately 7%; although this couldbe due to an isotrepic component of the source, it is at the marginof resolution. Time dependent moment tensor inversion of the body waves, which yields independent time histories for each moment tensor element, shows that the source orientations (fault plane and slip direction) change very little throughout therupture duration. Analysis at several frequencies shows that the seismic moment is verystable (3.0x10 TM Nm+ 6%)at periods greater than 150 seconds and that the isotropic moment is negligible given the resolution. We also modeled the P-waves for the January 10, 1994 event,the strong aftershock of June 9, 1994 andthe August 8, 1994 earthquake.Rotation of the P axes for events after the mainshock may suggest a possible stress adjustment. Introduction The Bolivian earthquake of June 9, 1994 (M,•=8.3, depth=640 km) is thelargest "deep" earthquake to be recorded in the era of modern digitalseismic instrumentation. The event was well recorded globally on the IRIS GSN(Global Seismic Network), which allows a detailed analysis of the earthquake in a broad period band. The unprecedented quality of therecords allows us to search for the isotropic component in the seismic source whichsome phase-change models for deep earthquakes predict [Bridgman, 1945; Benioff, 1963]. Further, the relatively long duration of the earthquake allows us to investigate temporal evolution of the fault geometry. Studies on the aftershock location and source mechanisms reported elsewherein this issue [Myers eta/., 1995; Tinker et al., 1995] together with the teleseismicanalysis done here, provide the bestconstraints yet available on the character and dynamics of a de6p earthquake. The 1994 deepBolivia earthquake sequence occurred on a section of the subducted Nazcaslab whichhadno majordeep seismicity in the 30 years after the 1958 (July, 26) and 1963 Copyright 1995 by the American Geophysical Union. Paper number 95GL02232 0094-8534/95/95GL-02232503.00 (Aug.15, depth 543 km, m b=6 ) largedeep earthquakes [Chandra, 1970; Gilbert and Dziewonski, 1975; Fukao and Kikuchi, 1987]. This segment of the slab is between two active segments trending north-west north of 12øS and roughly north-south south of 17uS. The Nazca plate is thought to dip steeplybelow 500 km on thesesegments [Kirby et al., 1995]. These two zones are offset by approximately 700 km in southern Peru-central Bolivia (between 12 oand 17øS), a region which has undergone more shortening than anywherein western South America [Winchester et al., 1994]. Based on analysis of earthquake locations and focal mechanism solutions, Cahill and lsacks [1992] suggested that the southern Peru(13øS - 15øS) is-one of thethree transitions of subduction angle beneath the central Andes. In this narrow section the slab undergoes an abrupttransformation southward from nearly horizontal subduction toa descend about 30 ø. Deep seismicactivity in the central section started on Jan. 10, 1994 with a moderate-sized earthquake (mb=6.4, focal depth=600 km) located near the 1958 and 1963 deepearthquakes which were approximately 200 km west of the June 9 mainshock (See Figure 1). We were able to perform bodywave inversions for the January 10 event, June 9 mainsh0ck, June 9 aftershock -5' -10' -15' -25' ß Bolivia Sequence o -5' d>4•lan 150< d <399 -10' 285' 290' 295' 300' -15' Figure 1' Focal mechanisms of Bolivia deepearthquake sequence with deep seismicity (1974-1994, PDE catalog) in the Peru-Bolivia border region. The epicenter of the 1958and 1963deep events aretaken fromlsacks andMolnar (1971) The large event approximately 7 minutes after the M 8.3 event (Myers et al., 1995 ) is marked as M6. The shading represents topography measured in meters. 2237