Yoshio Fukao

Whole mantle P-wave travel time tomography

Abstract A method of tomographic inversion to obtain three-dimensional velocity perturbations in the Earths whole mantle has been developed, and applied to more than two million P-wave arrival time data reported by International Seismology Center (ISC). The model is parameterized with 32,768 blocks; the divisions in latitude, longitude, and radius are 32, 64, and 16, respectively. Horizontal cell size is 5.6° × 5.6°. The layer thicknesses vary with depth; 29 km just below the surface and 334 km just above the core-mantle boundary. Starting from a spherically symmetric Earth model, we obtained a three-dimensional model using the following iterative procedures. First, we relocated all the events; second, we backprojected the residual into the whole mantle; third, we refined the spherically symmetric Earth model. The solutions have been converged in five iterations. We adopted the following techniques in the backprojection procedure. The first order smoothness was introduced as a damping, which makes the solution independent of the starting model and its apparent fluctuation minimal. The basic equations for delay times and smoothness are solved using the conjugate gradient method, an iterative method which guarantees the convergence of solution into the exact least squares solution. The weight on the smoothness, i.e., the damping factor, was objectively determined by a simplified cross validation technique. The final solution was obtained as an average of the ten solutions, each of which was derived from one tenth of the total data set. The reliability of the solution is examined in two ways: (1) mapping the resolution given by the reconstruction of checkerboard patterns, and (2) mapping the variance given from the gaussian noise input. This is the first result which delineates the whole mantle structure with its reliability mapping. The longest wavelength anomalies of the lower mantle are similar to those of previous studies by Dziewonski (1984) and Hager and Clayton (1989). However, the shorter wavelength patterns have many discrepancies. The fast anomalies of the uppermost mantle beneath the shield and the slow anomalies beneath the active tectonic regions are clearly seen. We also revealed the l = 2 pattern at the transition zone reported by the previous surface wave and free oscillation studies. The most striking feature is that of the fast anomalies lying along the extension of the large subducting plate up to depths of about 1000 km or more. The resolving powers in these regions are generally good.

Iterative deconvolution of complex body waves from great earthquakes. The Tokachi-Oki earthquake of 1968

Abstract A method of body-wave inversion is developed in an attempt to extract the information about asperities or barriers in a fault zone. A sequence of point sources, each being characterized with the seismic moment, the onset time and the location, are iteratively derived from observed records at multi-stations, where the two-dimensional extent of the source location is taken into account. A modification is made of the iterative method of Kikuchi and Kanamori on the formulation of inversion procedure to facilitate the computation. Using this method, we analyse long period P waves of the Tokachi-Oki earthquake of 1968 (Mw = 8.2) and obtain several significant subevents with time durations of ∼ 10 s. Their spatio-temporal distribution shows that the rupture process consists of three characteristic stages: (A) a stage of introductory rupture, (B) a stage of main rupture and (C) a stage of aftershocks. The main rupture takes place in the form of clustering around a few sites of the fault plane. The largest subevent occurs in the northwestern corner. The stress drop associated with this event is estimated to be ∼ 200 bars, one order of magnitude higher than the stress drop averaged over the entire fault plane. The sum of the seismic moments of the individual subevents amounts to 2.3 × 1028 dyn. cm which approximately coincides with the one estimated from the analysis of long-period surface waves. This implies that the source of the Tokachi-Oki earthquake consists of several major subevents with time durations of ∼ 10 s in addition to other minor subevents.

Inversion of long-period P-waves from great earthquakes along subduction zones

Abstract Long-period P-waves from subduction zone earthquakes in the moment range from 7×1027to 8×1029dyn · cm are analyzed using an iterative deconvolution method. In addition to discrete subevents which effectively contribute to the observed seismograms, the long-period component of the moment rate function is derived with the constraint that the total seismic moment is equal to the moment determined from long-period surface waves. It is shown that great earthquakes can be described by several subevents with a duration of about 10 s. which changes little from earthquake to earthquake in the above moment range. The long-period component of the moment rate function becomes larger for a greater earthquake. The greatest earthquakes of Mw ~ 9 additionally involve subevents with another duration scale (much longer than 10 s). It is also shown that the local stress drop is 1 order of magnitude larger than the stress drop averaged over the entire fault plane. The constancy of the local stress drop as well as the average stress drop holds among various magnitudes of earthquakes. These observations suggest an essential similarity in the mode of dislocation distribution.

Mechanism of large earthquakes along the eastern margin of the Japan sea

Abstract The source mechanisms of large earthquakes occurring along the eastern margin of the Japan Sea are investigated. The earthquakes studied are those off Oga (May 7, 1964), Sakhalin (Sep. 5, 1971) and Shakotan (Aug. 1, 1940) and, though supplementary, the Niigata (Jun. 16, 1964) earthquake. The focal mechanisms of these four earthquakes are predominantly dip-slip reverse faulting on a plane dipping relatively steeply. The strikes of the fault planes are parallel to the general trend of the eastern margin of the Japan Sea. The parameters for the dislocation are determined on the basis of the long-period surface waves and the tsunami heights at the coasts. The stress drop of 126 bars for the Niigata earthquake is significantly large compared to those for the other three shocks. The smallest is the stress drop of 17 bars for the earthquake off Shakotan. This earthquake was found to be especially interesting in two respects. First, the relative excitation of the short- and long-period seismic waves and the tsunami suggests that the dislocation of 1.1 m took place with an abnormally long duration. Second, the fault plane involves an exceptionally large area of 50 × 170 km 2 . This fault plane cuts through the entire thickness of the lid on the low-velocity zone developing beneath the Japan Sea. The fault plane may manifest a pre-existing fractured zone by which the Japan Sea lithosphere is separated from the northern Japan arc, as far as the west of Hokkaido is concerned.

Source retrieval for mantle earthquakes by iterative deconvolution of long-period P-waves

Abstract Long-period P-waves of eighteen deep and intermediate-depth earthquakes in a moment range 1×10 26 −3× 10 28 dyn · cm are analysed by the method of Kikuchi and Fukao (this issue) to iteratively derive point dislocations on a plane. The fault plane is defined as the plane of point dislocations that gives the best fit between the synthetic and observed waveforms. For dip-slip type earthquakes the fault plane is found to dip more steeply than the auxiliary plane. In general, point dislocations tend to cluster into one or a few subevents. We characterize the largest subevent by its maximum moment rate ( M 0 ) m and its half duration τ from which the associated stress drop Δσ can be roughly estimated. For earthquakes shallower than 400 km, half duration τ is approximately constant, 5 s, so that stress drop Δσ increases from 1000 bar as ( M 0 ) m increases. The value of τ for the great Banda Sea earthquake of 1963 ( h = 100 km ) is exceptionally large (τ~10 s, for which Δσ ~ 700 bar). For earthquakes deeper than 400 km, τ increases from 3 s to 9 s as ( M 0 ) m increases so that Δσ is approximately constant, 700 bar. Stress drop so obtained approximates to an overall stress drop with a possible exception of the several largest earthquakes. The distribution of point dislocations and the moment rate function indicate that rupture most often expands in a fan-shaped fault area and that a large part of the rupture front ceases propagation in a coherent manner. A sharp contrast in stress drop behavior across a depth of 400 km may be related to the olivine-spinel transition within the downgoing slab of lithosphere.

Seismic Tomogram of the Earth's Mantle: Geodynamic Implications

Recent seismic tomography of the Earths mantle has revealed a large-scale pattern of mantle convection comprising upwelling columnar plumes in the Pacific and Africa and downwelling planar sheets along the Circum Pacific. Upwelling and downwelling occur most extensively under the south Pacific and west Pacific, respectively. High-resolution image of plate subduction has been obtained from the dense seismic networks around Japan. Japanese seismologists are in the best position to resolve the internal structure of downwelling current as an integral part of the whole convection system.

Hierarchy in earthquake size distribution

Abstract We propose a fault model in which there are a finite number of characteristic lengths between the smallest and the largest scales, e.g., between a grain size and a size of great earthquakes. A fault plane in one scale is segmented into blocks of statistically similar sizes by the barriers of statistically similar heights. Rupture expands by successively “climbing” the barriers. The average barrier height is proportional to the average block size which defines the characteristic length of the relevant scale. Stresses before and after rupturing are assumed to be independent of earthquake size. With these assumptions the probability for rupture to grow further is shown to be independent of the scale of barrier structure. The rupture in one scale, however, tends to grow more easily as it grows so that once the rupture exceeds some critical size, it no longer stops until it encounters the barriers in the next larger scale. Rupture grows by repeating this cycle, which starts with rupturing of a unit block in one scale and ends with one in the next larger scale. Our model involves two parameters which must be determined empirically. One is the length ratio r between two successive scales. The other is the critical number of blocks, n c , which defines the critical size for unstable rupture growth in one scale. We suggest values of ∼ 5 for both r and n c . With these values our model explains a diversity of earthquake phenomena such as the magnitude-frequency relation, the seismic gap and the magnitude gap, the magnitude difference between a main shock and its largest aftershock, the occurrence of multiplet earthquakes, the generation of high frequency seismic waves, the incoherency of the near field acceleration and the recent observation by Furumoto and Nakanishi that the main rupture is often preceded by a precursory rupture whose time duration, in general, scales with seismic moment.

Foreshocks and multiple shocks of large earthquakes

Abstract The rupture process of the Tokachi-Oki earthquake (May 16, 1968) and that of the Kurile Islands earthquake (August 11, 1969) were investigated primarily on the basis of the long-period seismograms of WWSSN. These two large earthquakes are of the similar magnitude ( M = 7.8–7.9) and are only about 460 km apart from each other. The detailed analyses of the long-period P waves and the multiple surface waves of G2, G3, and G4 revealed the following features. 1. (1)The fracture at the earlier stages of the Tokachi-Oki event occurred in the form of multiple-shock activity. The largest two shocks were generated 57 km S and 93 km WSW of the initial hypocentre at 32 and 45 sec after the initial break respectively. 2. (2)The main faulting of the Tokachi-Oki event commenced about 80 km S of the initial epicentre at 40–45 sec after the initial break and propagated in the northerly to northwesterly directions, the rupture velocity being about 4 km/sec. 3. (3)No multiple event nature is seen for the Kurile event. Its initial break represents the commencement of the main-fault propagation. This earthquake was, however, preceded by an intense foreshock activity. The above features indicate that the multiple-shock activity of the Tokachi-Oki event and the foreshock activity of the Kurile event are the same type of precursory phenomena for the main phase of fault propagation. They are understood as the process for the nucleation of cracks at weak spots along the fault plane and their subsequent coalescence to grow larger cracks. This is the self-accelerated process resulting in the climax which marks the initiation of the main-fault propagation. We suggest that the rupture process of other large earthquakes can be described in a very similar way.

Seismic moment of great deep shocks

Abstract The long-period Rayleigh waves were investigated for the largest four deep shocks in 1963–1973 to determine the seismic moment by the same technique as used for shallow earthquakes. The results could be used for a quantitative comparison of source parameters between shallow and deep events. Three of the four shocks occurred beneath the South American continent (the Colombia earthquake, 1970; the western Brazil earthquake, 1963; the Peru—Bolivia border earthquake, 1963) and the other beneath the Japan Sea (1973). The focal depths are 653, 576, 593 and 575 km, respectively. The largest value of seismic moment was obtained as 2.1 · 1028 dyncm for the Colombia earthquake. This value is still about forty times smaller than that for the great Alaskan earthquake. A slight inconsistency was found between the first-motion diagram and the Rayleigh wave radiation pattern for the Colombia earthquake and the Peru—Bolivia border earthquake.

Degree‐2 pattern of attenuation structure in the upper mantle from apparent complex frequency measurements of fundamental spheroidal modes

The authors present preliminary results of an analysis of global lateral variations in elastic and anelastic structure of the upper mantle. Using the Sompi method, they have analyzed IDA records to measure apparent complex frequencies of fundamental spheroidal modes {sub 0}S{sub 21}-{sub 0}S{sub 34}, which are primarily sensitive to the structure of the transition zone. The apparent complex frequencies have then been inverted into local frequency and attenuation maps using the new inversion method. The authors have obtained consistent degree-2 patterns in the local attenuation maps as well as in the local frequency maps. The high-attenuation anomalies correlate with regions of slow P-wave velocity in the lower mantle and with regions of high hotspot density. This may represent a continuation of lower-mantle upwelling into the transition zone.