Shamita Das

Source mechanism of volcanic tremor: fluid-driven crack models and their application to the 1963 kilauea eruption

Abstract We propose a model for the mechanism of magma transport based on a fluid-filled tensile crack driven by the excess pressure of fluid. Such a transport mechanism can generate seismic waves by a succession of jerky crack extensions, if the fracture strength of rock varies in space, or if there is a difference between the dynamic and static values of the critical stress intensity factor. We also find that the opening and closing of a narrow channel connecting two fluid-filled cracks may be a source of seismic waves. Using the finite-difference method, we calculated the vibration of dry and fluid-filled cracks generated by: (1) a jerky extension at one end or at both ends and (2) a jerky opening of a narrow channel connecting two cracks. We then calculated the far-field and near-field radiation from these vibrating cracks. The spectra show peaked structures, but interestingly, most high-frequency peaks are only present in the near-field and cannot be transmitted to the far-field. The spectral features described above are often observed for volcanic tremors and in some cases for seismic signals associated with hydraulic fracturing experiments. We first consider as a model of volcanic tremor randomly occurring jerky crack extensions, and derive a formula relating the tremor amplitude to the excess pressure in the magma, the incremental area in each extension, and the frequency of extensions. These parameters are also constrained by other observations, such as the rate of magma flow. Our model was tested quantitatively against observations made in one of the best-described case histories of volcanic tremor: the October 5–6, 1963 Kilauea flank eruption. We found that a single, long crack extending from the summit to the eruptive site cannot explain the observations. The model of a steadily expanding crack ran into difficulties when quantitative comparisons were made with observations. The extension of crack area needed to explain the amplitude of volcanic tremor should accompany a large increase in tremor period which was not observed. Our second model is a chain of cracks connected by narrow channels which open and close. The length of each crack is around 1 km, the channel area connecting neighboring cracks is about 103m2, and the channel opens jerkily with the magmatic excess pressure of about 20 bars. The frequency of jerky opening of each channel is about once in 15 seconds. The channel is closed after each jerky opening, as soon as magma is moved through the channel.

Earthquake rupture stalled by a subducting fracture zone.

We showed that the rupture produced by the great Peru earthquake (moment magnitude 8.4) on 23 June 2001 propagated for ∼70 kilometers before encountering a 6000-square-kilometer area of fault that acted as a barrier. The rupture continued around this barrier, which remained unbroken for ∼30 seconds and then began to break when the main rupture front was ∼200 kilometers from the epicenter. The barrier had relatively low rupture speed, slip, and aftershock density as compared to its surroundings, and the time of the main energy release in the earthquake coincided with the barriers rupture. We associate this barrier with a fracture zone feature on the subducting oceanic plate.

The great March 25, 1998, Antarctic Plate earthquake: Moment tensor and rupture history

We use broadband body and mantle wave data to study the 1998 Antarctic intraplate earthquake. The centroid moment tensor (CMT) has a large non-double-couple component. There exist two pure double-couple constrained solutions that fit the data almost equally well. The frequent practice of taking the “best double-couple” gives a far from optimal solution. We use P and SH body waves to determine the rupture parameters of the first and larger of the two observed subevents. The best rupture plane, with strike 96°, dip 69°, and rake −18°, is compatible with only one of the two CMT solutions: strike 96°, dip 64°, rake −23°, centroid location (63.1°S, 148.4°E, 10 km depth), centroid time 0313:02 UT, and M0 = 1.3 × 1021 N m (Mw = 8.0). The first subevent is a simple, primarily westward propagating ∼140-km rupture, of ∼45-s duration, with average velocity ≳3 km s−1; it has a seismic moment of 1.2×1021 N m (Mw = 8.0), with 75% of its moment released between 10 and 27 s, and a stress drop of ∼240 bars. The rupture is physically bounded by two fracture zones at 147.5°E and 150°E. The second subevent lasted from 70 to 90 s on a fault extending from 210 to 270 km west of the epicenter, with a moment of 0.3–0.6×1021 N m (Mw = 7.6–7.8). This is a spectacular example of dynamic stress triggering over a 100-km separation distance with a time delay of ∼40 s. The complex pattern of aftershocks is primarily controlled by preexisting fracture zones on the ocean floor.

On the inverse problem for earthquake rupture: The Haskell-type source model

In order to gain insight into how to invert seismograms correctly to estimate the details of the earthquake rupturing process, we perform numerical experiments using artificial data, generated for an idealized faulting model with a very simple rupture and moment release history, and solve the inverse problem using standard widely used inversion methods. We construct synthetic accelerograms in the vicinity of an earthquake for a discrete analog of the Haskell-type rupture model with a prescribed rupture velocity in a layered medium. A constant level of moment is released as the rupture front passes by. We show that using physically based constraints, such as not permitting back slip on the fault, allows us to reproduce many aspects of the solution correctly, whereas the minimum norm solution or the solution with the smallest first differences of moment rates in space and time do not reproduce many aspects for the cases studied here. With the positivity of moment rate constraint, as long as the rupturing area is allowed to be larger than that in the forward problem, it is correctly found for the simple faulting model considered in this paper, provided that the rupture velocity and the Earth structure are known. If, however, the rupture front is constrained either to propagate more slowly or the rupturing area is taken smaller than that in the forward problem, we find that we are unable even to fit the accelerograms well. Use of incorrect crustal structure in the source region also leads to poor fitting of the data. In this case, the proper rupture front is not obtained, but instead a “ghost front” is found behind the correct rupture front and demonstrates how the incorrect crustal structure is transformed into an artifact in the solution. The positions of the centroids of the moment release in time and space are generally correctly obtained.

On the tectonics of Asia

Abstract This report describes an interpretation of the tectonics of central Asia made from seismic and geologic data. It is suggested that central Asia is not a tectonically passive unit, as previously proposed by others, responding solely to the convergence of the Indian plate with Asia. We postulate that the tectonics of central Asia can be represented by the motion of a few continental blocks which are influenced by spreading from the Baikal rift zone as well as compression due to the collision of the Indian plate. Here, a block is defined as a tectonic unit, within a continental plate, with boundaries delineated by broad zones of high seismicity with respect to the interior of the unit. Five tectonic units are postulated for central Asia. These are: the Siberian block, the East and West China blocks, the Southeast Asian block; and the Indian plate. An unusual phenomenon is noted along the boundary between the Siberian and West China blocks. There is general horizontal crustal compression along this boundary from the Hindu Kush north-eastward to the southern tip of Lake Baikal; however, there is general horizontal extension eastward from Lake Baikal through the Stanovoy range. Thus the West China block, to the south of this boundary, seems to be turning clockwise about a point near the southern tip of Lake Baikal. The major known faults within this block, which strike mainly northwest-southeast, may be interpreted as shear zones where interior stresses, due to the block rotation, are released. We cannot support this suggestion with an analytical model because of the uncertainties in various model parameters and geometries. The suggested model gives a possible explanation of why India, to the south of the Himalayas, is almost completely aseismic while the regions to the north and northeast have higher seismicity.

Diversity of solutions of the problem of earthquake faulting inversion; application to SH waves for the great 1989 Macquarie Ridge earthquake

Abstract We use the SH waves generated by the great 1989 Macquarie Ridge earthquake to investigate the non-uniqueness of the inverse problem of earthquake faulting. The details of the method of linear programming developed for this problem of solution of a constrained set of equations is described. Protective measures necessary to remove loss of accuracy when solving the problem for large numbers of unknowns and constraints are developed. It is shown that solutions which look very different fit the data almost equally well. A method is presented for choosing solutions with desirable properties (e.g. with the most uniformly distributed moment along the fault, or with the most concentrated moment release on the fault) from among available solutions. The results show that the data may equally well be interpreted as representing an asperity model of faulting or a propagating crack model of faulting. The common feature of all the models is that the hypocentral region is a region of low moment release, with regions of higher release being located to its northeast and its southwest along the fault. Clearly, this earthquake did not initiate by the rupture of an asperity.

Effect of Subducting Seafloor Topography on the Rupture Characteristics of Great Subduction Zone Earthquakes

Improvements in the quality and quantity of seismological data, together with technological advances in marine geophysics, mean that we are now able to examine in detail the infl uence of sea fl oor topography on the rupture process of great subduction earthquakes. Subducting seamounts were first suspected to affect the rupture process of a great earthquake in the 1986 M w 8.0 Andreanof Islands earthquake, where large slip was seen in isolated round patches in the direction of the plate subduction. Since then, we have been able to show that a ridge and trough feature on the subducting oceanic plate stalled the rupture process of the 2001 M w 8.4 Peru for 30 s, and then broke, thereby resulting in the third largest earthquake worldwide since the 1960s. An important question is how much of a subducting oceanic plate bathymetric feature remains intact after subduction and how it affects earthquake rupture on the subduction plane. Recent high quality bathymetric and seismic surveys from the Middle America trench, for example, shows both the scars on the hanging wall associated with the subduction of a seamount, as well as large, clear, subducted seamounts after subduction. This paper discusses the rupture histories of four great subduction earthquakes in the Indian and Pacifi c oceans and examines the relationships between these histories and bathymetric features on the subducting oceanic plate.

Fine details of the Wadati‐Benioff zone under Indonesia and its geodynamic implications

Using about 5200 handpicked P, S, pP, sP, PcP, and ScP phases from digitally recorded seismograms, together with International Seismological Centre reported phases, we obtain improved hypocentral locations for ∼1790 earthquakes, ∼983 of them having 90% confidence limits 50 km, in the period 1962 to September 1996, along the Indonesian subduction zone. We use an improved joint hypocenter determination method in which the solution is more stable than in the original one. We show that for deep earthquakes, use of the core-reflected phases is as good as use of depth phases, and since such core phases often have larger amplitudes and sharper onsets than the depth phases, their arrival times can be read very accurately. The positions of the relocated hypocenters show that (1) a portion of the Indonesian arc between ∼110°E and 123°E longitude, and deeper than ∼500 km, is dipping southward at an angle of ∼75°, that is, in a direction opposite to the upper part of the north dipping slab, and (2) east of about 108°E, the seismic zone is wider near 670 km than near the 500 km depth. The first suggests southward lateral flow in the mantle, relative to the plate motion vector here. From the contortion of the seismic zone along the eastern portion of the Indonesian arc, we can estimate the average lateral shear strain rate in the 300 to 670 km depth range to be of the order of 10−16s−1, over the last 10–20 Myr.

The 1987–1992 Gulf of Alaska earthquakes

Abstract We present a study of the 1987–1992 Gulf of Alaska earthquake sequence using relocated seismicity data together with body wave analysis of selected larger events. Most of the sequence is located on a N-S-trending fault, directly south of the Yakataga seismic gap and consists of four main events of Mw = 7.2, 7.8, 7.7 and 6.8 and associated aftershocks. The first earthquake is of left-lateral strike-slip type on an ENE-WSW-trending fault. The second event is a right-lateral strike-slip earthquake on the N-S-trending fault. The fault plane of the third strike-slip event was identified as an ENE-WSW-trending fault, located to the south of that for the first event, in a previous study using body wave analysis. We show that though the body wave study cannot unambiguously identify the fault plane, the temporal development of the seismicity together with the pattern of aftershock distribution on the conjugate fault suggests that this event also occurred on the N-S-trending fault and is of right-lateral type. The seismicity on the conjugate fault is interpreted as being triggered by the increase of the shear stress in the direction of the normal to the fault plane due to the main shock. The occurrence of the fourth main shock, a right-lateral strike-slip event in 1992, which itself can be considered an aftershock of the 1987–1988 sequence with epicentres distributed along a N-S-trending fault, favours this conclusion. The first three events of the sequence have been described in earlier studies as having too short a rupture length for their seismic moment. If the rupture lengths inferred from the aftershock zones in this study are used we find that this is not the case. The events in the 1987–1992 sequence lie along magnetic anomaly 13 and related conjugate fracture zones, indicating that the oceanic crust is rupturing along pre-existing zones of weakness in response to plate boundary stresses. At the southern end, the seismic activity terminates near a seamount, which rises 1 km above an otherwise relatively low relief ocean floor, and close to a fracture zone perpendicular to the N-S trend of the earthquake sequence, suggesting that the N-S-striking rupture along the N-S-trending fault is terminated here by a barrier. At the northern end, the sequence terminates at the Alaska trench near the Yakataga seismic gap, where there is a right-lateral offset in the bathymetry along the Transition fault, this offset also being visible on the gravity map of the region. Thus, this offset has the same right-lateral sense as the N-S-trending strike-slip fault planes of three of the main shocks and suggests that though there was no significant seismic activity in this region of the Gulf of Alaska since 1964, faulting here has been active in the past.

Stability of earthquake moment tensor inversions: effect of the double-couple constraint

Abstract We show that spurious large non-double-couple components can be obtained in inversions for the full deviatoric moment tensor for shallow crustal earthquakes due to inaccurate Earth models. The traditional “best double-couple” solution does not in general provide an optimal estimate of a double-couple mechanism, and is only reliable when the non-double-couple component of the full deviatoric solution is small. The inverse problem for the moment tensors of the 1998 Antarctic Plate and 2000 Wharton Basin strike-slip earthquakes is shown in each case to have two well-fitting minima in the misfit function of pure double-couple solutions. Such pairs of solutions are most likely to exist for earthquakes which are close either to vertical strike-slip or to dip-slip on a fault plane dipping at 45°. It is shown theoretically that these pairs of solutions arise from the combination of the pure double-couple constraint and the instability of two elements of the moment tensor. No significant non-double-couple component is found for the shallow thrusting 1996 Biak, Indonesia earthquake.