Sumatra-Andaman Islands Earthquake Research Articles

Quantifying 10 Years of Improved Earthquake‐Monitoring Performance in the Caribbean Region

ABSTRACT Over 75 tsunamis have been documented in the Caribbean and adjacent regions during the past 500 years. Since 1500, at least 4484 people are reported to have perished in these killer waves. Hundreds of thousands are currently threatened along the Caribbean coastlines. Were a great tsunamigenic earthquake to occur in the Caribbean region today, the effects would potentially be catastrophic due to an increasingly vulnerable region that has seen significant population increases in the past 40–50 years and currently hosts an estimated 500,000 daily beach visitors from North America and Europe, a majority of whom are not likely aware of tsunami and earthquake hazards. Following the magnitude 9.1 Sumatra–Andaman Islands earthquake of 26 December 2004, the United Nations Educational, Scientific and Cultural Organization (UNESCO) Intergovernmental Coordination Group (ICG) for the Tsunami and other Coastal Hazards Early Warning System for the Caribbean and Adjacent Regions (CARIBE‐EWS) was established and developed minimum performance standards for the detection and analysis of earthquakes. In this study, we model earthquake‐magnitude detection threshold and P ‐wave detection time and demonstrate that the requirements established by the UNESCO ICG CARIBE‐EWS are met with 100% of the network operating. We demonstrate that earthquake‐monitoring performance in the Caribbean Sea region has improved significantly in the past decade as the number of real‐time seismic stations available to the National Oceanic and Atmospheric Administration tsunami warning centers have increased. We also identify weaknesses in the current international network and provide guidance for selecting the optimal distribution of seismic stations contributed from existing real‐time broadband national networks in the region.

Searching for Records of Past Earthquakes Under Water

European Science Foundation Research Conference: Submarine Paleoseismology—The Offshore Search of Large Holocene Earthquakes; Obergurgl, Austria, 11–16 September 2010; The 2004 Sumatra–Andaman Islands earthquake and tsunami highlighted the urgent need to better understand the long‐term history of underwater earthquakes and tsunamis and to better define hazards for the significant populations in coastal areas. As a result, a European Science Foundation Research Conference was convened to review research in the rapidly expanding field of marine and lake paleoseismology. The 4‐day‐long meeting brought together scientists engaged in interdisciplinary research from paleoseismology and marine science. Sixty‐six participants from 21 countries attended the conference, which took place at the Innsbruck University Centre, in Austria. The conference consisted of three main parts: (1) new techniques and results for on‐fault and off‐fault investigations, which summarized the most advanced marine methods for mapping, imaging, and dating underwater paleoseismic events; (2) case studies from different active tectonic continental margins, such as subduction zones, and from lakes or coastal environments that monitor tectonic systems; and (3) the contributions of marine and lake paleoseismology to seismic hazard assessment.

Free and forced polar motion and modern observations of the Chandler wobble

In the absence of excitation, the Chandler wobble is closely a prograde motion along a circular arc. For a step excitation, the centre of the arc shifts, giving a secular motion but an equal and nearly opposite contribution to the Chandler wobble occurs, giving only a second order discontinuity in the pole path. To detect excitation events, we fit circular arcs by least squares to the unequally spaced data, weighting by the inverse of the square of the accompanying standard errors. A break is determined if extrapolation along the circular arc leads to a forecast pole position for which the next measured position lies outside a circle of acceptance. We find that often for quite long periods of time, there seems to be relatively little continuous excitation, leading to the conclusion that much of the excitation comes from sudden events. In particular, we are encouraged that a break in the pole path was found 11 days before the December 26, 2004 Sumatra-Andaman Islands earthquake ( M = 9.1 ).

Storage in confined aquifer: Spectral analysis of groundwater responses to seismic Rayleigh waves

Summary It is frequently found that the fluctuations observed in aquifer-well systems demonstrate high coherence to nearby boundary disturbances, or more specific, in response to the vertical motion of seismic Rayleigh waves. This study derives the spectral relationship between groundwater head and vertical displacement disturbed by Rayleigh waves in the time-frequency domain. Essentially, the groundwater storage property, e.g. storage coefficient or specific storage, representing the amount of water released from storage when the head changes from water expansion and aquifer compression occur. They can be determined using the spectral representation of the factors associated to groundwater head and Rayleigh waves. For this study, the Sumatra-Andaman Islands Earthquake with magnitude 9 at UT 2004/12/26 00:58:53.45 is used as the seismic source while a groundwater monitoring well and a seismological station located on the east of Taiwan acted as the receivers. The time series data associated with the observed groundwater head and vertical displacement of Rayleigh waves are used to evaluate spectral response analyzing the autospectral density, cross-spectral density, and coherence between them. The resultant spectral estimates can be used to determine the storage coefficient and specific storage of the case site to be on the order of 10−3 and 10−4 (m−1), respectively. In addition, the matrix bulk modulus also demonstrated a linear trend with porosity by taking the value of standard water modulus. It suggests that the storage of an aquifer can be robustly estimated using spectral representations and responses from head changes in groundwater and vertical displacement of Rayleigh waves.

Global earthquake casualties due to secondary effects: a quantitative analysis for improving rapid loss analyses

This study presents a quantitative and geospatial description of global losses due to earthquake-induced secondary effects, including landslide, liquefaction, tsunami, and fire for events during the past 40 years. These processes are of great importance to the US Geological Survey’s (USGS) Prompt Assessment of Global Earthquakes for Response (PAGER) system, which is currently being developed to deliver rapid earthquake impact and loss assessments following large/significant global earthquakes. An important question is how dominant are losses due to secondary effects (and under what conditions, and in which regions)? Thus, which of these effects should receive higher priority research efforts in order to enhance PAGER’s overall assessment of earthquakes losses and alerting for the likelihood of secondary impacts? We find that while 21.5% of fatal earthquakes have deaths due to secondary (non-shaking) causes, only rarely are secondary effects the main cause of fatalities. The recent 2004 Great Sumatra–Andaman Islands earthquake is a notable exception, with extraordinary losses due to tsunami. The potential for secondary hazards varies greatly, and systematically, due to regional geologic and geomorphic conditions. Based on our findings, we have built country-specific disclaimers for PAGER that address potential for each hazard (Earle et al., Proceedings of the 14th World Conference of the Earthquake Engineering, Beijing, China, 2008). We will now focus on ways to model casualties from secondary effects based on their relative importance as well as their general predictability.

Displacement analysis of the GPS station of Sampali, Indonesia

Abstract The displacement of the SAMP GPS station located in Medan City, Indonesia, is analyzed by means of an on-line point positioning method, the Canadian Spatial Reference System-Precise Point Positioning (CSRS-PPP). Based on the comparison of the results obtained with those from previous studies, we propose that CSRS-PPP can be applied to analyses of the displacement of a GPS station. Previous studies have focused solely on the “Sumatra-Andaman Islands Earthquake of December 26, 2004”; in contrast, we provide here an in-depth analysis of the crustal movements at the SAMP station for an expanded period of 2.5 years. CSRS-PPP, an Internet data processing service of the Department of Natural Resources Canada (NRCan), was used to process the data obtained at the SAMP station from January 2004 to July 2006. The data show a clear displacement in the southwestern direction from December 26, 2004 to March 28, 2005 when two major earthquakes occurred. However, after the midpoint of 2005, the data show displacement at a regular speed. In particular, the “Sumatra-Andaman Islands Earthquake (M w = 9.0) of December 26, 2004” led to a displacement of 0.1387 m (dn = −0.0122 m, de = −0.1382 m) to the southwest. The earthquake (M w = 8.7) that occurred on March 28, 2005 led to a displacement of 0.1921 m (dn = −0.1400 m, de = −0.1315 m) to the southwest. Starting from December 26, 2004, displacement to the southwest continued. From April 2005, however, the speed of the displacement gradually slowed down. The dn variation shows a displacement at a regular rate (−55.69 mm/year) from April 28, 2005 to July 2006, while the de variation shows a displacement at a regular rate (−23.66 mm/year) from July 5, 2005 to July 2006.

Earthquake Trend Around Sumatra Indicated by a New Implementation of LURR Method

The current implementation of LURR (Load/Unload Response Ratio) theory has shown promise in intermediate to short-term earthquake forecasting through past practice, but has also met difficulties at the same time. In this paper a new implementation of LURR method is presented to address one of the problems. The major change in the new implementation is that LURR values will result from a calculated maximum faulting orientation (MFO) instead of any focal mechanism selected by users. After comparison with the world stress map, the calculated MFO has been found to be in good agreement with the observation from the regional tectonic stress regime. The MFO pattern in the Indonesia region has a special feature which may be related to the unique subduction complexity. The MFO pattern in the Sumatra region in general is different from that in the Java region after the 2004 M 9.0 Sumatra Andaman Islands earthquake. This phenomenon may be supported by the evidence of the recent observation that a section in the southern part of the Sumatran arc is locked. Furthermore, the MFO pattern before the 2004 main shock is different from that after the event. Retrospective examination of the Indonesia region by means of this implementation can produce significant LURR anomaly not only prior to the 2004 main shock but also before the 2006 M 7.7 South Java earthquake. Therefore future great earthquakes might favorably be forecasted if the LURR anomaly detected by MFO method could be considered a precursor.

Coupling Between Seismic Activity and Hydrogeochemistry at the Shillong Plateau, Northeastern India

Transient hydrogeochemical anomalies were detected in a granite-hosted aquifer, which is located at a depth of 110 m, north of the Shillong Plateau, Assam, India, where groundwater chemistry is mainly buffered by feldspar alteration to kaolinite. Their onsets preceded moderate earthquakes on December 9, 2004 (MW = 5.3) and February 15, 2005 (MW = 5.0), respectively, 206 and 213 km from the aquifer. The ratios [Na+K]/Si, Na/K and [Na+K]/Ca, conductivity, alkalinity and chloride concentration began increasing 3–5 weeks before the MW = 5.3 earthquake. By comparison with field, experimental and theoretical studies, we interpret a transient switchover between source aquifers, which induced an influx of groundwater from a second aquifer, where groundwater chemistry was dominantly buffered by the alteration of feldspar to smectite. This could have occurred in response to fracturing of a hydrological barrier. The ratio Ba/Sr began decreasing 3–6 days before the MW = 5.0 earthquake. We interpret a transient switchover to anorthite dissolution caused by exposure of fresh plagioclase to groundwater interaction. This could have been induced by microfracturing, locally within the main aquifer. By comparison with experimental studies of feldspar dissolution, we interpret that hydrogeochemical recovery was facilitated by groundwater interaction and clay mineralization, which could have been coupled with fracture sealing. The coincidence in timing of these two hydrogeochemical events with the only two MW ≥ 5 earthquakes in the study area argues in favor of cause-and-effect seismic-hydrogeochemical coupling. However, reasons for ambiguity include the lack of similar hydrogeochemical anomalies coupled with smaller seismic events near the monitoring station, the >200 km length scale of inferred seismic-hydrogeochemical coupling, and the potential for far-field effects related to the Great Sumatra–Andaman Islands Earthquake of December 26, 2004.

Size and duration of the high-frequency radiator in the source of the 2004 December 26 Sumatra earthquake

SUMMARY We recover the gross space–time characteristics of high-frequency (HF) radiator of the great Sumatra-Andaman islands earthquake of 2004 December 26 (Mw= 9.1–9.3) using the time histories of the power of radiated HF P waves. To determine these time histories we process teleseismic P waves at 36 BB stations, using, in sequence: (1) bandpass filtering (four bands: 0.4–1.2, 1.2–2, 2–3 and 3–4 Hz); (2) squaring wave amplitudes, making ‘power signals’ for each band and (3) stripping the propagation-related distortion (P coda, etc.) from the power signal and thus recovering source time function for HF power. In step (3) we employ an inverse filter constructed from an empirical Green's function, which is estimated as the power signal from an aftershock. For each ray we thus obtain signals with relatively well-defined end and no coda. From these signals we extract: total duration (joint estimate for all four bands) and temporal centroid of signal power for each band. Through linear inversion, the set of duration values for a set of rays delivers estimates of the rupture stopping point and stopping time. Similarly, the set of temporal centroids can be inverted to obtain the position of the space–time centroid of HF energy radiator. The quality of inversion for centroid is acceptable for lower-frequency bands but deteriorates for higher-frequency bands where only a fraction of stations provide useful data. For the source length and duration the following joint estimates were obtained: 1241 ± 224 km, 550 ± 10 s. The estimated stopping point position corresponds to the northern extremity of the aftershock zone. Spatial HF radiation centroids are located at distances 350–700 km from the epicentre, in a systematic way: the higher is the frequency, the farther is the centroid from the epicentre. Average rupture propagation velocity is estimated as 2.25 km s–1.

Effect of the M 9.3 Sumatra–Andaman islands earthquake of 26 December 2004 at several permanent and campaign GPS stations in the Indian continent

The effect of the 26 December 2004 Sumatra–Andaman earthquake on the Indian continent has been estimated from the analysis of GPS data from permanent and campaign GPS sites in the Indian continent. Co‐seismic displacements at these sites have been determined for 11 permanent GPS stations of the national network, five campaign sites in southern India, and four campaign sites in Andaman and Nicobar Islands. The results indicate co‐seismic eastward displacements of 12–20 mm in southern India almost directly west of Andaman, 1.8–6 mm in Central India and insignificant displacement in the Himalayas. Permanent sites in north‐east India which lie almost towards the northward extension of the rupture plane show smaller co‐seismic displacements ranging from 5 to 10 mm southward. Four campaign sites in the Andaman and Nicobar Islands show large horizontal co‐seismic displacements of 1.6–6.49 m WSW and SW. Vertical displacement varies from an uplift of 0.6 m in north Andaman to 1.1 m subsidence at Car Nicobar. The observed GPS displacements are modelled using coulomb 2.6, and the slip on the four segments of the rupture plane (450 km×175 km; 250 km×140 km; 250 km×100 km; 150 km×100 km) that best fits both the far and near field displacements is estimated to be predominantly 12 m reverse in the southernmost segment, which slowly translates to an oblique slip of 7 m in the northernmost segment of the rupture plane. The seismic moment of the Sumatra–Andaman earthquake for the above rupture plane and slip is M o = 5.21×1022 Nm, which corresponds to a moment magnitude of M w = 9.1.

Seismicity Associated with the Sumatra–Andaman Islands Earthquake of 26 December 2004

AbstractThe U.S. Geological Survey/National Earthquake Information Center (usgs/neic) had computed origins for 5000 earthquakes in the Sumatra–Andaman Islands region in the first 36 weeks after the Sumatra–Andaman Islands mainshock of 26 December 2004. The cataloging of earthquakes of mb (usgs) 5.1 and larger is essentially complete for the time period except for the first half-day following the 26 December mainshock, a period of about two hours following the Nias earthquake of 28 March 2005, and occasionally during the Andaman Sea swarm of 26–30 January 2005. Moderate and larger (mb ≥5.5) aftershocks are absent from most of the deep interplate thrust faults of the segments of the Sumatra–Andaman Islands subduction zone on which the 26 December mainshock occurred, which probably reflects nearly complete release of elastic strain on the seismogenic interplate-thrust during the mainshock. An exceptional thrust-fault source offshore of Banda Aceh may represent a segment of the interplate thrust that was bypassed during the mainshock. The 26 December mainshock triggered a high level of aftershock activity near the axis of the Sunda trench and the leading edge of the overthrust Burma plate. Much near-trench activity is intraplate activity within the subducting plate, but some shallow-focus, near-trench, reverse-fault earthquakes may represent an unusual seismogenic release of interplate compressional stress near the tip of the overriding plate. The interplate-thrust Nias earthquake of 28 March 2005, in contrast to the 26 December aftershock sequence, was followed by many interplate-thrust aftershocks along the length of its inferred rupture zone.

Review of the source characteristics of the Great Sumatra–Andaman Islands earthquake of 2004

The December 26, 2004 Sumatra–Andaman Island earthquake, which ruptured the Sunda Trench subduction zone, is one of the three largest earthquakes to occur since global monitoring began in the 1890s. Its seismic moment was M 0 = 1.00 × 1023–1.15 × 1023 Nm, corresponding to a moment-magnitude of M w = 9.3. The rupture propagated from south to north, with the southerly part of fault rupturing at a speed of 2.8 km/s. Rupture propagation appears to have slowed in the northern section, possibly to ∼2.1 km/s, although published estimates have considerable scatter. The average slip is ∼5 m along a shallowly dipping (8°), N31°W striking thrust fault. The majority of slip and moment release appears to have been concentrated in the southern part of the rupture zone, where slip locally exceeded 30 m. Stress loading from this earthquake caused the section of the plate boundary immediately to the south to rupture in a second, somewhat smaller earthquake. This second earthquake occurred on March 28, 2005 and had a moment-magnitude of M w = 8.5.

A new method for the computation of global viscoelastic post-seismic deformation in a realistic earth model (I)-vertical displacement and gravity variation

We introduce a new method by which to compute global post-seismic deformation (PSD) in a spherically symmetric, self-gravitating viscoelastic earth model. Previous methods are based on simplified earth models that neglect compressibility and/or the continuous variation of the radial structure of Earth. This is because the previous mode summation technique cannot avoid intrinsic numerical difficulties caused by the innumerable poles that appear in a realistic earth model that considers such effects. In contrast, the proposed method enables both of these effects to be taken into account simultaneously. We carry out numerical inverse Laplace integration, which allows evaluation of the contribution from all of the innumerable modes of the realistic earth model. Using this method, a complete set of Green's functions is obtained, including functions of the time variation of the displacement, gravity change, and the geoid height change at the surface for strike-slip, dip-slip, horizontal and vertical tensile point dislocations. As an earth model, we employ the preliminary reference earth model (PREM) and a convex viscosity profile. Further, we investigate the effects of fine layering of the viscoelastic structure and compressibility on a time-series of PSD using the Green's function for a dip-slip fault. The result indicates that the effect of increasing number of layers is saturated at several tens of layers even when compressibility is taken into account and that the effect of compressibility is detectable with modern observational techniques for a shallower large earthquake (Mw∼ 8). As an application, the PSD due to the Sumatra-Andaman Islands earthquake (Mw= 9.3) is estimated. We show that the rate of post-seismic vertical displacement and gravity change is possibly detected in the far field where the epicentral distance exceeds 400 km.

The very-broad-band long-base tiltmeters of Grotta Gigante (Trieste, Italy): Secular term tilting and the great Sumatra-Andaman islands earthquake of December 26, 2004

The horizontal pendulums of the Grotta Gigante (Giant Cave) in the Trieste Karst, are long-base tiltmeters with Zöllner type suspension. The instruments have been continuously recording tilt and shear in the Grotta Gigante since the date of their installation by Prof. Antonio Marussi in 1966. Their setup has been completely overhauled several times since installation, restricting the interruptions of the measurements though to a minimum. The continuous recordings, apart from some interruptions, cover thus almost 40 years of measurements, producing a very noticeable long-term tiltmeter record of crustal deformation. The original recording system, still in function, was photographic with a mechanical timing and paper-advancing system, which has never given any problems at all, as it is very stable and not vulnerable by external factors as high humidity, problems in power supply, lightning or similar. In December 2003 a new recording system was installed, based on a solid-state acquisition system intercepting a laser light reflected from a mirror mounted on the horizontal pendulum beam. The sampling rate is 30 Hz, which turns the long-base instrument to a very-broad-band tiltmeter, apt to record the tilt signal on a broad-band of frequencies, ranging from secular deformation rate through the earth tides to seismic waves. Here we describe the acquisition system and present two endline members of the instrumental observation, the up to date long-term recording, and the observation of the great Sumatra-Andaman Islands earthquake of December 26, 2004, seismic moment magnitude M w = 9.1–9.3 [Lay, T., Kanamori, H., Ammon, C.J., Nettles, M., Ward, S.N., Aster, R.C., Beck, S.L., Bilek, S.L., Brudzinski, M.L., Butler, R., DeShon, H.R., Ekström, G., Satake, K., Sipkin, S., 2005. The Great Sumatra-Andaman Earthquake of 26 December 2004. Science. 308, 1127–1133.]. The secular-term observations indicate an average tilting over the last four decades towards NW of 23.4 nrad/year. We find evidences that this tilting is regional and has been going on since at least 125 ka. The recent earthquake of December 26, 2004 was well recorded by the pendulums. We show that the free oscillation modes were activated, including the lowest modes as e.g. 0 T 2 , 0 T 3 , 0 T 4 , 0 T 5 and 2 S 1 , 0 S 3 , 0 S 4 , 1 S 2 .

A Possible Detection of the 26 December 2004 Great Sumatra-Andaman Islands Earthquake with Solution Products of the International GNSS Service

The main goal of this work is to critically review the IGS solution products and Precise Point Positioning (PPP) in order to demonstrate their potential to contribute to studies of large earthquakes such as the one that devastated Southeast Asia on December 26th, 2004. In view of a possible detection of the Mw 9.0 Sumatra-Andaman Islands Earthquake of December 26, 2004, position solutions, ranging from intervals of years to one second, of four International GNSS Service (IGS) stations within 3000 km of the epicenter were examined. The IGS combined, cumulative solution product (IGS04P51), consisting of epoch and station velocity solutions and based on data spans of several years prior to the earthquake, was used as a reference. Four IGS combined weekly position solutions (igs04P1301-4), two weeks before and after the earthquake, were utilized for the weekly solution resolution. PPP static and kinematic solutions with IGS Final combined orbits and clocks were used for the mean daily and instantaneous 5-min and 1-sec epoch solutions, respectively. The most significant changes, detected by both weekly and daily solutions occurred in longitude. The nearest IGS station ntus, about 1000 km east of the epicenter, moved westward about 15 mm, while the more distant Indian station iisc (∼ 2300 km NW from the epicenter), shifted about 15 mm eastward. In spite of position errors caused by interpolation of the 5-min IGS clocks, the 1-sec solutions, based on separate data sets, available only for two stations (iisc, dgar), still showed seismic surface waves, in particular at the Indian station iisc. Precise daily IGS combined polar motion and length-of-day products, after correcting for the atmospheric effects, also likely detected, statistically significant, anomalistic excitations on December 26, 2004 that could be caused by this great earthquake.

Breaking Omori's Law of Public Awareness

As seismologists dedicated to studying one of the Earth's most dangerous natural phenomena, we are inured to the reality that public interest in our field surges after a significant earthquake and decays rapidly afterward to a rather low background level. This would be little more than a curiosity were it not for the fact that understanding earthquakes and their effects could make a big difference to the lives and livelihoods of much of that public. Moreover, the political will to make pre-emptive investments in monitoring, research, preparedness, and mitigation wanes with decreasing public interest. Ironically, we can point with increasing confidence to areas of habitation and economic production most exposed to earthquakes and other natural hazards, and we can even treat the issue in economic terms, evaluating risk reduction as an investment. What will it take to change the disaster management culture from one that relies on emergency response to one that promotes mitigation? Will sustained interest in natural hazard reduction accelerate the translation of research outcomes into practice? Will political interest continue to follow public awareness? Justified as they may still be, simply asking these old rhetorical questions is not enough. The Sumatra-Andaman Islands earthquake and Indian Ocean tsunami struck deeply into the heart of some of the central questions of our time: the gap between rich and poor, between those who benefit from advanced understanding of our natural world and those who haven't yet seen the impact of knowledge, between those who contemplate the relationship between man and nature and those who ignore it. This time may be different, and the science enterprise has the responsibility to recognize and act on the evidence …