S51A-01
Focal Mechanism of a Catastrophic Earthquake of the Last Rococo Period (1783) in Southern Italy Retrieved by Inverting Historical Information on Damage
Using geophysical inversion to discover the fault source of a blind earthquake, that took place before the invention
of the seismograph, seemed impossible. We demonstrated that sometimes it is possible using our simplified
KF model (Sirovich, 1996) through automatic genetic inversion (Gentile et al., 2004 in BSSA; Sirovich and
Pettenati, 2004 in JGR), and determined it conclusively by treating the Coalinga 1983, Loma Prieta 1989, and
Northridge 1994 earthquakes (Pettenati and Sirovich, 2007 in BSSA). KF is able to simulate the body-wave
radiation from a linear source, and eleven source parameters are retrieved: the three nucleation coordinates, the
fault-plane solution, the seismic moment, the rupture velocities and lengths along-strike and anti-strike, the shear
wave velocity in the half-space. To find the minima on the hypersurface of the residuals in the multi-parameter
model space, we use a genetic process with niching since we have already shown that the problem is bimodal
for pure dip-slip mechanisms. The objective function of the nonlinear inversion is the sum of the squared
residuals (calculated-minus-observed intensity at all sites). Here, we use the very good intensity data provided in
the MCS scale by the INGV of Italy for the M 6.9 earthquake of Feb. 5, 1783 (see the Italian intensity data bank on
http:emidius.mi.ingv.it/DOM/consultazione.html). The data of 1783 were created by seismologists and historians
who interpreted the reports of the time and many other historical sources. Given the limitations of the KF
approach, we limited our inversion to a square area of 200 by 200 km around the most heavily damaged zone.
341 surveyed towns and hamlets received intensity degrees by INGV (we discarded 6 of them as statistical
outliers according to the classical Chauvenet method). Thus, 335 data were inverted. The match between
experimental and synthetic isoseismals is really noteworthy. The found mechanism is almost pure dip-slip and,
thus, the problem is bimodal. In fact, two source models score almost the same objective function, and they
coincide with the auxiliary planes of the same solution. The best source is: nucleation latitude [degrees] 38.28,
longitude 15.95, depth 13.1 km; strike angle 210, dip 31, rake 269(±180); seismic moment 2.7 * 10e19 Nm,
3.1 km/s rupture velocity along-strike (3.3 anti-strike), 22.3 km rupture length along-strike (20.7 anti-strike), 3.9
km/s shear wave velocity in the half-space. The second solution, having the same nucleation coordinates and a
one-point difference in the objective function (117 against 116) is: strike angle 30, dip 60, rake 270(±180);
seismic moment 2.6 * 10e19 Nm, 3.3 km/s rupture velocity along-strike (3.0 anti-strike), 19.4 km rupture length
along-strike (21.9 anti-strike), 3.9 km/s shear wave velocity in the half-space. Note the symmetry with the former
solution. In other words, we find one fault source having a low-angle dip toward the Tyrrhenian Sea; its virtual
intersection with the topographical surface should be found close to the Jonian coast. On the contrary, its
symmetric solution strikes SW, has a high-angle dip toward the Jonian Sea, and the fault should outcrop close to
the Tyrrhenian coast. Both sources are compatible with the orientation of the principal tectonic structures in the
area, however. The final choice between them will hopefully come from tectonic interpretation. This kind of study
seems especially promising for southern Europe where a lot of documents on seismic damage caused by old
earthquakes exist.
http:emidius.mi.ingv.it/DOM/consultazione.html
S51A-02
Temporal variation of seismicity in the particular case of Vrancea (Romania) intermediate- depth earthquakes
The aim of the study is to investigate the temporal behavior of some common seismicity parameters and define seismicity pattern for the Vrancea seismogenic zone, located at the South-Eastern Carpathians arc bend, in Romania. A persistent and unusually strong cluster of earthquakes is generated here in a confined volume at intermediate depths, between 60 and 180 km. The atypical geometrical configuration of the hypocenters, elongated along NE-SW direction and close to a planar distribution, the persistence of the earthquake generation in time (around 15 events/month with M greater than 3 and around 3 events/century with M greater than 7), the predominance of the focal mechanism, rise a lot of questions and debates in connection with this seismic area. We use a representative and homogeneous catalog of Vrancea intermediate-depth events recorded by the seismic network of National Institute for Earth Physics between 1994 and 2006 and a routine catalog for a 70-year time interval. The refined catalog is complete for duration magnitudes above 3 and contains only small and moderate earthquakes (Mw 5.8 was the largest observed magnitude). A JHD technique applied for hypocenters determination allows high accuracy locations and detailed and complex analyses related to seismicity evolution in time and space. We analyze the space and time variation of the seismic activity, frequency-magnitude distribution, fractal dimension, deformation accumulation curve (Benioff's curve), on different time and space windows and test any possible earthquake precursor. The study area, situated between 60 and 180 km depth, is divided in two active segments, one centered around 90 km depth, other centered around 140 km depth. The particular configuration of the foci along NE-SW direction allows a 2D approach. The evolution of seismic activity shows alternative accelerating and decelerating deformation release in the upper segment and lower segment, respectively, of the subducting lithosphere. The different seismicity behavior in the two segments of the seismic active volume and the apparent interconnection between them can be speculated to predict the most probable future particular seismic hazard pattern. The largest events are the most infrequent, but the most important to understand, since they control the evolution of the system and are the most destructive events. Our detailed pattern analysis suggests that recognizable patterns of smaller, more frequent events can be used to detect the generation of the next major event although reliable and efficient forecasting of the largest events is still questionable. We identify the characteristics of the preparation process of the strong subcrustal events originating in Vrancea region and analyze how they can be incorporated in a time-dependent seismic hazard assessment.
S51A-03
Seismicity of the Subducted Caribbean Plate in Panama
The Panama microplate is bounded by convergent boundaries to the north (North Panama deformed belt; NPDB) and south (South Panama deformed belt; SPDB and Middle American Trench; MAT), suturing with South America along the Atrato valley to the east and the Central Costa Rica deformed belt (CCRDB) to the west. GPS measured convergence rates between the NPDB and the Caribbean plate are to the southwest at 7 mm/a. Seismicity of the portion of the NPDB along the isthmus of Panama is poorly understood, even though the largest recorded earthquake in Panama (7 September, 1882 (M7.9)) occurred on this boundary. This earthquake caused extensive damage in central Panama and produced a tsunami that killed 75 people in Comarca de San Blas. A preliminary study of seismicity in the area, using high quality seismic data recorded by local stations between 1996-2006, suggests that most events were of intermediate depth (40-70km) with a geometry indicative of a subducted plate. We also note that focal mechanisms show a complex stress field in this area possibly due to activation of pre-existing faults and internal deformation of the Caribbean plate being subducted. The existence of a well-defined Wadati-Bennioff zone below northern Panama should be included in future risk assessment evaluations.
S51A-04
The 2004 Sumatra Earthquake Mw 9.3: Seismological and Geophysical Investigations in the Andaman-Nicobar Islands
The December 26, 2004 Sumatra-Andaman earthquake (MW 9.3) is the fourth largest event (M>9.0) in the world during the last 100 years. It occurred by thrust faulting on the interplate thrust zone of the subducting India plate and overriding Burma platelet. The main shock rupture, ~1300 km long and ~200 km wide, propagated from north of Sumatra to Andaman - Nicobar Islands; the slow rupture generated Tsunami which killed about 300,000 people. The epicenter of the earthquake is located at 3.90N and 94.260E with a focal depth at 28 km (USGS). This mega seismic event triggered giant tsunamis that devastated the coastal regions of Indonesia, Malaysia, Thailand, Sri Lanka, India, Maldives and even the east coast of Africa. The impact of the tsunami was quite severe in India, in the coasts of Andaman and Nicobar Islands. The Air-base in the Car- Nicobar island was totally devastated by the tsunami and killed about 200 people. Macroseismic survey was carried out by different teams of GSI in North Andaman, Middle Andaman, South Andaman, Havelock Hut Bay and also in the Nicobar Islands. A maximum intensity VIII was recorded in the Andaman Islands. The mega thrust event was followed by an intense aftershock activity spreading over an area extending between 30-140N along the Andaman - Nicobar - Sumatra Island arc region. The aftershocks are distributed northwards from the epicenter of the main shock suggesting a unilateral rupture propagation. The aftershock (M >4.5) area covers a length of about 1300 km and a width of about 200 km, in a 'banana' shape. The national network (IMD) recorded almost all aftershocks M >5.0; about 350 were recorded till 31.01.2005. The Geological Survey of India (GSI) deployed six temporary seismograph stations in the Andaman and Nicobar Islands and also in Havelok and Narkunda (volcanic) islands. About 20,000 aftershocks (M >3.0) were recorded until end of March, 2005. About 1000 aftershocks (M >3.0) located by the GSI network until January 31, 2005 are studied. The aftershocks are still continuing; frequency of occurrence is, however, reduced now. Fault plane solutions suggest predominant thrust faulting in the fore arc region, and normal/strike ship in the back arc region, consistent with the regional tectonics. Crustal deformation study was carried out by various organizations. Pre- and -post earthquake vectors clearly show that islands have moved 2 to 6 meters in horizontal position towards mainland, and also there is anti-clockwise rotation. The GPS stations move southwesterly, 2 to 3 m in the Andaman Islands and 5 to 6 m in the Nicobar islands. Tidal observations indicate that there is a rise in local mean sea level of an order of 1.05 m at the Port Blair observatory. This observation is conformable with the GPS/levelling measurements that show a subsidence of the observatory to an extent of 1.1 m. The uplift and subsidence are explained by the thrust faulting involving reverse slip; uplift at the up dip edge and subsidence at the down dip on the coseismic rupture.
S51A-05
The Intraplate Seismogenic Zone of Porto dos Gauchos in the Amazon Craton, Brazil
The Porto dos Gauchos Seismogenic Zone (PGSZ), in the center north of Mato Grosso State, in the contact between the southern Amazonian Craton and northern Parecis Basin, represents one of the most important area of seismic activity in Brazil, with the largest magnitude ever observed in the stable continent of the South America plate (6.2 mb on January 31, 1955). Focal mechanism studies indicated a pure reverse faulting regime with compressional SHmax. oriented in SE-NW direction. After the 1955 earthquake, located in Serra do Tombador, a recurrent seismicity has been detected in Porto dos Gauchos, 100 km northeast of Serra do Tombador. No recent events have been detected in the area of the 1955 epicenter, suggesting a long recurrence time or mislocation of Serra do Tombador earthquake. The Porto dos Gauchos recurrent seismicity has been observed since 1959, when a 4.5 macroseismic estimated magnitude was felt by local inhabitants settled in that remote area two years earlier. In subsequent years, with deployment of regional stations in Brazilian Amazon region earthquakes were detected in 1981 (3.8 mb), 1989 (3.3 mb), 1993 (3.8 mb), 1996 (4.4 mb), 1997 (3.3 mb), and finally on March 10, 1998 (5.1 mb). The aftershocks of 1998 main shock were studied with a local network with up to eight 3- component stations. Such network detected more than 2500 events until December of 2002, when the network was deactivated, but only 100 were accurately located. Based on this set of events and a controlled source experiment we determined a 1-D velocity model for the area, a composite focal mechanism with P wave polarities, spectral analysis studies to estimate the source dimension, stress drop and moment magnitudes for the main shock and some others events of the set. On March 23, 2005 another shock occurred in the same seismogenic area of Porto dos Gauchos, with magnitude 4.7 to 5.0. One week later five stations were installed again to monitor the aftershock activity, detecting more than 2000 events in three months. Hypocentral determination, with different hypocentral location programs (Hypocenter, Hypo71, Hypoinverse, Velest and NonLinLoc) and a composite focal mechanism indicated the same source location and similar fault regime (dextral transcurrent fault for both 1998 and 2005 sequences). In this work, besides a brief review of the seismotectonic of the PGSZ we intend to present additional results of better constrained composite focal mechanisms including the P, SH, and SV amplitudes ratios for the two sequences and an attempt to get focal mechanism solution for the main shock of March 1998 based on regional stations.
S51A-06
New Seismicity Map for Central Part of Saudi Arabia
Riyadh city and central part of Saudi Arabia are located in the Arabian Plat which is known as a relativity stable platform. However, we have been able to determine some events that were caused by local faults. In addition, we have been able to record some regional Earthquakes that cussed by tectonic movements. Most of these local events are not felt. The Institute of Astronomy and Geophysics Research at King Abdulaziz City for Science and Technology installed nine earthquake portable stations in central part of Saudi Arabia around Riyadh city to record local and micro- seismicty events. We are using digital recorders (RefTek 72A) for data acquisition, and SAISAN, HYPOINVERSE software for analyses. Currently, I am using different types of velocity models, and I am developing a map that shows some micro-earthquake events for that region. In addition, it includes some regional events. The objective of this study is to define the active faults in central part of Saudi Arabia, and this study will not only yield additional information regarding the tectonic setting, but also revised hazard assessments for the region.
S51A-07
Intraplate Earthquake Swarm in Belo Jardim, NE Brazil: Reactivation of a Major Late Proterozoic Shear Zone (Pernambuco Lineament)
Intraplate earthquakes in stable continental areas have been explained basically by reactivation of pre-existing zones of weakness, stress concentration, or both. Zones of weakness are usually identified as sites of the last major orogeny, provinces of recent alkaline intrusions, or stretched crust in ancient rifts. Stress concentrations can occur due to basement heterogeneities (including lithospheric thinning), thermal expansion/contraction, and isostatic/flexural deformation. Stresses from continental/oceanic crustal transition and flexure from sediment load in the continental margin have been shown to be a major factor to explain the observed seismicity along the continental margin in NE Brazil. Although Northeastern Brazil is one of the most seismically active areas in the country (magnitudes 5 roughly every 5 years), with hypocentral depths shallower than ~10 km and seismic zones as long as 30-40km, no clear relationship with the known surface geology can be established with confidence, and a clear identification of zones of weakness has not been possible. The 2004 earthquake swarm of Belo Jardim (magnitudes up to 3.0) and the recurrent activity in the nearby town of Caruaru (magnitudes up to 3.8), both located along the major Late Proterozoic shear zone (known as Pernambuco Lineament), are the first clear indication of seismic activity occurring as reactivation of an old structure. A seismic network of 12 stations, operated in Belo Jardim during 20 days in November 2004, showed that the activity occurred by normal faulting on a North dipping, E-W oriented fault plane. This fault plane is in close agreement with the E-W trending structures within the Pernambuco Shear zone. The Belo Jardim activity was concentrated in a 2 km (E-W) by 2 km (downdip) fault area, with an average depth of 5 km. The nearby Caruaru activity occurs as both strike-slip and normal faulting, also consistent with the Pernambuco Lineament. The focal mechanisms of Belo Jardim and Caruaru indicate E-W compressional and N-S extensional stresses. The orientation of this local stress regime cannot be explained only by forces originated near the continental margin, implying that some additional factors are necessary to understand the causes of the seismic reactivation of the old Pernambuco Lineament
S51A-08
Seismic activity in northeastern Brazill-new perspectives
Northeastern Brazil is the most seismic active region in the country. Some earthquakes with magnitude above 5.0
and intensity VII MM associated with swam-like seismic activity lasting for many years are a serious social
concern. Since the 1980's macroseismic and instrumental surveys have been carried out in this region and they
are an important data archive which allows the composition of a reliable catalogue of seismic activity for this
region. Among the many scientific results it was possible to identify the main seismogenic areas, obtain reliable
hypocentres and focal mechanisms. As a consequence, it was possible also to analyse the relationship between
seismicity and geological features. It was also possible to determined maximum horizontal stress direction for
the region. An important induced seismic activity case has also been reported in the area as being a classical
example of pore pressure diffusion triggering mechanism. The majority of the results were obtained using
analogic data. Recently, a new research project is being conducted and will allow us to provide a regional scale
monitoring with 6 broad-band stations and a new portable six station digital seismic network equipped with short-
period sensors. Thus, with the continuous seismic activity in the area we trust that the results of this project will
increase the present knowledge of seismic activity in northeastern Brazil.
http:www.dfte.ufrn.br
S51A-09
Finite-Fault Analysis Using Regional Broadband Waveforms: The 2004 Parkfield, California Earthquake
Global improvements in seismic-station coverage and data quality have resulted in the increased use of broadband waveforms in finite-fault studies of the earthquake source. These improvements have also prompted the use of regional waveforms in rapid or real-time derivations of the earthquake rupture history. We examine the applicability of regional records to identify the properties of an extended source by conducting a finite-fault analysis of the Mw 6.0 Parkfield, California, earthquake of 28 September 2004. This earthquake was recorded by more than 80 broadband stations at distances less than 5 degrees from the epicenter. We invert three- component displacement waveforms recorded at the 14 stations located within 2 degrees of the epicenter using a general Northern California crustal structure and a bandpass filter with corners of 0.2 and 1.0 Hz. The resulting rupture model shows coseismic slip extending primarily to the northwest with a peak slip of 60 centimeters at a distance of about 15 km from the hypocenter. This model is compared against results obtained using strong motion and geodetic data. We also examine the use of waveforms recorded at more distant sites by comparing them against synthetic waveforms predicted by the inferred slip model and conduct several tests using different stations at differing azimuths to evaluate the effects of station distribution on the finite-fault results. These tests are useful for assessing the limitations of finite-fault inversion schemes in the routine derivation of the earthquake rupture history using real-time regional waveform data.
S51A-10
Prediction of Earthquakes by Lunar Cicles
Prediction of Earthquakes by Lunar Cicles Author ; Guillermo Rodriguez Rodriguez Afiliation Geophysic and Astrophysicist. Retired I have exposed this idea to many meetings of EGS, UGS, IUGG 95, from 80, 82.83,and AGU 2002 Washington and 2003 Niza I have thre aproximition in Time 1º Earthquakes hapen The same day of the years every 18 or 19 years (cicle Saros ) Some times in the same place or anhother very far . In anhother moments of the year , teh cicle can be are ; 14 years, 26 years, 32 years or the multiples o 18.61 years expecial 55, 93, 224, 150 ,300 etcetc. For To know the day in the year 2º Over de cicle o one Lunation ( Days over de date of new moon) The greats Earthquakes hapens with diferents intervals of days in the sucesives lunations (aproximately one month) like we can be see in the grafic enclosed. For to know the day of month 3º Over each day I have find that each 28 day repit aproximately the same hour and minute. The same longitude and the same latitud in all earthquakes , also the littles ones . This is very important because we can to proposse only the precaution of wait it in the street or squares Whenever some times the cicles can be longuers or more littles This is my special way of cientific metode As consecuence of the 1º and 2º principe we can look The correlation between years separated by cicles of the 1º tipe For example 1984 and 2002 0r 2003 and consecutive years include 2007...During 30 years I have look de dates. I am in my subconcense the way but I can not make it in scientific formalisme
S51A-11
Teleseismic Body Wave Analysis for the 27 September 2003 Altai, Earthquake (Mw7.4) and Large Aftershocks
We investigate the kinematics of the rupture process for the September 27, 2003, Mw7.3, Altai earthquake and its associated large aftershocks. This is the largest earthquake striking the Altai mountains within the last 50 years, which provides important constraints on the ongoing tectonics. The fault plane solution obtained by teleseismic body waveform modeling indicated a predominantly strike-slip event (strike=130, dip=75, rake 170), Scalar moment for the main shock ranges from 0.688 to 1.196E+20 N m, a source duration of about 20 to 42 s, and an average centroid depth of 10 km. Source duration would indicate a fault length of about 130 - 270 km. The main shock was followed closely by two aftershocks (Mw5.7, Mw6.4) occurred the same day, another aftershock (Mw6.7) occurred on 1 October , 2003. We also modeled the second aftershock (Mw6.4) to asses geometric similarities during their respective rupture process. This aftershock occurred spatially very close to the mainshock and possesses a similar fault plane solution (strike=128, dip=71, rake=154), and centroid depth (13 km). Several local conditions, such as the crustal model and fault geometry, affect the correct estimation of some source parameters. We perfume a sensitivity evaluation of several parameters, including centroid depth, scalar moment and source duration, based on a point and finite source modeling. The point source approximation results are the departure parameters for the finite source exploration. We evaluate the different reported parameters to discard poor constrained models. In addition, deformation data acquired by InSAR are also included in the analysis.