Search

Abstract Ambient noise cross-correlation has been widely used to observe post-earthquake temporal velocity variations. Comparative studies are essential for assessing seismic hazards and clarifying the relationship between velocity variation and magnitude. However, very few comparative studies by earthquake magnitude have been conducted, particularly for magnitudes smaller than 6. In 2016 and 2017, southeastern Korea experienced the two most damaging earthquakes of the past 250 years: the Pohang earthquake (M 5.4) and Gyeongju earthquake (M 5.8). This study compared the coseismic velocity variations of these earthquakes with those observed in the M 7.3 Kumamoto and M 6.6 Tottori earthquakes, which occurred in Japan in 2016. In this study, the moving window cross-spectral analysis revealed different curves for bidirectional cross-correlations, a difference that has not been addressed in previous studies. This discrepancy was most likely caused by data deficiencies due to prolonged aftershocks or by instability from small-magnitude earthquakes recorded by surface seismometers in Korea. The average velocity decrease between forward and reverse order pairs correlated well with earthquake magnitude, except in the case of the Korean earthquake. Our devised measure based on aftershock activity suggests that the large decrease in the Pohang earthquake was due to non-coseismic noise in addition to instability.

Seismogenic deformations of Cape Shartlay represent a very young fault system on the northwestern coast of Lake Baikal. Their study is providing an important opportunity to measure earthquake magnitudes, to identify areas where earthquakes are more likely to occur, and to estimate the probability of earthquake occurrence as applied to seismically active Baikal region. In this connection, the present work was aimed at characterizing in detail the structure, displacements, and reconstruction of the rupture propagation model. The study is based on photogrammetric processing and interpretation of the unmanned aerial survey data, as well as on morphostructural analysis of the displacement profiles and georadiolocation (GPR) data. It has been found that seismogenic ruptures of Cape Shartlay formed under prevailing extension conditions during no less than two earthquakes with magnitudes Mw≥7.0, Ms≥7.2. Seismic rupture propagation was primarily northward. The main rupture with displacement amplitude of more than 2 m contributed 39 to 93 % to the total surface displacement depending on the amount of dislocations on the transverse profile. It is shown that the length of a certain rupture increased almost instantaneously, then displacements along some of the ruptures stopped. A significant elongation of ruptures is primarily due to their merging. The present-day seismogenic zone is highly permeable. According to the tectonophysical model of formation of inner structure of the fault zone, the development of the seismogenic rupture system of Cape Shartlay corresponds to the late disjunctive stage. This means that the rupturing process in this segment of the North Baikal fault may not have stopped yet, and the lack of large earthquakes in the instrumental record implies the accumulation of stress in its southern part. The obtained results provide an opportunity to reconstruct the development of large fault zones by studying the displacement profiles and, therefore, to localize more precisely the places where future earthquakes may occur.

In Egypt, there were many earthquakes activities in seismo-tectonic regions. In the last century most of earthquakes concentrated at Gulf of Suez and Gulf of Aqqaba. Also, Egypt affected by large earthquakes in the northern Mediterranean Sea. In this study, we are trying to evaluate the crustal displacement associated with the large earthquakes activities by using of GNSS measurements. GNSS observations can measure the horizontal and vertical displacement within mm of accuracy. Earthquakes with magnitude more than 4 have been chosen. The crustal displacements have been computed using GNSS data from Egyptian Permanent Geodetic Network (EPGN). The results show that before and during and after earthquakes activities, the displacement of GNSS stations gives different behavior depending on its distance from earthquakes epicenter. In some studied GNSS station, there was a signal of deformation before the earthquake found. This displacement can used in establishment of early warning system for earthquakes using GNSS observations.

This study discusses the temporal distribution of earthquake magnitudes in the city of Bingöl, near Karlıova Triple Junction. We determine the probability distributions and return periods of earthquakes for all districts of Bingöl. Bingöl has eight districts; namely Adaklı, Central, Genç, Karlıova, Kiğı, Solhan, Yayladere, and Yedisu. In six of them, active faults were mapped previously (Adaklı, Central, Genç, Karlıova, Solhan, and Yedisu). We consider 5 time-dependent probability distributions for analysis. Using the annual maximum earthquake magnitudes, the best fit arises from the Gumbel distribution for Central, Karlıova, and Adaklı Districts. For the Genç District, where the least maximum earthquake magnitude is reported, the Weibull distribution gives the best fit. The return period and maximum annual earthquake magnitude relations suggest the following results. For the Central and Karlıova Districts along which maximum earthquake magnitudes are reported, every 250 years a 6.7 M, and 7.2 M occurs respectively. These results are compatible with the results of paleo-seismological data reported along the NAFZ and the EAFZ. For a 10-year return period, earthquake magnitudes reach 3.9 and 5.1 in all districts. It is important to note that in the Yedisu District, the maximum earthquake magnitudes seem as 5.1 M for the 1000-year return period, incompatible with previously published findings probably because of low quality seismic data in this region.

Abstract The reliable estimation of earthquake magnitude and stress drop are key in seismology. The novel technology of distributed acoustic sensing (DAS) holds great promise for source parameter inversion owing to the measurements' high spatial density. In this study, I demonstrate the robustness of DAS for magnitude and stress drop estimation using the empirical Green's function deconvolution method. This method is applied to nine co‐located earthquakes recorded in Israel following the 2023 Turkey earthquakes. Spectral ratios were stacked along the fiber, and fitted with a relative Boatwright source spectral model. Excellent fits were obtained even for similar sized earthquakes. Stable seismic moments and stress drops were calculated assuming that the moment of one earthquake is known. DAS derived estimates were found to be more stable and reliable than those obtained using a dense accelerometer network. The results demonstrate the great potential of DAS for source studies.

Abstract Large variations in the maximum earthquake magnitude ( $$M_{{\text{max}}}$$ M max ) have been observed among the world’s subduction zones. There is still no universal relationship between $$M_{{\text{max}}}$$ M max and a given subduction-zone parameter, such as plate age, plate dip angle, or plate velocity, which suggests that multiple parameters control $$M_{{\text{max}}}$$ M max . Here, we conduct exhaustive variable selections that are based on three evaluation criteria; leave-one-out cross-validation errors (LOOCVE), Akaike information criterion (AIC), and Bayesian information criterion (BIC) to determine the combination of subduction-zone parameters that best explains $$M_{{\text{max}}}$$ M max . Multiple linear regression analyses are applied using 18 subduction-zone parameters as potential candidates for the explanatory variables of $$M_{{\text{max}}}$$ M max . The minimum BIC is obtained when five variables (trench sediment thickness, existence of an accretionary prism, upper-plate crustal thickness, bending radius of the subducting oceanic plate, and trench depth) are selected as explanatory variables; each variable contributes positively to $$M_{{\text{max}}}$$ M max . Minimum LOOCVE and AIC values are obtained when eight variables (the five parameters for BIC, plus the along-strike plate convergence rate, age of the subducting plate, and maximum depth of the subducting plate) are selected. Our selection of the trench sediment thickness and plate bending radius contributing to $$M_{{\text{max}}}$$ M max is consistent with previous studies. The results show that increasing upper-plate crustal thickness results in a large $$M_{{\text{max}}}$$ M max . In addition to smoothing the subducting-plate interface via subducted sediments, along-dip extension of the crustal area along the convergent plate boundary would be important for generating a large earthquake. Graphic Abstract

We collected a data set of mainshocks and their respective largest foreshocks of 38 earthquake sequences in Taiwan. The plot of local magnitude, ML, of a mainshock (denoted by MLm) versus ML of its largest foreshock (denoted by MLf) shows an increase in MLm with MLf. This indicates that for Taiwan’s earthquakes the bigger the largest foreshock is, the larger the mainshock is. The plot of the epicentral distance, Δ (in km), from the largest foreshock to the mainshock versus ML of the mainshock exhibits a weak increase in Δ with MLm as Δ1 day.

This research presents a new approach which addresses the conversion of earthquake magnitude as a supervised machine-learning problem through a multistage approach. First, the moment magnitude (M w) calculations were extended to lower magnitude earthquakes using the spectral P-wave analyses of the vertical component seismograms to improve the scaling relation of M w and the local magnitude (M L) of 138 earthquakes in northeastern Egypt. Second, using unsupervised clustering and regression analysis, we applied the k-means clustering technique to subdivide the mapped area into multiple seismic activity zones. This clustering phase created five spatially close seismic areas for training regression algorithms. Supervised regression analysis of each seismic area was simpler and more accurate. Conversion relations between M w and M L were calculated by linear regression, general orthogonal regression (GOR), and random sample consensus (RANSAC) regression techniques. RANSAC and GOR produced better results than linear regression, which provides evidence for the effects of outliers on regression accuracy. Moreover, the overall multistage hybrid approach produced substantial improvements in the measured-predicted dataset residuals when individual seismic zones rather than all datasets were considered. In 90% of the analyzed cases, M w values could be regarded as M L values within 0.2 magnitude units. Moreover, predicted magnitude conversion relations in the current study corresponded well to magnitude conversion relations in other seismogenic areas of Egypt.

Aim. Study of the dependence of mortality from CVD on helio-seismic indicators in the Shamakhi region of the Republic of Azerbaijan.Materials and methods. 352 case histories of patients who died in 2013 from various diseases were reviewed. The relationship was established between the number of deaths, their causes, distribution by sex and age with the magnitude of earthquakes, the depth of the epicenter and seismological activity by months. The data obtained were statistically processed using the Statistica 12.0 for Windows software package (Statsoft Inc., USA). To establish correlations, Pearson Chi-Square Tests was calculated, the results were considered reliable at p

Finding a suitable method for the accurate and reliable estimation of earthquake return periods is an important challenge for research in the field of earthquake engineering. Usually, the Gutenberg-Richter recurrence law is used to determine the average frequency of earthquakes with a magnitude greater than or equal to the specific value. The Gutenberg-Richter recurrence law can provide the value of the average return period, and cannot be used to determine the full details of return periods including the short, medium and long periods. Therefore, we considered the possibility of utilizing the discrete Fourier transform method to develop the frequency model for earthquake magnitudes. For this purpose, we used a series of sine functions with different values of amplitude, frequency, and phase as a prediction frequency model. To implement the proposed method, we used a catalog of earthquakes with magnitudes greater than 3.8 occurred within a radius of 50 kilometers around the Bojnord since 1990 to 2019. In addition to having the overall consistency with the frequencies of the Gothenburg-Richter recurrence law, the developed model provides more details of earthquake frequencies for different magnitudes. Additionally, there is a high correlation between the frequencies obtained by the frequency model and the tidal frequencies of the moon and sun. The correlation coefficient between the earthquake and tidal frequencies was obtained to be 93.7%. Considering the frequencies of the developed model reveals that the main return periods of the earthquakes in this region are 2, 6 and 10 months and 1 and 25 years.

Reliable short-term tsunami forecast on the Kuril Islands when earthquakes occur in the Kuril-Kamchatka Trench is the most difficult. Forecasting by the traditional magnitude method often leads to false tsunami alarms. Based on the examples of the events of 2006, 2007 and 2020 on the Kuril Islands, as well as the event of 2018 in Alaska, it was shown that according to the data of the ocean level measuring stations closest to the tsunami source (tsunami travel time is 10–20 minutes) it is possible to adequately predict the tsunami near the coast. Calculations of tsunami waveforms near the coast from data containing seismic noise have shown that the resulting waveforms contain high-frequency oscillations. However, these fluctuations do not interfere with the assessment of the real waveform and the danger of the expected tsunami. In contrast to forecast methods based on the magnitude criterion, the applied method of short-term tsunami forecast makes it possible to calculate the waveform: the amplitudes of the first, maximum waves, their arrival time at a given point and the estimated duration of the tsunami. The proposed method can become a tool that will improve the quality of operational tsunami warning, significantly reducing the number of false tsunami alarms.

Displacement based earthquake observations are becoming more common and application fields are becoming larger. The determination of the earthquake magnitude in a short time after the earthquake plays an important role both for Earthquake Early Warning Systems (EEW) and the estimation of the losses and coordinating the rescue efforts. However, magnitude estimation from the displacement waveforms for the 1999 Chi-Chi (Mw7.6), 1999 Hector Mine (Mw7.1) and 2003 Tokachi-Oki (Mw8.3) earthquake shows that while they are more precise, the earthquake magnitudes were underestimated (Brown et al. 2011; Crowell et al. 2013). There are well known bottlenecks with the velocity measuring seismometers due to the fact that they get saturated nearby large earthquakes or they do not provide sufficient time for early warning when they are away from the earthquake source. Geodetic measurements and accelerometer records provide data from the close vicinity of the earthquake source and enable to determine to static offsets and the earthquake magnitude based on displacements since GNSS receivers operate in a non-inertial reference frame. Nowadays, earthquake early warning systems based on seismogeodetic observations are being established and becoming more common in countries subject to earthquake hazard such as USA and Japan. In this study, earthquake magnitudes were estimated for several recent earthquakes in Turkey only by using the GNSS displacement waveforms. GNSS_based Re-estimation of the earthquake magnitudes for Gökova, Gökçeada and Van earthquakes resulted in the magnitudes of 6.54, 7.21 ve and 7.22 Mw, respectively, corresponding to the original magnitudes of 6.6, 6.9, 7.2. These results show that earthquake magnitudes based on GNSS waveforms provide high precision and their potential for use in earthquake early warning.

Predicting the earthquake magnitude is a complex problem, which has been carried out in recent years. The machine learning, geophysical, and regression methods were used to predict earthquake magnitude in literature. Iran is located in a highly seismically active area; thus, earthquake prediction is considered as a great demand there. In this study, two time series algorithms are utilized to predict the magnitude of the earthquake based on previous earthquakes. These models are autoregressive conditional heteroscedasticity (GARCH), autoregressive integrated moving average (ARIMA), and the combination of ARIMA and GARCH by multiple linear regression (MLR) technique (model 3). The 9017 events are used to train and predict earthquake magnitude. On the other hand, 6188 events are applied for training models, and then 2829 events are utilized for testing it. The statistical parameters, such as correlation coefficient, root mean square error (RMSE), normalized square error (NMSE), and fractional bias, are calculated to evaluate the accuracy of each model. The results demonstrate that the ARIMA and model 3 can predict future earthquake magnitude better than other models.

The Finite-Fault Rupture Detector (FinDer) algorithm computes rapid line-source rupture models from high-frequency seismic acceleration amplitudes (PGA). In this paper, we propose two extensions to FinDer, called FinDerS and FinDerS+, which have the advantage of taking into account a geological property of the source fault, its structural maturity, as well as its relation to the earthquake slip distribution. These two new algorithms calculate real-time earthquake slip profiles by backprojecting seismic and/or geodetic displacement amplitudes onto the FinDer line-source. This backprojection is based on a general empirical equation established in previous work that relates dynamic peak ground displacement (PGD) at the stations to on-fault coseismic slip. While FinDerS projects PGD onto the current FinDer line-source, FinDerS+ allows the rupture to grow beyond the current model extent to predict future rupture evolution. For an informed interpolation and smoothing of the estimated slip values, FinDerS and FinDerS+ both employ a generic empirical function that has been shown to relate the along-strike gradient of structural maturity of the ruptured fault, the earthquake slip distribution, and the rupture length. Therefore, while FinDer derives magnitudes from a relatively uncertain and general empirical rupture length-magnitude relations, FinDerS and FinDerS+ provide alternate and better informed magnitude estimates using the mean slip of the profiles derived from the integration of fault source maturity. The two new algorithms can incorporate both seismic strong-motion and geodetic displacement data. In order to recover PGD from strong-motion instruments, we double-integrate and high-pass filter (> 0.075 Hz) the seismic acceleration records. Together, the three algorithms exploit the full spectrum of ground-motions, including high frequencies to derive a source fault model (FinDer) and low frequencies to determine the static offsets along this model (FinDerS and FinDerS+). We test the three algorithms for the 2019 MW 7.1 Ridgecrest (California), 2016 MW 7.0 Kumamoto (Japan), and 2008 MW 7.9 Wenchuan (China) earthquakes. Conclusively, low-frequency PGD data and integration of the fault maturity gradient do not speed-up calculations for these events, but provide additional information on slip distribution and final rupture length, as well as alternative estimates of magnitudes that can be useful to check for consistency across the algorithm suite. The FinDer algorithms systematically outperform previously established real-time PGD-based magnitude estimates in terms of speed and accuracy. The resulting slip distributions can be useful for improved ground-motion prediction given the observed relationship between seismic radiation and fault maturity.

Predicting the parameters of upcoming earthquakes has always been one of the most challenging topics in studies related to earthquake precursors. Increasing the number of sensors and satellites and consequently incrementing the number of observable possible earthquake precursors in different layers of the lithosphere, atmosphere, and ionosphere of the Earth has opened the possibility of using data fusion methods to estimate and predict earthquake parameters with low uncertainty. In this study, a Mamdani fuzzy inference system (FIS) was proposed and implemented in five case studies. In particular, the magnitude of Ecuador (16 April 2016), Iran (12 November 2017), Papua New Guinea (14 May 2019), Japan (13 February 2021), and Haiti (14 August 2021) earthquakes were estimated by FIS. The results showed that in most cases, the highest number of anomalies was usually observed in the period of about one month before the earthquake and the predicted magnitude of the earthquake in these periods was slightly different from the actual magnitude value. Therefore, based on the results of this study, it could be concluded that if a significant number of anomalies are observed in the time series of different precursors, it is likely that an earthquake of the magnitude predicted by the proposed FIS system within the Dobrovolsky area of the studied location will happen during the next month.

Modern seismic design and technologies have undergone tremendous developments. In modern design codes, building structures subjected to high earthquake loads are allowed to experience plastic deformations without collapsing, and the design is permitted up to the ultimate strength. According to comparative results in Japan, the number of humans injured due to earthquakes is higher than the number of deaths/missing. Likewise, the number of residential buildings that collapsed are less than the partially damaged buildings. This outcome implies that residential buildings designed based on the revised seismic standards have good earthquake resistances. It also infers that the human deaths/injury casualties were not a result of the collapsed the structure, but due to the strong vibrations originated from the earthquake, yielding in the collapse of non-structural elements such as ceilings and bookshelves. This paper presents a proposed design philosophy that attempts to implement the effect of earthquakes to non-fatal human casualties

Modern seismic design and technologies have undergone tremendous developments. In modern design codes, building structures subjected to high earthquake loads are allowed to experience plastic deformations without collapsing, and the design is permitted up to the ultimate strength. According to comparative results in Japan, the number of humans injured due to earthquakes is higher than the number of deaths/missing. Likewise, the number of residential buildings that collapsed are less than the partially damaged buildings. This outcome implies that residential buildings designed based on the revised seismic standards have good earthquake resistances. It also infers that the human deaths/injury casualties were not a result of the collapsed the structure, but due to the strong vibrations originated from the earthquake, yielding in the collapse of non-structural elements such as ceilings and bookshelves. This paper presents a proposed design philosophy that attempts to implement the effect of earthquakes to non-fatal human casualties

Nepal, located above the convergent India-Eurasia plate boundary, has repeatedly experienced devastating earthquakes. During the 2015 magnitude 7.8 Gorkha earthquake, an often-reported experience was that people were not aware of the threatening seismic hazard and had an insufficient level of preparedness. An important source of the problem is that earthquake-related topics are not part of the school curriculum. Earthquake education reaching a broad group of the population early in their lives is therefore strongly needed. We established an initiative in Nepal to introduce seismology in schools, with a focus on education and citizen seismology. We have prepared educational materials adapted to the Nepali school system, which we distributed and also share on our program’s website: http://seismoschoolnp.org. In selected schools, we also installed a low-cost seismometer to record seismicity and to allow learning-by-doing classroom activities. Our approach was very well received and we hope it will help make earthquake-safe communities across Nepal. The seismic sensor which we installed in schools is a Raspberry Shake 1D (RS1D), this was selected based on its performance in laboratory tests and suitability for the field conditions. At a test site in Switzerland we were able to record magnitude 1.0 events up to 50 km distance with a RS1D. In Nepal, 22 such seismometers installed in schools create the Nepal School Seismology Network providing online data openly. The seismometer in each school allows students to be informed of earthquakes, visualize the respective waveforms, and estimate the distance and magnitude of the event. For significant local and regional events, we provide record sections and network instrumental intensity maps on our program’s website. In 6 months of network operation, more than 194 local and teleseismic earthquakes of M ≥ 4 have been recorded. From a local and a global catalog, complemented with our own visual identifications, we have provided an earthquake wave detectability graph in distance and magnitude domain. Based on our observations, we have calibrated a new magnitude equation for Nepal, related to the epicentral distance D [km] and to the observed peak vertical ground velocity PGVV [μm/s]. The calibration is done to best fit local catalog magnitudes, and yields the following equation: M = 1.05 × log10(PGVV) + 1.08 × log10(D) + 0.75.

In earthquake engineering, acceleration has played a major role, while wave energy has rarely been considered as a demand in design. In order to understand earthquake damage mechanism in terms of energy, the demand in terms of wave energy in surface soil layers is studied here, assuming one-dimensional SH wave propagation by using a number of vertical array records during nine strong earthquakes in Japan. A clear decreasing trend of the energy demand with decreasing ground depth and decreasing surface soil stiffness has been found as well as a propensity of incident energies calculated at bedrocks being roughly compatible with empirical formulas. How the energy demand is correlated with structural damage is also discussed in simplified models to show that induced structural strain is governed by upward energy flux, degree of structural resonance, and impedance ratio between structure and ground and structural stiffness. In low-damping brittle superstructures, wave energy flux in resonance and associated predominant frequency are decisive in determining the damage, while cumulative wave energy determines the damage in high damping ductile soil and massive concrete structures. The trend of lower energy demand in softer soil sites may not be contradictory, with a widely accepted perception that softer soil sites tend to suffer heavier earthquake damage as far as geotechnical damage is concerned.

In this paper, a heterodyne laser interferometer, which is used as a sensor for high-precision displacement measurement, is introduced to measure ground vibration and seismic waves as a seismometer. The seismic wave is measured precisely through the displacement variation obtained by the heterodyne laser interferometer. The earthquake magnitude is estimated using only the P-wave magnitudes for the first 3 s through the total noise enhanced optimization (TNEO) model. We use data from southern California to investigate the relationship between peak acceleration amplitude ( P d ) and the earthquake magnitude ( M g ). For precise prediction of the earthquake magnitude using only the P d value, the TNEO model derives the relation equation between P d and the magnitude, considering the noise present in each measured seismic data. The optimal solution is obtained from the TNEO model based objective function. We proved the performance of the proposed method through simulation and experimental results.

An integrated web application, referred to as ER2 for rapid risk evaluator, is under development for a user-friendly seismic risk assessment by the non-expert public safety community. The assessment of likely negative consequences is based on pre-populated databases of seismic, building inventory and vulnerability parameters. To further accelerate the computation for near real-time analyses, implicit building fragility curves were developed as functions of the magnitude and the intensity of the seismic shaking defined with a single intensity measure, input spectral acceleration at 1.0 s implicitly considering the epicentral distance and local soil conditions. Damage probabilities were compared with those obtained with the standard fragility functions explicitly considering epicentral distances and local site classes in addition to the earthquake magnitudes and respective intensity of the seismic shaking. Different seismic scenarios were considered first for 53 building classes common in Eastern Canada, and then a reduced number of 24 combined building classes was proposed. Comparison of results indicate that the damage predictions with implicit fragility functions for short (M ≤ 5.5) and medium strong motion duration (5.5 < M ≤ 7.5) show low variation with distance and soil class, with average error of less than 3.6%.

Large-scale liquefaction features (e.g., sand blows, lateral-spread fissure vents) that can be recognized on remote-sensing imagery and photography have been of great utility in developing chronologies of paleo-earthquakes. In areas where large-scale features are obscured on imagery by forest cover and Holocene exposure is lacking, small-scale liquefaction features (e.g., convoluted bedding, clastic intrusions, foundered and suspended blocks, water-escape structures) offer an alternative data source that can be investigated in meter-scale excavations. In order to assess the geographic extent of Holocene sand blow fields in southeast Arkansas that were previously mapped on river terraces and flood plains using aerial photography, we investigated the distribution of small-scale liquefaction features in alluvium along streams within a forested region between the sand blow fields. Our results suggest that the fields are not continuous and do not reflect a single large liquefaction field related to paleo-earthquakes >M 6.5. Features at one of our sites suggests the Desha County sand blow field may be larger than presently mapped, and that the distance from the center of the field to the farthest liquefaction may be ~30 km. The empirical relationship of magnitude and distance to farthest liquefaction suggests a field of this size could have been produced by a M 6.3 earthquake. We also found Holocene liquefaction features that we interpret as resulting from ground shaking near previously documented Pleistocene and Holocene surface ruptures of the Saline River fault zone. Liquefaction during a paleo-earthquake (~M 5.5) may have coincided with movement on that fault zone ~ AD 1700.

Examination of 20th century data from the U.K. shows a good correlation between felt area and instrumental magnitude, producing good magnitude values for historical earthquakes where felt area is known; this is a more reliable method than using I0. The equations: ML = 1.03 log A3:; -0.19; ML = 0.92 log A4 + 0.71 give magnitude from areas within isoseismals 3 and 4 MSK. Dcpths for historical earthquakes have also been obtained using a modification of the well-known Sponheuer method, using the program MACDEP; the regional value for (á is 0.003 and depths for events with magnitude above 4 ML range from 3 to 25 km.

In the very early morning at 03 h 53.7 m on 15 April 1909 (local time), a large earthquake occurred in northern Taiwan. In all, 9 persons were killed and 51 injured; 122 houses collapsed along with damage to another 1050 houses. This earthquake was one of the largest and most damaging events of the 20th century for the Taipei Metropolitan Area. The epicenter estimated by Hsu (1971) was determined to be 25¢XN, 121.53¢XE and its focal depth and earthquake magnitude evaluated by Gutenberg and Richter (1954) were ~80 km and MGR = 7.3, respectively. The event took place underneath the Taipei Metropolitan Area and might be located at the western edge of the subduction zone of the Philippine Sea plate. In this study, the magnitudes of the earthquakes determined by others will also be described.