Marijan Herak

ModelHVSR-A Matlab ® tool to model horizontal-to-vertical spectral ratio of ambient noise

ModelHVSR is a user-friendly collection of Matlab^(R) routines performing a number of tasks related to analysis of microtremor data. Its distinguishing features include:*Computation of amplification spectra for horizontally stratified soil models, for vertically incident P- and S-waves. Q-values may be frequency dependent. For constant-Q models dispersion of body-waves is taken into account; *Computation of theoretical horizontal-to-vertical spectral ratios (HVSR) for body-waves, using estimates of both horizontal and vertical amplification of soil; *Computation of average observed HVSR using standard procedures or Landweber-filtered spectra; *Computation of linear amplification factors (DAF) of the peak-ground acceleration for any soil model and target earthquake; *Inversion of the observed HVSR spectra by Monte Carlo perturbation of initial model parameters, to obtain the best-fitting family of models; *Computation of confidence regions for all inverted model parameters (including DAF). The program may be used during microzonation or similar studies, when one needs to either verify the existing geotechnical models by comparing theoretical HVSR to the observed one, or invert the observed ambient noise measurements to obtain the most-likely geotechnical models of the soil. It is also useful to conduct sensitivity analyses and to explore possible biases between model parameters.

Eurasia‐Africa Plate Boundary region yields new seismographic data

The tectonic plate boundary between Eurasia and Africa is complex, in that it cannot be characterized as a single discrete plate boundary Deformation near this plate boundary varies from trans-tensional in the Azores archipelago, through strike-slip in the eastern Atlantic basin, to overall compressional between the European and African continents, with extensional sub-domains in the Mediterranean Sea. This complex pattern of deformation, related plate motion, and underlying driving forces leads to strong variations in seismic hazard throughout the region. A better understanding of the plate boundary processes requires knowing crust and upper mantle structure in the region, which is best investigated with three-component, broadband seismic data. To investigate the regions three-dimensional crust and upper mantle structure, we are carrying out a multiinstitutional project (MIDSEA) involving seismologists from 10 countries on the northern, southern, and western sides of the plate boundary.

Seismic Zoning of Croatia

The territory of Croatia and neighboring regions is divided into 17 seismic source zones, considering available seismological and geological data. On this basis, seismic hazard elements (seismicity rate, maximum magnitude, b-value, probabilities of exceedance and return periods for a predefined set of magnitudes) are computed using the maximum likelihood method appropriate for treating data-sets with variable completeness thresholds. The values of long term expected peak horizontal acceleration obtained by using a combination of the deterministic and the probabilistic procedure are the highest in the Dubrovnik zone, while the Zagreb zone has the highest earthquake hazard in the continental part of the country.

Gutenberg's surface-wave magnitude calibrating function: Theoretical basis from synthetic seismograms

The surface-wave magnitude, as originally defined by Gutenberg (1945), appears to be the most stable quantity rating the strengths of seismic sources. It is based on the ground displacement produced by a nearly monochromatic wave train. The calibrating function underlying the magnitude definition was obtained from observations of the amplitudes as a function of epicentral distance for a wealth of earthquakes. No distinction was made as to focal depths. Thus, the calibrating function depends on one parameter—the epicentral distance-only. Subsequent attempts to introduce a focal depth correction were not successful in practice, and one and the same calibrating function continues to be applied to events with focal depths ranging from 0 to about 70 km. Assuming a strike-slip point source located inside a standard Earth model, synthetic seismograms of the fundamental mode SH-(Love-) wave are computed for a range of focal depths and azimuths. Amplitude-distance curves are obtained, and averaged over the azimuths. The curves are the basis for improved calibrating functions. Average depth corrections are obtained and applied to some populations of seismic events. The corrected surface-wave magnitudes are internally more consistent than the uncorrected ones usually published. Applying the depth correction to published surface wave magnitudes and considering themb-Ms relation for a sample of earthquakes and explosions, it has been established that, for a given intensity of short-periodic radiation, the 20-s radiation intensity can vary by orders of magnitude. While it is known that the 20-s radiation intensity is generally lower for explosions than for earthquakes, it turns out to be even lower if the depth correction is applied, yielding a clearer separation between both classes of seismic events.

A Contribution to Seismic Hazard Assessment in Croatia from Deterministic Modeling

Abstract—Some of the elements of regional seismic hazard in Croatia are assessed by computing synthetic accelerograms at a predetermined set of sites. The input dataset consists of structural models, parameters of seismic sources, and an updated earthquake catalog. Synthetic strong-motion time series for frequencies below 1 Hz are computed on a grid of sites using the modal summation technique. The long-period hazard is described by the distribution of estimated peak values of ground displacement, velocity and acceleration, while the short-period hazard is represented by the map of design ground acceleration values (DGA). The highest values of DGA exceeding 0.35 g on the base-rock level are found in the southeastern coastal part of the country, in the greater Dubrovnik area.

Amplification of strong ground motion in the city of Zagreb, Croatia, estimated by computation of synthetic seismograms

A hybrid technique consisting of modal summation and subsequent finite differences modelling is applied for the computation of synthetic accelerograms along a profile crossing the city of Zagreb, the capital of Croatia. Assuming the source geometry is known, the amplification properties of the underlying soil may be determined by comparison of synthetics and their response spectra computed for a bedrock model with the ones obtained under the assumption of a realistic laterally varying local model. The peak ground acceleration is larger by a factor of up to 3.5 than the value obtained for the bedrock model. The amplification of the response spectra is most prominent for frequencies below 2 Hz, and increases sharply to the SW from the mapped fault running through the centre of the city.

Body-wave velocities in the circum-Adriatic region

Abstract The paper presents body-wave velocities in the crust and upper mantle of the circum-Adriatic region obtained by analyzing travel times of regional phases of 419 selected earthquakes with epicenters in the central part of the External Dinarides. The studied region was divided into five areas and two kinds of velocity models were determined for each of them. The first kind ( inter-area models ) gives average velocities that characterize the zone between and including the epicentral region and the area in question, while the second kind ( intra-area models ) presents velocities within the respective area. The locations of hypocenters considerably improved when the new models were used instead of the standard one for the Balkan region. The large-scale variation of average velocities obtained by inter-area modeling was found to be relatively small both in the crust and in the upper mantle. On the other hand, the magnitude and distribution of station corrections and the resulting intra-area velocities point to relatively pronounced lateral velocity variation within some of the studied regions. The obtained velocities in most cases do not differ significantly when compared to the values reported in other studies. The most important exceptions occur in the epicentral area (central part of the External Dinarides) where the P-wave velocity in the upper crust is much higher then elsewhere, while the velocity in the upper mantle is found to be lower than that determined by DSS experiments. There is also an indication of pronounced variation of the Poisson ratio in this area.

Theoretical and Observed Depth Correction for Ms

Abstract - Modal summation technique is used to generate 5000, three-component theoretical seismograms of Love and Rayleigh waves, assuming modified PREM (PREM-C) and AK135F global earth models. The focal depth h and the geometrical fault parameters are randomly chosen so as to uniformly cover possible source mechanisms and obtain uniform distribution of log h in the interval 1 < h < 600 km. The amplitudes of 20 ± 2-s waves measured on each of the synthetic seismograms yield curves of amplitude vs. depth, and consequently the theoretical surface waves magnitude depth correction for all three components of ground motion. Predicted surface wave amplitudes are practically constant for h < 20 km, then decrease with the focal depth. This decrease is not uniform, and depends on the excitation level of higher modes. For PREM-C model and for shallow sources, computed Love wave amplitudes are nearly an order of magnitude larger than those of Rayleigh waves, which is why the AK135F model is given preference over PREM-C. The theoretical depth correction ranges between zero, for shallow sources, to about +1 magnitude unit, for the deepest ones. The theoretical results are compared with 74,480 individual station measurements of 20-s surface wave amplitudes reduced to the same distance, period and seismic moment. It is found that empirical data closely match the predictions made by using the AK135F model. Considering both theoretical modeling and observational data, we propose the introduction of a piecewise log-linear depth correction for M s of the form: ΔM s (h) = 0 for h < 20km, ΔM s (h) = 0.314 1og(h) - 0.409 for 20 ≤ h < 60km, ΔM s (h) = 1.351 log(h) - 2.253 for 60 ≤ h < 100km, ΔM s (h) = 0.400 log(h) - 0.350 for 100 ≤ h < 600 km.

Lapse-time dependent Qc-spectra observed in the Dinarides region (Yugoslavia)

Abstract The paper presents results of the coda quality factor (Qc) measurements in the central Outer Dinarides and the central Adriatic Sea regions. The main objective of this study was to analyse quantitatively the frequently observed Qc increase with lapse time, t, 108 earthquakes were chosen for analysis, with lapse times between 30 and 130 s. After computing the coda amplitude spectra A(ƒ) for 0.6 ⩽ ƒ ⩽ 5.0 Hz , the curve 1n(At) = a + btc was fitted to observed data for each earthquake and for each frequency. Denoting its slope by p(t), Qc was then estimated as a function of t as Q c (t|ƒ) = −πƒ p(t) . In this way coda decay rate measurements were done in a consistent way, regardless of the lapse time window used. These individual values were then weighted and averaged in order to calculate the regional Q c (ƒ,t) . The results indicate that in the region studied Qc increases both with frequency and lapse time, while the degree of frequency dependence decreases with lapse time. If Qc is expressed as Q c (t) = Q 0 (t) ƒ n(t) then for lapse times between 30 and 130 s Q0 increases from 45 to 275 while n decreases from 1.0 to 0.6.

Attenuation of Peak Horizontal and Vertical Acceleration in the Dinarides Area

Peak acceleration attenuation relations for horizontal and vertical components are presented for the Dinarides region, based on 145 3-component accelerograms related to 46 earthquakes with local magnitudes of 4.5 or greater and with epicentral distances of less than 200 km as recorded on 39 recording sites in the greater Dinarides region. The attenuation functions were obtained by two-stage stratified regression on the local magnitude and epicentral distance as independent variables. The predicted peak acceleration values within the distance range covered by the data are comparable to the ones obtained for stiff-soil or rock sites when selected reference relations are used. The rather large average residuals are caused mostly by the lack of information on local site conditions and by the use of epicentral distance instead of fault distance.