Silvia Castellaro

Performance of Different Regression Procedures on the Magnitude Conversion Problem

Through an extensive set of simulations we investigate the performance of different linear regression procedures commonly used to convert magnitudes from one type into another one, an operation that also has strong influence on the slope of the frequency-magnitude (the b -value of the Gutenberg–Richter) distribution. It has already been demonstrated that a general orthogonal regression provides the most reliable results. However, questions arise when the ratio between the variances of the magnitudes to be related (the knowledge of which is required to apply the general orthogonal regression) cannot be computed. We therefore systematically investigate the biases introduced by the classical standard least-squares regressions and the orthogonal regressions (or similar procedures) as a function of the true slope between magnitudes, of the ratio η between magnitude variances, and of the absolute variances of magnitudes. We compute such biases through simulations very close to the real cases inferred from the German and Chinese broadband networks. We observe that for 0.7 < 1.8 the orthogonal regression under the η = 1 assumption performs better than standard regressions. For values outside this interval neither procedure is capable of correct estimates. Therefore it is recommended to estimate the absolute errors and their ratio from empirical data and apply the general orthogonal regression. This requires that a seismological data center publish average estimates of event magnitudes and also their related standard deviations. Regrettably, this is not yet a common practice, thus impeding the derivation of optimal magnitude conversion relations. Online material: Graphics illustrating the performance of different regression procedures on magnitude conversion. [1]: /embed/inline-graphic-1.gif

Chinese National Network Magnitudes, Their Relation to neic Magnitudes, and Recommendations for New iaspei Magnitude Standards

We investigate the linear regression relationships between common seismic magnitudes determined by the Chinese Earthquake Network Center (CENC) and compare them with related magnitude determinations for the same events at the U.S. Geological Surveys National Earthquake Information Center (NEIC). Despite their generally good agreement some systematic differences are revealed. These dif- ferences are due to differences in seismograph response (shape and bandwidth), the time window for measurement of maximum P-wave amplitudes, the period and dis- tance ranges used, and, in part, also the different calibration functions applied. Chi- nese broadband body-wave magnitude mB, compared with the NEIC short-period P- wave magnitude mb, is much less prone to magnitude saturation. Thus it is more suitable to assess the size of large earthquakes from P waves. Also, following Inter- national Association of Seismology and Physics of the Earths Interior (IASPEI) recommendations of 1967, Chinese surface-wave magnitude MS is determined in a wider distance (1 D 180) and period range (3 sec T 30 sec) than MS(20) at NEIC (20 D 160 and 18 sec T 22 sec, respectively). Chinese MS for small and medium earthquakes at regional distances between 2 D 10 scales well with local magnitude ML. In contrast, NEIC MS(20) tends to underestimate the magnitude of regional events when the IASPEI-recommended MS calibration function by Vane yk et al. (1962) is used. These findings support some of the new standards for magnitude measurements from digital data adopted at the IASPEI meeting in 2005. They include, complementary to band-limited m b and M S (20), the determination of m B and M S (BB) measured on unfiltered broadband records.

The different response of apparently identical structures: a far-field lesson from the Mirandola 20th May 2012 earthquake

Twin structures, that is structures very similar in terms of geometry, materials, mass distribution etc., founded on the same soil and set at very close distance, are rationally expected to have an identical response to earthquakes. When this does not occur, a role is usually played by factors like the interaction with the surrounding structures or by other anomalies hidden behind the apparent similarity. We present the case of two apparently twin towers that showed a very different response to the 2012 Mirandola (Italy) earthquake ground shaking: one remained perfectly intact while the other had a wide set of fractures on secondary walls. This resulted to be the effect of several contributing factors: the stiffness of the two structures, experimentally measured, provided unexpected differences. This reflected into different modal frequencies for the two towers, with the first and second modes of the damaged tower coincident or very close to the soil resonance. The final result was a coupled soil-structure resonance, implying a much higher displacement of one tower compared to the other, under the same input motion. In Italy, insurance against earthquake damage will probably become compulsory in the near future. This case suggests that the specific soil-structure and structure-structure interaction will have to be carefully evaluated since they can critically affect even apparently identical structures.

Simplified seismic soil classification: the Vfz matrix

Site effect assessment studies aim at predicting the effect of seismic shaking on structures by modeling the subsoil as an oscillator coupled to another oscillator representing the construction. The resulting amplification functions and response spectra depend on so many strong assumptions and parameters that, in the standard engineering practice, simplified seismic classifications appear preferable to complex modeling procedures which can only offer an illusory better accuracy. Since stratigraphic seismic amplification is not properly related to the absolute rigidity of subsoil but to impedance contrasts, the standard simplified approaches based on the ‘average’ rigidity of subsoil in the first few meters (e.g. Vs30) can hardly be effective. Here it is proposed a simplified soil classification approach that takes into account the basic Physics of seismic amplification and its parameters, i.e. the average shear wave velocity of the cover layer, the resonance frequency and the impedance contrast between the cover and the bedrock, which we summarize as VfZ. A possible classification approach is illustrated through a set of examples.

Nondiffuse elastic and anelastic passive imaging.

The property at the basis of passive acoustic imaging is that, taken any two points, one of them can be seen as the source of the waves and the other as the recording station. This property, which was shown to hold also in nondiffuse fields, is here exploited: (1) to allow an undistorted passive imaging through the simple use of the statistical mode to estimate wave velocity, (2) to determine the azimuth of the instantaneous Huygens sources of the noise wavefield, and (3) to measure, provided that the noise bandwidth is wide with respect to that of the local system, the material dissipation constant as a function of frequency. The authors applied this theory to study the seismic noise field in the Ravenna, North-Central Italy, shore area and found it capable to provide velocity dispersion curves matching those of independent surveys, to track the sources of seismic noise to a few major firms in Ravenna port, with the prevailing source switching at the time scale of seconds, and to measure the dissipation quality factor Q at approximately 20 independent of frequency in the range 1-30 Hz.

Reconstructing seismic ground motion at reference site conditions: the case of accelerometric records of the Italian National Accelerometric Network (RAN)

Abstract To compare the outcomes of seismic hazard models with accelerometric records, the seismic ground motion at the reference soil conditions considered by the hazard estimates must be reconstructed. Here we illustrate the procedure proposed and operated to this purpose in the frame of the Italian DPC-INGV-S2 project. This includes two phases: the geological/geophysical characterization of the sites to be analysed and a numerical deconvolution analysis, carried on by considering specific records representative of the maximum ground motion observed at the relevant accelerometric sites. Particular attention was addressed to manage the uncertainty associated to this kind of analysis. Results are then compared with those obtained by simply applying the “correction” coefficients provided by the Italian National Seismic Code to account for site conditions.

Passive Seismic Survey for Cultural Heritage Landslide Risk Assessment

The present work aims at reconstructing the stratigraphy at the ancient citadel of Machu Picchu by using a passive seismic technique to implement a landslide stability model and a risk mitigation plan. After the warning launched in March 2001 by the scientific community about the potential collapse of the citadel as a consequence of potential landslide event, different studies have been promoted to reconstruct the landslide activity and to suggest mitigation measures for the protection and conservation of Machu Picchu cultural heritage.

Investigation of Composite Compacted Ground Using Microtremors

AbstractThis paper presents an interesting and unique case study of a composite compacted site where the upper section of dynamically compacted material achieved in the first stage of compaction was subsequently removed, reinstated, and recompacted in lifts in the second stage using conventional roller compaction. Dynamic compaction was employed initially in this area because of the need to densify deep fill materials. Although a number of mechanical methods are already available for assessing deep compaction, it has been particularly rare to find a cost-effective method that can be applied to a deep and extensive compacted site. Noninvasive techniques based on measurement of the horizontal-to-vertical spectral ratio (HVSR) of ambient vibrations (microtremors) are proposed in this paper to assist in a pilot appraisal of this area, which occupies a part of a deep and laterally extensive compacted site. First, the key features of the measured HVSR curves were interpreted to give a preliminary insight into t...

Estimates of Vs30 Based on Constrained H/V Ratio Measurements Alone

Most seismic codes adopt as a key quantitative parameter the average shear wave velocity in the first 30 m of subsoil, commonly calling it Vs30. Estimates of Vs30 are therefore required in most countries for microzonation both at large-scale as well as at the scale of the single buildings. We propose a fast and inexpensive technique to measure the Vs30 based on the horizontal to vertical spectral ratio (H/V) of microtremor recorded at a single station. The experimental H/V is fitted with a synthetic curve using as a constraint the thickness of the most superficial layer of the subsoil, which is always independently known in the geotechnical practice for building design approval. The fitting procedure consists of three steps: (1) identify the depth of the first shallow stratigraphic horizon from geotechnical data, (2) identify its corresponding H/V marker and (3) use it as a constraint to fit the experimental H/V with the synthetic one. The synthetic H/V curve is calculated by assuming a tremor wavefield with Rayleigh and Love waves in the fundamental mode in a stratified 1-D soil model. A validity check of this technique has been performed on a variety of geological settings in Northern Italy. The validity check compares: (1) the theoretical Rayleigh wave phase velocity dispersion curves calculated for the models derived from the H/V fits with the experimental curves measured with ESAC and ReMi array surveys, (2) the Vs30 estimates obtained with the proposed technique with those obtained at the same sites by using the latter array techniques. The inferred stratigraphy has been also compared with the geological knowledge. The proposed technique is found capable to provide, at a small fraction of the costs and survey times, Vs30 estimates coherent with those measured by ESAC and ReMi. Furthermore, the proposed technique is also found to be more informative than array techniques, allowing to detect deviations from 1-D subsoil geometry over lengths of a few meters, the correctness of which was confirmed by direct drilling.

Continuous monitoring of surface wave velocity at the Montevecchio earthflow (Forlì-Cesena Province, Northern Apennines)

The Montevecchio landslide is an active earthflow in the province of Forli-Cesena (Northern Apennines of Italy). In the last few years, this landslide reactivated several times. In order to investigate the landslide dynamics, two monitoring systems were installed by combining 3 permanent GPS stations (to measure landslide movement) and 4 geophones (to evaluate surface-wave velocity). In this paper we present the monitoring data collected during the last reactivation of the landslide (24th – 25th of May 2015). The geophones data show a clear relationship between surface-wave velocity (Vr) and displacement rate. In particular, a decrease of Vr value was observed just before an increase of velocity, while higher Vr values were detected when the landslide was slowing down. Moreover, a pressure sensor, buried at low depth beside the geophones, showed an increase in pore water pressure before the failure in correspondence of the critical rainfall event, and while the displacement rate increased the pore water pressure had a positive value (above the ground level), probably due to the undrained compression of the landslide material.