Haines, John

A dense 2D array (95 broad-band stations) was installed in a grid-like configuration over a 130x56 m area on the active landslide of Cavola in northern Apennines. Outcropping rocks consist of Monte Piano (fractured flysch and clayey melanges) and Ranzano (muddy and sandstone turbidite) formations of Eocene-Oligocene age. These units are covered by a 4-km long landslide with maximum thickness more than 60 m. In historical times there are three instances of activity triggered by meteorological events, in 1938, 1940, and 1960. The latter event was the most damaging, involving an area of 1.3 km2 and partial destruction of the Cavola village. Recent SAR interferometric analyses coupled to inclinometer data for the supposedly dormant period 1995-2001 show movements between 10-15 mm/yr affecting mainly areas close to the landslide side-scarps and to a new industrial district with active sliding surface at a depth of 10-15 m. The landslide structure beneath the array was reconstructed by means of active and passive seismic surveys, in-situ velocity measurements and geoelectrical tomography obtaining clay thickness ranging from 0 to 45 m and with average shear-velocity of 350 m/s. The inferred model is used for numerical 2D simulation of earthquake ground motion for different incidence angle and type of waves. The data from 11 of more than 100 local earthquakes recorded at the array have been used for the comparison with model results. The long data records (3 months of continuous recording) allow investigation of properties of ambient noise in terms of temporal stability and wavefield composition. We find that local industrial sources affect the H/V spectral ratios and their interpretation in terms of resonant frequencies. Analysis of noise polarization is also performed, revealing different patterns depending on the nature of the noise source.

We have studied two velocity-depth models with the aim of outlining the behavior of a velocity reversal in the top layer, which is associated with the stiff Brecce de L’Aquila unit (BrA). In this setting, the SMTH model is topped by a layer with about 2:1 impedance contrast with the underlying layer while the NORV model has no velocity reversal. We have simulated the propagation of SH and P-SV wavefields in the range 0–10 Hz for incidence 0◦ –90◦ . Earthquake spectral ratios of the horizontal and vertical components at six sites in L’Aquila downtown are compared to corresponding syn- thetics spectral ratios. The vertical component of P-SV synthetics enables us to investigate a remarkable amplification effect seen in the vertical component of the recorded strong motion. Sites AQ04 and AQ05 are best matched by synthetics from the NORV model while FAQ5 and AQ06 have a better match with synthetics spectral ratios from the SMTH model. All simulations show this behavior systematically, with horizontal and near-horizontal incident waves predicting the overall pattern of matches more clearly than vertical and near-vertical incidence. The model inferences are in agreement with new geological data reporting lateral passages in the top layer from the stiff BrA to softer sediments. Matches are good in terms of frequency of the first amplification peak and of spectral amplitude: the horizontal compo-nents have spectral ratio peaks predominantly at 0.5 Hz in the simulations and at 0.7 Hz in the data, both with amplitudes of 4, while the vertical component spectral ratios reach values of 6 at frequencies of about 1 Hz in both data and simulations. The vertical component spectral ratios are very well matched using Rayleigh waves with incidence at 90◦ . The NORV model without the velocity reversal predicts spectral ratio peaks for the horizontal components at frequencies up to 6 Hz. The reversal of velocity acts as a low-pass frequency filter on the horizontal components reducing the amplification effect of the sediment filled valley.

Experimental and modeling approaches fulfill complementary needs in the assessment of the seismic soil response. Here we present some results from 2D simulation performed for the L’Aquila basin (Central Appennines, Italy). The city of L’Aquila on April 6th, 2009 at 01:32 UTC was struck by a magnitude Mw=6.3 earthquake localized about 2 kilometers west of the city centre at hypocentral depth 9 km. The city of L’Aquila suffered wide spread destructive damage in its historical centre where housing is mainly 2 to 3 storey medieval masonry. Several reinforced concrete buildings built in the late 70s collapsed in the south-west section of the city and also monumental, historical churches were severely damaged. Because of its location, the ground motion recorded in the city is strongly influenced by the source rupture mechanism, nonetheless, local amplification are expected to have influenced the ground shacking. L’Aquila is indeed built over a Quaternary terraced alluvial-lacustrine basin with a rather complex lithology as well as surface topography. The 2D seismic modeling of L’Aquila terrace was already performed by several authors along transversally oriented (NE-SW) geological sections. In this study we present some new results obtained by the use of longitudinal cross sections (NW-SE) in order to better understand the role of lateral geological heterogeneities as derived by recent geological and geophysical data. The simulations have been performed using the impedance-operator-based numerical code. The models of L’Aquila terrace are based on geological and geophysical investigations performed in the framework of the micro-zoning activities of the city following the disastrous April 6th 2009 Mw=6.3 earthquake and in subsequent studies. The depth to the bedrock of the basin is constrained by gravimetric and deep borehole data with an estimated maximum depth of about 300 m. The basin is filled by silt and silty-clay of lacustrine origin topped by a breccia layer (BrA) of gravitational-fluvial origin. BrA does not extend continuously over the terrace and in particular in the southern area of the city it is locally replaced by silt and silty-clay of lacustrine origin with lens of BrA and gravel with silt. Locally, on the top of the terrace lens of less competent red silt were found by the recent deep borehole surveys performed in the micro-zoning activities. We have further constrained our 2D models using the resonance frequencies from noise and earthquake spectral ratios for selected sites. The wave velocities have been inferred by MASW and cross-hole analyses. We have compared the spectral ratios obtained from SH, P-SV and Rayleigh incident waves field in the range 0 -90 to the observed spectral ratios computed using the earthquake aftershocks recorded by the micro-zoning portable network. The modeling results are able to match the resonance frequency obtained by seismological data and to verify the role of the reversal in the velocity-depth profile and the lateral continuity of the top fast layer (BrA).

The city of L’Aquila is built over a terraced alluvial basin filled by silt and silty-clay of lacustrine origin with average S-wave velocity of about 725 m/s, topped by a breccias unit (BrA) with Vs of about 900 m/s. Coupled with a reversal of velocity, there are marked lateral heterogeneities. In the southern area the stiff BrA unit laterally passages into softer deposits and is topped by red silt, i.e. LRCA, with an average velocity of 350 m/s. Bedrock is limestone and marls and its estimated maximum depth is of about 300-400 m. Recent geological and geophysical surveys performed during microzoning activities constrain this model. Throughout the city measurements of horizontal ground motion (SSR and HVSNR) show a marked low frequency peak (< 1 Hz), with variable amplitude reaching its maximum in the southern area. However amplification level often exceeds two over the 2-10 Hz frequency range. Vertical ground motion has a remarkable peak at frequency higher than the horizontal one. Bordoni et al. [2011] in their simulation work have shown that: 1) the velocity reversal filters out the high frequency content and 2) synthetic Rayleigh waves can predict the vertical ground motion amplification. Here we further our investigations on 2D ground motion focusing on the behavior of lateral heterogeneity in the BrA layer coupled to reversal of velocity. Synthetic spectral ratios from 0° and 90° incidence angle from SH waves are compared to observations from sites investigated by the microzoning portable network.

The city of L’Aquila is built on a deep intramontane basin filled by silty clay of lacustrine origin with average S-wave velocity of about 725 m=s, topped by a calcareous breccias unit with a higher speed of about 900 m=s. New geologic and geophysical investigations following the 6 April 2009 Mw 6.3 L’Aquila, Italy, earthquake have allowed us to develop a new velocity model for the basin deposits that, coupled with the inversion of velocity, includes lateral velocity variations in the top breccia layer and a newly found red silts unit, the Limi Rossi del Colle dell’Aquila (LRCA). The city area where the LRCA unit outcrops is correlated with clusters of reinforced concrete buildings that collapsed in the nor- mal faulting earthquake. We simulate the 2D seismic response of this new velocity model along two orthogonal cross sections and compare the synthetic spectral ratios to the experimental ones evaluated using aftershock recordings. As a result, taking into account the full scatter about the mean of the observed spectral ratios, we are able to predict not only the main features of the primary observed low-frequency resonance peak, but also the level of amplification at high frequency at most of the sites investigated.

Horizontal-to-vertical spectral ratios using ambient noise (HVNSR) are commonly used in site effects studies. In the practice, many operators assume stability over time of HVNSR and base their analyses on few very short time windows. The availability of a long period of continuous microtremor recording allowed us to analyze three months of data coming from a dense array experiment performed at Cavola, a village in northern Apennines. This condition offers a good opportunity to check the validity of the stability assumption and to investigate variations of the local ambient noise wave-field composition. The Cavola site is characterized by landslide sediments over stiffer materials with a moderate impedance contrast and by a complex morphology. An intense industrial activity in the village contributes to the generation of seismic noise. After identifying this noise source in the time series, we evaluate its effects on HVNSR. The results indicate that the spectral peak of HVNSR varies in amplitude and frequency, posing a warning about stability in time. Analyzing the spectra we identify the anthropic activity as responsible for changes in the composition of the noise wave field. These variations affect HVNSR, including peak frequency and also ground-motion polarization.

Geophysical site investigations have been performed in association with deployment of a dense array of 95 3-component seismometers on the Cavola landslide in the Northern Apennines. The aim of the array is to study propagation of seismic waves in the heterogeneous medium through comparison of observation and modelling. The small-aperture array (130 m×56 m) operated continuously for three months in 2004. Cavola landslide consists of a clay body sliding over mudstone-shale basement, and has a record of historical activity, including destruction of a small village in 1960. The site investigations include down-hole logging of P- and S-wave travel times at a new borehole drilled within the array, two seismic refraction lines with both P-wave profiling and surface-wave analyses, geo-electrical profiles and seismic noise measurements. From the different approaches a consistent picture of the depths and seismic velocities for the landslide has emerged. Their estimates agree with resonance frequencies of seismic noise, and also with the logged depths to basement of 25 m at a new borehole and of 44 m at a pre-existing borehole. Velocities for S waves increase with depth, from 230 m/s at the surface to 625 m/s in basement immediately below the landslide.

A reference site has to be free of amplification or de-amplification effects, namely with no troughs and peaks in its Fourier amplitude spectrum. At the Cavola landslide we show that this spectrum is dependent on the direction of propagation of wave fronts for incidence angles in the range 30° to 90°. Our study is based on comparison of spectral ratios from observations and 2D numerical simulations. We have modeled propagation in a 2D profile for SH and SV waves with several incidence angles in the 0° ±90° range, where 0° and 90° are respectively vertical and horizontal incidence, except that ±90° denotes Rayleigh waves in the P-SV. We discuss in detail the result for angles of incidence of 0, ±20°, ±60°, ±90°. We have obtained observed horizontal-to-horizontal earthquakes spectral ratios using three reference sites. Two of these have matching receivers in the model, located at the opposite ends of the 2D profile. Overall observations are matched best when the reference site is located on the same side of the landslide as the incoming wave front. We also find general agreement of the observed H/H spectral ratios from earthquakes with H/V and H/H spectral ratios from noise, and the match between H/V values from noise and synthetic spectral ratios using an absolute, flat half-space reference is very good. On the other hand, 1D modeling performs poorly in comparison with 2D modeling in our case, for which the shape ratio h/D = 0.2 is intermediate between primarily-1D and strongly-2D wave propagation according to the classification of Bard and Bouchon (1985).