Geophysical techniques for urban environment monitoring

The research activities conducted in this thesis contributes, through the application of geophysical techniques, to the mitigation of seismic risk with the twofold objective of studying the interaction between the urban subsoil and the overlying-built heritage and carrying out a modal characterisation of a strategic infrastructure. The former objective was pursued by producing a map of the double soil-structure resonance levels of the Matera urban area, while the latter was achieved by setting up and applying an innovative multi-methodological geophysical approach on the Gravina Bridge. As part of the first study, I performed 230 single-station ambient seismic noise measurements on the main lithologies (134) and on the main typology of buildings (96) in reinforced concrete (RC) and unreinforced load-bearing masonry buildings (URM) of the Matera urban area. The ambient seismic noise recorded on the soil 12 min time duration and on buildings 14 min time duration was recorded with a compact digital seismometer and processed using a non-reference site method, the Horizontal-to-vertical noise spectral ratio technique, HVNSR. The measurements taken on the ground and buildings allowed the resonance frequencies and relative amplitudes of the fundamental peaks of the soil and the first elastic frequency of vibration of the buildings to be estimated. A deterministic interpolator (Inverse Distance Weight, IDW) was used in GIS environment to derive the iso-frequency and iso-amplitude maps of the urban area by using as variables the resonance frequencies and amplitudes of the soil HV ratios. A linear period-to-height relationship for the buildings was derived from the experimental results, allowing the fundamental elastic frequency to be estimated for all buildings in the study area. An intersection approach between soil and building frequency bands was used for the first time to derive a map of double soil-structure resonance levels in the linear elastic domain for the whole urban area. Matera represents an important case study since the elastic frequency of vibration for most of the buildings is quite close to that of the foundation soils. In the study area, 21% of the buildings show a high susceptibility to the effect of double soil-building resonance, 63% of the buildings could be characterised by a medium level of double resonance, while 16% could exhibit a zero or very low resonance level. The proposed approach also makes it possible to locate the areas of the city characterised by these different levels of double resonance. Therefore, the first part of the thesis work provided a contribution in assessing the soil – structure interaction effect (SSI, influence of built structures in modifying the ground motion during earthquake shaking) between urban soil and all the overlying buildings in the city of Matera by characterising all the foundation soils of the urban area and all the overlying buildings. A geo-database, the CLARA WebGIS portal (available at this link: https://smartcities-matera-clara.imaa.cnr.it/), for storing and sharing the data and results collected during my PhD activity has been implemented with 488 pre-existing geological, geotechnical, geophysical data. CLARA WebGIS is the first useful tool for predicting which and how many buildings could suffer higher damage due to the double soil-building resonance effect and is the first open geo-platform that shares the results of the double soil-building resonance from experimental data for an entire urban area. CLARA WebGIS addresses a wide range of end-users (local administrations, engineers, geologists, etc.) as support for the implementation of seismic risk mitigation strategies in terms of urban planning, seismic retrofit, and post-earthquake crisis management. The knowledge of the spatial distribution of the site effects (modifications of the ground motions due to changes in the shallow geological layers) in terms of amplification effect, the primary characteristics of buildings, and of soil-building resonance levels estimations, a three-part objective have been achieved: (i) through CLARA's WebGIS every citizen is aware of the characteristics of buildings and foundation soils, so this knowledge makes each individual citizen more resilient to the effects of a seismic event; (ii) preventing the potential losses in economic and social terms; (iii) reducing recovering phase time to facilitate the return of the urban system to equilibrium pre-existing conditions. A deepening of this first study was made by specialising the linear period-height relationship derived from the experimental results as a function of the construction typology and foundation soil for unreinforced load-bearing masonry buildings (URM) founded on rigid soil (Gravina calcarenite characterised by flat HVNSR curves). This relationship is more representative of the condition of a fixed-base masonry building. Variations in the dynamic response of masonry buildings due to soil-foundation-structure interaction at urban scale can be evaluated by simplified analytical approaches based on the traditional compliant-base oscillator model and on simplified assumptions about the geometry and mechanical properties of the soil and foundations. The experimental period-height relationship for URM buildings founded on Gravina calcarenite were integrated in a simplified analytical procedure extended to complex and more realistic stratified soils and irregular foundation geometry. The modified simplified procedure were applied at an urban scale to predict the fundamental period of seven masonry buildings studied in the historic centre of Matera, for which all soil and structural data necessary for the analytical model were available. The comparison of the fundamental periods obtained with the three approaches, traditional, simplified-modified, and experimental, shown that the adoption of the simplified-modified approach significantly improved the agreement between the experimental and analytical periods. This part of the thesis work therefore appears promising to encourage an extended application of the analytical and experimental techniques to other historic urban area characterised by similar characteristics of the built heritage and soil stratification. In the second study of the thesis, has been implemented a multi-methodological approach that allowed to estimate the main modal parameters of the Gravina bridge by analysing short duration ambient noise signals (less than two hours) recorded by low-cost and non-invasive sensors and by performing dynamic tests. The Gravina is an arch bridge located on outcropping limestone in the city of Matera and spans 144 m along a steel-concrete deck suspended by two tubular steel arches. Ambient seismic noise was recorded using two acquisition configurations on the deck and inside the arch. The noise signal data were processed by applying: the standard spectral analysis (FFT), to examine frequencies and energy content distribution, a spectral ratio method with reference station, the Standard Spectral Ratio (SSR) technique, to check and validate eigenfrequencies, the Operational Modal Analysis (OMA) technique, i.e., the Frequency Domain Decomposition (FDD) method, to derive eigenfrequencies and mode shapes, and a seismic interferometric method, the Ambient Noise Deconvolution Interferometry (ANDI), to derive the propagation velocity of ambient noise in the infrastructure. Six eigenfrequencies have been estimated on the deck. The examination of the energy content distribution played a key role for the interpretation of the mode shapes. The variation of the eigenfrequencies of the infrastructure with the seasons as a function of temperature (°C) were monitored: the frequency variations are less than 5% and the behaviour of the structure do not exhibit degradation since the Gravina Bridge is a newly constructed road infrastructure. Deconvolution interferometry has been applied on the ambient noise signals recorded on the deck deriving the wave propagation velocity on the infrastructure. The results presented showed that the ANDI method is sensitive to the distribution of infrastructure stiffness. The multi-methodological approach used in this part of the thesis is promising for (i) evaluating the behaviour of standard structure like buildings and critical infrastructure like a bridge at different scales (global and local), (ii) examining variation of eigenfrequencies, mode shapes and ambient noise waves propagation velocities as a result of aging, degradation, and/or occurrence of potential damage, (iii) controlling and validating outcomes comparing the results obtained from different techniques, (iv) supporting at an early stage as a quick, non-invasive, low-cost tool applied without either diverting, blocking the traffic flow, or stopping the infrastructure service.