Geospatial liquefaction models aim to predict liquefaction using data that is free and readily-available. This data includes (i) common ground-motion intensity measures; and (ii) geospatial parameters (e.g., among many, distance to rivers, distance to coast, and Vs30 estimated from topography) which are used to infer characteristics of the subsurface without in-situ testing. Since their recent inception, such models have been used to predict geohazard impacts throughout New Zealand (e.g., in conjunction with regional ground-motion simulations). While past studies have demonstrated that geospatial liquefaction-models show great promise, the resolution and accuracy of the geospatial data underlying these models is notably poor. As an example, mapped rivers and coastlines often plot hundreds of meters from their actual locations. This stems from the fact that geospatial models aim to rapidly predict liquefaction anywhere in the world and thus utilize the lowest common denominator of available geospatial data, even though higher quality data is often available (e.g., in New Zealand). Accordingly, this study investigates whether the performance of geospatial models can be improved using higher-quality input data. This analysis is performed using (i) 15,101 liquefaction case studies compiled from the 2010-2016 Canterbury Earthquakes; and (ii) geospatial data readily available in New Zealand. In particular, we utilize alternative, higher-quality data to estimate: locations of rivers and streams; location of coastline; depth to ground water; Vs30; and PGV. Most notably, a region-specific Vs30 model improves performance (Figs. 3-4), while other data variants generally have little-to-no effect, even when the “standard” and “high-quality” values differ significantly (Fig. 2). This finding is consistent with the greater sensitivity of geospatial models to Vs30, relative to any other input (Fig. 5), and has implications for modeling in locales worldwide where high quality geospatial data is available.
Overview of SeisFinder SeisFinder is an open-source web service developed by QuakeCoRE and the University of Canterbury, focused on enabling the extraction of output data from computationally intensive earthquake resilience calculations. Currently, SeisFinder allows users to select historical or future events and retrieve ground motion simulation outputs for requested geographical locations. This data can be used as input for other resilience calculations, such as dynamic response history analysis. SeisFinder was developed using Django, a high-level python web framework, and uses a postgreSQL database. Because our large-scale computationally-intensive numerical ground motion simulations produce big data, the actual data is stored in file systems, while the metadata is stored in the database. The basic SeisFinder architecture is shown in Figure 1.
Hybrid broadband simulation methods typically compute high-frequency portion of ground-motions using a simplified-physics approach (commonly known as “stochastic method”) using the same 1D velocity profile, anelastic attenuation profile and site-attenuation (κ0) value for all sites. However, these parameters relating to Earth structure are known to vary spatially. In this study we modify this conventional approach for high-frequency ground-shaking by using site-specific input parameters (referred to as “site-specific”) and analyze improvements over using same parameters for all sites (referred to as “generic”). First, we theoretically understand how different 1D velocity profiles, anelastic attenuation profiles and site-attenuation (κ0) values affects the Fourier Acceleration Spectrum (FAS). Then, we apply site-specific method to simulate 10 events from the 2010-2011 Canterbury earthquake sequence to assess performance against the generic approach in predicting recorded ground-motions. Our initial results suggest that the site-specific method yields a lower simulation standard deviation than generic case.
The influence of nonlinear soil-foundation-structure interaction (SFSI) on the performance of multi-storey buildings during earthquake events has become increasingly important in earthquake resistant design. For buildings on shallow foundations, SFSI refers to nonlinear geometric effects associated with uplift of the foundation from the supporting soil as well as nonlinear soil deformation effects. These effects can potentially be beneficial for structural performance, reducing forces transmitted from ground shaking to the structure. However, there is also the potential consequence of residual settlement and rotation of the foundation. This Thesis investigates the influence of SFSI in the performance of multi-storey buildings on shallow foundations through earthquake observations, experimental testing, and development of spring-bed numerical models that can be incorporated into integrated earthquake resistant design procedures. Observations were made following the 22 February 2011 Christchurch Earthquake in New Zealand of a number of multi-storey buildings on shallow foundations that performed satisfactorily. This was predominantly the case in areas where shallow foundations, typically large raft foundations, were founded on competent gravel and where there was no significant manifestation of liquefaction at the ground surface. The properties of these buildings and the soils they are founded on directed experimental work that was conducted to investigate the mechanisms by which SFSI may have influenced the behaviour of these types of structure-foundation systems. Centrifuge experiments were undertaken at the University of Dundee, Scotland using a range of structure-foundation models and a layer of dense cohesionless soil to simulate the situation in Christchurch where multi-storey buildings on shallow foundations performed well. Three equivalent single degree of freedom (SDOF) models representing 3, 5, and 7 storey buildings with identical large raft foundations were subjected to a range of dynamic Ricker wavelet excitations and Christchurch Earthquake records to investigate the influence of SFSI on the response of the equivalent buildings. The experimental results show that nonlinear SFSI has a significant influence on structural response and overall foundation deformations, even though the large raft foundations on competent soil meant that there was a significant reserve of bearing capacity available and nonlinear deformations may have been considered to have had minimal effect. Uplift of the foundation from the supporting soil was observed across a wide range of input motion amplitudes and was particularly significant as the amplitude of motion increased. Permanent soil deformation represented by foundation settlement and residual rotation was also observed but mainly for the larger input motions. However, the absolute extent of uplift and permanent soil deformation was very small compared to the size of the foundation meaning the serviceability of the building would still likely be maintained during large earthquake events. Even so, the small extent of SFSI resulted in attenuation of the response of the structure as the equivalent period of vibration was lengthened and the equivalent damping in the system increased. The experimental work undertaken was used to validate and enhance numerical modelling techniques that are simple yet sophisticated and promote interaction between geotechnical and structural specialists involved in the design of multi-storey buildings. Spring-bed modelling techniques were utilised as they provide a balance between ease of use, and thus ease of interaction with structural specialists who have these techniques readily available in practice, and theoretically rigorous solutions. Fixed base and elastic spring-bed models showed they were unable to capture the behaviour of the structure-foundation models tested in the centrifuge experiments. SFSI spring-bed models were able to more accurately capture the behaviour but recommendations were proposed for the parameters used to define the springs so that the numerical models closely matched experimental results. From the spring-bed modelling and results of centrifuge experiments, an equivalent linear design procedure was proposed along with a procedure and recommendations for the implementation of nonlinear SFSI spring-bed models in practice. The combination of earthquake observations, experimental testing, and simplified numerical analysis has shown how SFSI is influential in the earthquake performance of multi-storey buildings on shallow foundations and should be incorporated into earthquake resistant design of these structures.
This thesis presents the application of data science techniques, especially machine learning, for the development of seismic damage and loss prediction models for residential buildings. Current post-earthquake building damage evaluation forms are developed for a particular country in mind. The lack of consistency hinders the comparison of building damage between different regions. A new paper form has been developed to address the need for a global universal methodology for post-earthquake building damage assessment. The form was successfully trialled in the street ‘La Morena’ in Mexico City following the 2017 Puebla earthquake. Aside from developing a framework for better input data for performance based earthquake engineering, this project also extended current techniques to derive insights from post-earthquake observations. Machine learning (ML) was applied to seismic damage data of residential buildings in Mexico City following the 2017 Puebla earthquake and in Christchurch following the 2010-2011 Canterbury earthquake sequence (CES). The experience showcased that it is readily possible to develop empirical data only driven models that can successfully identify key damage drivers and hidden underlying correlations without prior engineering knowledge. With adequate maintenance, such models have the potential to be rapidly and easily updated to allow improved damage and loss prediction accuracy and greater ability for models to be generalised. For ML models developed for the key events of the CES, the model trained using data from the 22 February 2011 event generalised the best for loss prediction. This is thought to be because of the large number of instances available for this event and the relatively limited class imbalance between the categories of the target attribute. For the CES, ML highlighted the importance of peak ground acceleration (PGA), building age, building size, liquefaction occurrence, and soil conditions as main factors which affected the losses in residential buildings in Christchurch. ML also highlighted the influence of liquefaction on the buildings losses related to the 22 February 2011 event. Further to the ML model development, the application of post-hoc methodologies was shown to be an effective way to derive insights for ML algorithms that are not intrinsically interpretable. Overall, these provide a basis for the development of ‘greybox’ ML models.
