Earthquake detection is an important task, focusing on detecting seismic events in past data or in real time from seismic time series. In the past few decades, due to the increasing amount of available seismic data, research in seismic event detection shows remarkable success using neural networks and other machine learning techniques. However, creating high quality labeled data sets is still a manual process that demands tremendous amount of time and expert knowledge, and is stifling big data innovation. When compiling a data set, it is unclear how many earthquakes and noise are mislabeled. Another challenge is how to promote the general applicability of the machine learning based models to different geographical regions. The models trained by data sets from one location should be applicable to the detection at other locations. This thesis explores the most popular deep learning model, convolutional neural networks (CNN), to build a single location detection model. In addition, we build more robust generalized earthquake detection models using transfer learning and meta learning. We also introduce a process for generating high quality labeled datasets. Our technique achieves high detection accuracy even on low signal to noise ratio events. The AI techniques explored in this research have potential to be transferred to other domains that utilize signal processing. There are a myriad of potential applications, with audio processing probably being one of the most directly relevant. Any field that deals with waveforms (e.g. seismic, audio, light) can utilize the developed techniques.

The Kumamoto earthquake of April 2016 produced two foreshocks of moment magnitude 6.0 and 6.2 and a mainshock of 7.0, which should have been followed by widespread and intense soil liquefaction. A Geotechnical Extreme Events Reconnaissance team (GEER) led by Professor Rob Kayen of UC Berkley was dispatched to the Kumamoto Plain--which is in Kumamoto Prefecture, the southern main island of Japan--immediately following the earthquake. The Japanese and U.S. engineers in the GEER team observed mostly minor and sporadic liquefaction, which was unexpected as the local site geology, known soil stratigraphy, and intensity of the seismic loading made the Kumamoto Plain ripe for soil liquefaction. The paucity and limited scale of liquefaction shows a clear gap in our understanding of liquefaction in areas with volcanic soils. This study is a direct response to the GEER team's preliminary findings regarding the lack of significant liquefaction. An extensive literature review was conducted on the Kumamoto Plain and its volcanic soil. The liquefaction of the 2016 Kumamoto Earthquake was also researched, and several sites were selected for further analysis. Four sites were analyzed with SPT, CPT, and laboratory testing during the spring of 2017. A slope stability analysis and undisturbed testing were performed for specific sites. The results of the analysis show a general over-prediction of SPT and CPT methods when determining liquefaction hazard. The Youd et al. (2001) NCEES method was the most consistent and accurate in determining liquefaction. The soils in the area including sands and gravels had high levels of fines, plasticity, and organic matter due to the weathering of volcanic ash and pyroclastic material. The volcanically derived coarse-grained soils may also have exhibited some crushability, which gave lower resistance readings. Filled river channels had the worst liquefaction with natural levees and the Kumamoto flood plains having only minor liquefaction. Publicly available boring logs rarely showed laboratory test data of bore holes which led to a general inaccurate soil classification. Boring logs were also not updated with laboratory classifications and data. Undisturbed cyclic triaxial testing of soils at one site showed that volcanic soils had relatively high resistance to soil liquefaction, though drying of samples may have compromised the results. Embankment cracking at one test location was calculated a lateral spread and a seismic slope failure along the pyroclastic flow deposit.

Thesis advisor: John E. Ebel / Determination of the source parameters of a local earthquake from full seismic waveforms requires seismograms with clear body-wave signals from the earthquake source. Coherence of the earthquake body-wave seismograms recorded at two different receivers can be used to estimate the signal-to-noise ratio (SNR) of the body-wave energy radiated by the source. In this study, the coherence of earthquake body waves recorded in the Northeastern United States and Southeastern Canada (NEUSSEC) is measured as a function of frequency, interstation distance, and ambient SNR, and then used as an estimate of body-wave SNR. Seismograms from the CN, IU, LD, N4, NE, TA, and US arrays were used to measure coherence between stations with a mean separation of 70 km. Seismograms from the Acton Littleton Seismic Array (ALSA) were used to measure coherence at 5 km mean station separation. Coherence is measured at frequencies between 0.05-10 Hz for Pn and Sn phases from NEUSSEC earthquakes with magnitudes (M) between 0.0 and 4.7 at epicentral distances between 180-1800 km as well as at frequencies between 0.05-10 Hz for the first arrivals of P and S waves from earthquakes M>6 at distances >2500 km. The teleseismic P waves display values of coherence greater than 0.9 out to interstation distances of 1500 km at frequencies <0.8 Hz, but as frequency increases, the interstation distance at which coherence falls below 0.9 decreases. Teleseismic S and regional Pn and Sn waves display coherence values around 0.5, suggesting the amplitudes of the body-waves are smaller than those of the noise, which likely is the result of converted and reflected or refracted P waves and/or smaller signal amplitudes. These coherence values are compared to the coherence values of ambient noise. For any two P, S, Pn or Sn waveforms recorded in the NEUSSEC at 3-5 Hz there is a 50% or greater chance of those two waveforms containing coherent energy that is not ambient noise; these frequencies are where this percent chance is greatest for all seismic phases. At frequencies between 3-5 Hz the effects of scattering are most pronounced on the coherence values of regional seismic phases, suggesting that most scattering in the crust of the NEUSSEC takes place at these frequencies. Teleseismic seismic phases do not include as much scattered energy as the regional seismic phases at 3-5 Hz, and must therefore encounter fewer scattering heterogeneities along their travel path than the regional seismic phases. / Thesis (MS) — Boston College, 2021. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Earth and Environmental Sciences.

This research consists of full-scale cyclic shake table tests to investigate liquefied residual strength of #2/16 Monterey Sand. A simple shear testing apparatus was mounted to a full-scale one-dimensional shake table to mimic a confined layer of saturated sand subjected to strong ground motions. Testing was performed at the Parson’s Geotechnical and Earthquake Laboratory at California Polytechnic State University, San Luis Obispo. T-bar penetrometer pullout tests were used to measure residual strength of the liquefied soil during cyclic testing. Cone Penetration Testing (CPT) was performed on the soil specimen throughout testing to relate the laboratory specimen to field index test data and to compare CPT results of the #2/16 Monterey sand before and after liquefaction. The generation and dissipation of excess pore pressures during cyclic motion are measured and discussed. The effects of liquefied soil on seismic ground motion are investigated. Measured residual strengths are compared to previous correlations comparing liquefied residual strength ratios and CPT tip resistance.

Soil liquefaction, or loss of strength due to excess pore water pressures generated during dynamic loading, is a main cause of damage during earthquakes. When a soil liquefies (referred to as triggering), it may lose its ability to support overlying structures, deform vertically or laterally, or cause buoyant uplift of buried utilities. Empirical liquefaction models, used to predict liquefaction potential based upon in-situ soil index property measurements and anticipated level of seismic loading, are the standard of practice for assessing liquefaction triggering. However, many current models do not incorporate predictor variable uncertainty or do so in a limited fashion. Additionally, past model creation and validation lacks the same rigor found in predictive modeling in other fields. This study examines the details of creating and validating an empirical liquefaction model, using the existing worldwide cone penetration test liquefaction database. Our study implements a logistic regression within a Bayesian measurement error framework to incorporate uncertainty in predictor variables and allow for a probabilistic interpretation of model parameters. Our model is built using a hierarchal approach account for intra-event correlation in loading variables and differences in event sample sizes that mirrors the random/mixed effects models used in ground motion prediction equation development. The model is tested using an independent set of case histories from recent New Zealand earthquakes, and performance metrics are reported. We found that a Bayesian measurement error model considering two predictor variables, qc,1 and CSR, decreases model uncertainty while maintaining predictive utility for new data. Two forms of model uncertainty were considered – the spread of probabilities predicted by mean values of regression coefficients (apparent uncertainty) and the standard deviations of the predictive distributions from fully probabilistic inference. Additionally, we found models considering friction ratio as a predictor variable performed worse than the two variable case and will require more data or informative priors to be adequately estimated.

<p>У овом раду анализирани су утицају интеракције конструкција-темељ-<br />тло током сеизмичких дејстава. Примењене су нелинеарне анализе<br />временске историје (TH) и пушовер метода. Анализирани су модели<br />гредних мостова, на каркатеристичним локацијама и са оптерећењима и<br />материјалима према БАБ&#39;87. Интеракција је третирана преко<br />еластичних опруга и као нелинеарнa преко линк елемената p-y кривама.<br />Криве су моделоване вишелинијским Тaкеда хистерезисним моделом.<br />Проучено је стање конструкција у току и након земљотреса, укључиво и<br />други удар земљотреса. У мањем обиму су анлизиране конструкције<br />зграда према ЕЦ2 и ЕЦ8.</p> / <p>U ovom radu analizirani su uticaju interakcije konstrukcija-temelj-<br />tlo tokom seizmičkih dejstava. Primenjene su nelinearne analize<br />vremenske istorije (TH) i pušover metoda. Analizirani su modeli<br />grednih mostova, na karkaterističnim lokacijama i sa opterećenjima i<br />materijalima prema BAB&#39;87. Interakcija je tretirana preko<br />elastičnih opruga i kao nelinearna preko link elemenata p-y krivama.<br />Krive su modelovane višelinijskim Takeda histerezisnim modelom.<br />Proučeno je stanje konstrukcija u toku i nakon zemljotresa, uključivo i<br />drugi udar zemljotresa. U manjem obimu su anlizirane konstrukcije<br />zgrada prema EC2 i EC8.</p> / <p>The thesis studies the effects of the structure-foundation- soil interaction<br />during seismic action using non-linear analysis methods Time History and<br />PushOver. Analysis is made of beam bridges, in their typical locations and<br />with loads and construction materials based on BAB&rsquo;87/-Yu Code. The<br />interaction is treated as linear using elastic spring model, and non-linear as<br />link element on p-y curves which are modeled using the multi-linear Takeda<br />hystheresis model. Studies are made on the performance of the<br />constructions during and after an earthqake including aftershocks. The work,<br />to a lesser extent, also includes similar analyses and studies for buildings,<br />based on EC2 and EC8 codes.</p>

Studying recent volcanic and tectonic events in the Red Sea region is important for improving our knowledge of the Red Sea plate boundary and for regional geohazard assessments. However, limited information has been available about the past activity due to insufficient in-situ data and remoteness of some of the activity. In this dissertation, I have used satellite remote sensing to derive new information about several recent volcanic and tectonic events in the Red Sea region. I first report on three volcanic eruptions in the southern Red Sea, the 2007-8 Jebel at Tair eruption and the 2011-12 & 2013 Zubair eruptions, which resulted in formation of two new islands. Series of high- resolution optical images were used to map the extent of lava flows and to observe and analyze the growth and destructive processes of the new islands. I used Interferometric Synthetic Aperture Radar (InSAR) data to study the evolution of lava flows, to estimate their volumes, as well as to generate ground displacements maps, which were used to model the dikes that fed the eruptions. I then report on my work of the 2009 Harrat Lunayyir dike intrusion and the 2004 Tabuk earthquake sequence in western Saudi Arabia. I used InSAR observations and stress calculations to study the intruding dike at Harrat Lunayyir, while I combined InSAR data and Bayesian estimation to study the Tabuk earthquake activity. The key findings of the thesis are: 1) The recent volcanic eruptions in the southern Red Sea indicate that the area is magmatically more active than previously acknowledged and that a rifting episode has been taken place in the southern Red Sea; 2) Stress interactions between an ascending dike intrusion and normal faulting on graben-bounding faults above the dike can inhibit vertical propagation of magma towards the surface; 3) InSAR observations can improve locations of shallow earthquakes and fault model uncertainties are useful to associate earthquake activity with mapped faults; 4). The successful application of satellite remote sensing technologies in studying the recent volcanic and tectonic processes in the Red Sea region implies that remote sensing data are an important resource for the local authorities to monitor geohazards.

Please note: creative writing theses are permanently embargoed in OpenBU. No public access is forecasted for these. To request private access, please click on the locked Download file link and fill out the appropriate web form. / Poetry / 2031-01-01

This qualitative study utilized the protective action decision model to explore the risk perceptions and hazard adjustments to the earthquake risk of residents in El Reno, Oklahoma.

No description available.