Abstract: Currently, even in socioeconomically deprived areas, useful data for earthquake risk assessment is present. The problem in fact is that those data need extensive preprocessing before being used for risk analyses. Although this situation preclude their use for quick response in emergency situations, detailed studies can be conducted with ample time allowed. An important issue in making these data useful is the support and participation of all institutions managing relevant data, because their improvement depends on our capacity to combine and interrelate a variety of data sources. Here imagination and data processing capacities are important skills. After the April 16th 2016 earthquake, a detailed study was carried out at Portoviejo (Ecuador). Among others, one of the aims of this study was to understand the earthquake impact in Portoviejo’s buildings inventory. This work is focused in the methodologies developed to prepare the data for risk analysis. Such methodologies, in addition to ad-hoc corrections, includes algorithms to a) correct the geographical coordinates related to the damage assessment, b) get the exposure and reposition costs for economic losses, c) characterize the buildings (their plan shape and regularity, level of isolation from other buildings, orientation). The procedures were applied mostly in the urban area, where roughly 80000 structures are present, being conducted at the individual level, namely building by building and it took one and a half year. Data processing applications were developed in ANSI C code language, combined with scripts in Python and R, which allows to recalculate the results very efficiently. The software also compute basic statistics, calculations used to describe the before-the-earthquake city and to correlate the damage with different cadastral, morphometric and on-site variables. This study also highlights the damage assessment and form damage scenario calculations.
Slidecast:
Abstract: In this study we investigate details of spatial and temporal evolution of the stress field and damage at a pre-existing fault surface during laboratory stick-slip friction experiments performed on Westerly Granite samples. Specimens were deformed at constant strain rate of 3×10-6 s-1 and confining pressures of 150 MPa. Here we analyze a series of 6 macroscopic slip events, each associated with intense microseismic activity. The Acoustic Emission (AE) events were recorded using a 16-channel transient recording system. Monitoring and mapping AE properties allowed recovering spatiotemporal damage and stress evolution. We investigated source characteristics (magnitude, seismic moment tensors and focal mechanisms) as well as the statistical properties (b-, c-, d- value) of microseismicity to unravel the micromechanical processes governing nucleation and propagation of slip events. In addition, the calculated AE focal mechanisms were used to derive time-dependent local stress orientations, stress shape ratio, and additional parameters such as scaled shear traction, quantifying proximity to failure of individual fault patches. The calculated characteristics are used to evidence the clear complexity of the preparatory and post-slip damage and stress evolution framing the macroscopic slip in the microscale. The observed fault processes and characteristics are discussed in the context of global strain and stress changes, fault surface maturation (roughness), and earthquake stress drop.
Poster:
WED.Brickell.1515.Bohnhoff
Abstract: The Hayward Fault is an active strike-slip fault trending southeast to northwest through the east bay region of California’s San Francisco Bay Area, which has caused approximately 100 km of offset during the last 10 Ma. Along a 30 km segment, between the City of Oakland in the north and the City of Fremont in the south, the Hayward Fault cuts through a tectonic body of Mesozoic gabbroic and volcanic rocks known as the San Leandro Block. In 2013, the United States Geological Survey (USGS) and the California State University East Bay (CSUEB) collaborated on the East Bay Seismic Experiment (EBSE) in order to examine the structure of fault zones in the vicinity of the CSUEB main campus. Sparse-data P-wave tomography from the EBSE survey, across the San Leandro Block at a study area on an undeveloped parcel in the City of Hayward, identifies the probable orientation of the Hayward Fault and an unnamed fault in the shallow subsurface of that study area. However, the USGS geophysics group from the Menlo Park Earthquake Science Center is interested in developing high-resolution seismic velocity models and tomographic images across the Hayward Fault Zone at the same study area. On November 18, 2017, the USGS geophysics group in collaboration with researchers from CSUEB coordinated a P-wave seismic survey across the study area. We hand-augured 63 ~ 45 cm deep boreholes across a 310 m long survey line as placement locations for a Betsy Seisgun seismic source; we then hand-augured 3 additional ~ 2 m deep boreholes as placement locations for larger charge. 4.5 Hz P-wave geophones were placed at each of the 63 Betsy source locations. Our 2017 survey produced a much larger data set than the previous 2013 EBSE survey, which used only 3 geophones. Assignment of P-wave first arrivals was completed by graduate student Collin Quesenberry using SeisImager™ PickWin software supplied by CSUEB. The resultant P-wave velocity data will be processed into a tomographic velocity model with the aid of the USGS. One of our immediate conclusions is the topsoil at our location is highly unconsolidated and caused unusual damping of Seisgun generated P-waves. In response to this first conclusion, we are planning a similar seismic survey across the same study area using a more powerful seismic source than the Betsy Seisgun. Results of this project will be shared between the USGS and CSUEB, and used for Collin Quesenberry’s graduate thesis.
Media has not been submitted for this Presentation
Abstract: The Mw 8.2 September 8 earthquake occurred in the middle of the “Tehuantepec Gap,” a segment of the Mexican subduction zone that has no historical mentions of a large earthquake. It was, however, not the expected subduction megathrust earthquake, but rather an intraplate, normal faulting event, in the subducting oceanic Cocos plate. We inverted for the slip on the fault plane, using; 1) local strong motion and high rate GPS records and 2) teleseismic body and surface waves, together with static GPS offsets. From the hypocenter at a depth of 50 km, the rupture propagated NW on a near-vertical plane, breaking towards the surface. Most of the slip was concentrated in the distance range 30-100 km from the hypocenter and at depth between 15 and 50 km, with maximum slip of ~15m. The earthquake seems to have broken the entire lithosphere, estimated to be 35 km based on the plate age. The strike of the fault is sub- parallel to the trench, aligned with the existing fabric on the incoming plate, suggesting a reactivation of previous structures. We relocated the aftershocks and found that they occurred along the fault plane during the first day after the event, with activation of other parallel structures within the subducting plate, towards the east, as well as in upper plate, in the following days. Coulomb stress modeling suggests that the stress on the plate interface, updip of the earthquake, is reduced. There are several other examples of large intraslab normal faulting earthquakes, near the downdip edge (1931 Mw 7.8 and 1999 Mw 7.5, Oaxaca) or directly below (1997 Mw 7.1, Michoacan) the coupled plate interface, along the Mexican subduction zone. The possibility of events of similar magnitude to the 2017 earthquake occurring close to the coastline, all along this part of the subduction zone, cannot be ruled out.
Slidecast:
Abstract: This article describes the developing of a computer tool to evaluate the liquefaction potential in seismic zones. The tool consists of: i) a pre-processing script, ii) calculation or process script, and iii) a post-processing script. The pre and post-processing stages are carried out by using the Matlab software through the GUIDE development environment, while the process uses OpenSees as a calculation engine. The layered deposit of cohesionless soil subject to a seismic base excitation is represented by a 2D finite element model using nine-node quadrilateral elements by considering the displacement and pressure of pores as degrees of freedom. The validation process showed that the generated soil column model properly represents the problem posed. The application of this tool was done by modeling 1325 samples corresponding to 14 theoretical scenarios in which the following variables were considered and permuted: depth of the soil column, stratigraphy, groundwater level, relative soil density, and intensity of seismic load. All this, in order to know the degree of influence of each of these parameters on the liquefaction phenomenon. The results showed that the phreatic level and the relative density of the soil are the most sensitive variables, which are in accordance with the theoretical foundations studied. Additionally, there was geotechnical information from two soil profiles of the Tarqui parish, Ecuador (area affected by soil liquefaction during the 2016 Ecuador earthquake). This information was used to evaluate and verify the susceptibility of these soils before the aforementioned phenomenon. Finally, It is worth mention that this tool is part of the Virtual Laboratory of the Technical University of Loja (UTPL), which is managed by its Research Group of Seismic Engineering and Seismology (GRISSUTPL)(http:www.ingenieriasismica.utpl.edu.ec/) keywords: Ecuador, Liquefaction of soils, earthquake, OpenSees, numerical modeling.
Poster:
Edwin Duque - Presentación_Edwin Duque_ENG
Abstract: Lightning often occurs during ash-producing eruptive activity and its detection is now being used in volcano monitoring for rapid alerts. We report on infrasonic and sonic recordings of the related, but previously undocumented, phenomenon of volcanic thunder. We observe volcanic thunder during the waning stages of two explosive eruptions at Bogoslof volcano, Alaska, on a microphone array located 60 km away. Thunder signals arrive from a different direction than co-eruptive infrasound generated at the vent following an eruption on June 10, 2017, consistent with locations from lightning networks. For the March 8, 2017 eruption, arrival times and amplitudes of high frequency thunder signals correlate well with the timing and strength of lightning detections. In both cases, the thunder is associated with lightning that continues after significant eruptive activity has ended. Thus, the optimal observation time is during the minutes immediately after the cessation of eruptive activity when volcanic lightning continues in the detached plume – when the thunder signal is not masked by eruption noise. Of all the events in the 2016-17 Bogoslof eruption sequence, the March 8 and June 10 eruptions had the clearest volcanic thunder signals. Observations for other eruptive events with lightning detections were hampered by higher levels of background noise, non-optimal wind conditions, and volcanic activity that did not abruptly terminate. Further work is needed on the characteristics of volcanic thunder and its relation to lightning properties, such as frequency content of thunder and the nature of the lightning (intercloud or cloud-to-ground). These infrasonic and sonic observations of volcanic thunder offer a new avenue for studying electrification processes in volcanic plumes, provide ways of distinguishing volcanic thunder from eruption acoustic signals, and motivate improved methods of measuring volcanic thunder at close range and during explosive activity.
Poster:
WED.Brickell.1715.Haney
