Abstract: In 2015, the Central American countries (Guatemala, El Salvador, Honduras, Nicaragua, Costa Rica and Panama) agreed to establish a Central American Tsunami Advisory Center (CATAC) at INETER in Managua, Nicaragua. This proposal was also approved by the Intergovernmental Oceanographic Committee (IOC) of UNESCO and the Intergovernmental Coordination Groups (ICG) of the Pacific Ocean Tsunami Warning Systems (PTWS) and the Caribbean Sea (CARIBEEWS). CATAC will provide scientific real-time technical tsunami services related to the Pacific Ocean and the Caribbean Sea to the Scientific Institutions and Civil Protection Agencies of the Central American countries. The issuance of tsunami advice to the population remains the responsibility of the national governments. In 2016, INETER started to establish CATAC. The Nicaraguan government agreed with the Japan International Cooperation Agency (JICA) to execute a technical assistance project to strengthen CATAC; this project began in October 2016. Progress was made in the development of the data processing system, the monitoring networks, the construction of the tsunami database, the training of personnel and the development of Standard Operating Procedures (SOP).
Poster:
2018.05_LACSC-SSA_Furukawa_etal
Abstract: On September 7th, 2017, a Mw 8.2 earthquake occurred off the coast of Chiapas (Mexico), inside the subducting Cocos slab. The normal mechanism event started at a depth of 50 km (SSN, 2017), causing a small tsunami. The latter was recorded at the central and southern stations of the Servicio Mareografico Nacional of Mexico (SMN), with a maximum height of 3.42 m at Puerto Chiapas station. We simulated the tsunami using the Carnegie Mellon Finite Element Toolchain Hercules (Tu et al., 2006) to obtain the dynamic seafloor deformation (SFD) for an area of 235 km2 around the rupture area; considering the finite source model by Melgar et al. (2017). The SFD was introduced into the GeoClaw software (Clawpack Development Team, 2016) to compute the tsunami wave propagation. Comparisons of the observed and computed tsunami waveforms reveal a fair fit at the Salina Cruz and Puerto Madero stations. Additionally, we performed two simulations for shallow inverse faults, and perform a preliminary comparison of the tsunami inundation against the one caused by the September 7th, 2017 quake at the Chiapas-Oaxaca coast.
Slidecast:
https://vimeo.com/276518667
Abstract: The 19 September 2017 earthquake, which occurred in Central Mexico, caused severe damage to important cities of Central Mexico including Mexico City. Authorities reported that 369 persons were killed by the earthquake (more than 60% in the Mexico City) and hundreds of buildings collapsed or were seriously damaged. Hypocentral location reveals an intermediate-depth earthquake located inside the subducted Cocos plate. Using regional broad band data (400 km < R < 900 km) we inverted for the Regional Moment Tensor (RMT) and obtained the following solution: NP1 =303°, =45°, =-79°; NP2 =107°, =47°, =-101°, similar to that reported by GCMT: NP1 =300°, =44°, =-83°; NP2 =149°, =46°, and =-97°. RMT yields a scalar moment of 4.22×1026 dyn-cm as compared to 6.51×1026 dyn-cm reported by GCMT. Three portable broad band stations were installed surrounding the epicenter of the mainshock within 48 hrs. Only 12 small earthquakes (M<4.1) were recorded in the next 2 months. We found that the hypocentral locations are better correlated with the fault plane dipping to northeast (=303°, =45°, and =-79). Local network also recorded shallow triggered seismicity which were located northwest of the epicenter at a distance < 100 km.
Slidecast:
https://vimeo.com/276522081
Abstract: For years, scientists have tried to determine whether periodicities exist in earthquake catalogs. These studies have ranged from global to local and looked for influence due to both diurnal tides and seasonality related to snowmelt or rainfall cycles. However, the standard test for periodicity, the Schuster Test, is valid only under the condition of homogeneity in the catalog. When the fundamental earthquake rate has changed over time, as is the case with many locations experiencing an increase in earthquakes due to wastewater injection from oil production activity, the heteroscedasticity prevents a direct application of the Schuster Test for interpretable results. Herein we formulate and validate a modified Schuster Test that is appropriate for use when the background rate of seismicity is changing. When applied to seismicity catalogs of induced seismicity in Oklahoma, we rule out a signature of tidal triggering and identify a number of spurious periodicities whose significance would be overestimated by the traditional Schuster Test.
Slidecast:
https://vimeo.com/276910716
Abstract: The rate of earthquakes across the United States mid-continent has dramatically increased since 2009. The historically high seismicity rates across the mid-continent have been largely driven by substantial increases in seismicity occurring within Oklahoma, including several magnitude 5.0+ earthquakes. Prior to 2009, background seismicity rate was about 2 M 3.0+ earthquakes per year; this increased to 579 and peaked at 903 M 3.0+ earthquakes in 2014 and 2015, respectively. The increase in Oklahoma seismicity was roughly coincident with an oil and gas boom focused around the Mississippian Limestone and Hunton Limestone. Those plays contain substantial amounts of co-produced formation brines. Common disposal practices involved disposing of wastewater in deep underground injection wells completed into upper parts of the basement and the karstic Arbuckle Group, which directly overlies the basement over most of Oklahoma. From 2010 to late 2014, statewide disposal rates increased from ~30 million bbls/month to ~90 million bbls/month. The increase in seismicity rate roughly correlates to the increase in monthly injection rates, though with a lag greater than a year or so in many sub-regions of Oklahoma. The statewide disposal is down to ~40 million bbls/month and the earthquakes have also declined, due to targeted and statewide reductions in permitted disposal directed by the Oklahoma Corporation Commission, to economic factors, and to changes in disposal practices, i.e. injecting into shallower units. As injection and the seismicity rate have declined in 2016 and 2017, we have observed continued seismic activity at greater well-to-earthquake distances, particularly long-duration foreshock and aftershock sequences, and aftershock sequences that seem depleted in larger magnitude aftershocks. In light of the observed time-lagged coherence between wastewater injection and seismicity rate, several groups strive to estimate the future seismic hazard in Oklahoma and consensus has yet to be reached. However, it appears that the seismic hazard likely remains high in the next several years and elevated over the next several decades. Thus, Oklahoma still needs significant scientific infrastructure and intellectual investment in order to make progress in understanding and mitigation of induced seismicity.
Slidecast:
https://vimeo.com/276526212
Abstract: Over the past 50 years global seismic detection has matured considerably due to the importance of differentiating seismic events: man-made (mining explosions) from natural phenomena (earthquakes, volcanoes). Where there is still some prevalent concern with these seismological advancements is in separating these classes. This concern derives from the fact that a seismic signature lacks easily defined features which complicates the discrimination process for a variety of detection methods. This paper aims to address this issue, by evaluating state-of-the-art machine learning algorithms, and compare several deep classification techniques. We target deep learning methods (Deep Neural Networks, Convolutional Neural Networks, and Long-Short Term Memory Networks) due to their ability to mimic human-like neural connections and support data-specific feature extraction to classify waveform signatures. Our proposed algorithms show noteworthy accuracy, opening up the realization of automated network seismic discrimination. Our results show evidence that deep learning should be considered the leading candidate for classifying seismic waveforms and will further advance the gap between machine learning and geological sciences.
Media has not been submitted for this Presentation
