2.6 Seismic Applications (SEI) - pdfs.semanticscholar.org€¦ · 2.6 Seismic Applications (SEI) The Seismic research area continued analysis of data captured by the Middle America

2.6 Seismic Applications (SEI)The Seismic research area continued analysis of data captured by the Middle America Seismic Experiment (MASE), analysis of the ongoing Peru Subduction Zone Experiment (PeruSZE), and successful testing of GeoNet, the Reftek ENSBox platform for both structural and seismic applications.MASE and PeruSEIn 2008 our wireless network that was developed and installed across Mexico (MASE Middle America Seismic Experiment) was shipped to Peru (PeruSE Peru Seismic Experiment) and installed along a line (Line 1) across the Andes between the cities of Mollendo and Puno, where the Nazca plate is subducting beneath the west coast generating devastating earthquakes and tsunamis (Figure 5). Graduate and undergraduate students were involved in the installation of the 49 station network. Most stations at the end of the summer were recording on-site. Subsequently Richard Guy and Igor Stubailo installed the networking that links the stations across the Andes. In the summer of 2009, along with colleagues from Caltech, 50 stand-alone Caltech stations were installed along the Puno-Cusco line (Line 2) along the Altiplano in the Andes. In late 2010, 25 stations from the Mollendo-Cusco line were moved to a new line (Line 3) Pisco-Cusco that parallels the first and completes a U shaped network of 100 stations linking the Altiplano to the coast. Data is available from line 1: 2008-present, Line 2: late 2009-present, Line 3: late 2010-present. It is anticipated the full network will run for at least another year.The Peru experiment (field work funded by the Caltech Moore Foundation grant) provided an opportunity to redesign our networking protocols based on our networking experience in Mexico. Our Delay Tolerant Shell (DTS) was improved. A new website for hourly system status was designed and implemented. The data was input to LabView. The various synergies of CENS have combined to make a significant remote area networking product. The data is radioed across Peru to Internet drops. It is then transmitted to UCLA over the Internet. The Atacama desert, where the network is located, is one of the more remote parts of the world. The facility that has been developed to install a remote wireless network over 250 km, and have the data transmit back to the laboratory, as well as duplex control on the instrumentation in the field, has application worldwide where remote network sensing is required.Exciting science has been discovered in the MASE data by Mexican graduate students, Luis Antonio Dominguez and Igor Stubailo, and in the Peru data by Emily Foote. All three presented papers at the Fall meeting of the American Geophysical Union.CENS Development of the Reftek EnsBox with application to GeoNet-SHMnet-ShakeNet-FlexiRAMPCENS has provided the infrastructure and technology to link UCLA departments in a single development with and industrial partner (Reftek Refraction Technology of Texas). It is the culmination of our experience in wireless networking to design a node that satisfies the digitizing and wireless networking requirements of the following groups to improve their science:

• GeoNet (geophysical monitoring) Paul Davis, Department of Earth and Space Sciences, UCLA, earthquake and tectonic networks.

• SHMnet (Structural Health monitoring) John Wallace, Civil Engineering, UCLA

• ShakeNet (Monitoring civil structures for shaking after earthquakes) Monica Kohler, CENS and Caltech, Ramesh Govindan, USC, Department of Computer science

Figure 5. Chile 8.8 earthquake in 2010 February 27 recorded by the Peru network. This unique on-scale recording of a huge earthquake by an array of broadband seismometers provided seismologists with an unprecedented view of the development of a mega-rupture.

2011 Annual Report! 117Center for Embedded Networked Sensing! 2.6 Seismic Applications

• FlexiRAMP, A flexible network for Rapid Array Mobilization Procedures, IRIS (Incorporated Research Institutions for Seismology) Richard Allen Berkeley, Marcos Alvarez, IRIS, and Paul Davis, UCLA. in association with Deborah Estrin, CENS

Through CENS, in a multidisciplinary project involving four UCLA Departments (Computer Science, Electrical Engineering, Civil Engineering, and Earth and Space Sciences), the PIs have designed a novel seismic node and have had two built by the leading manufacturer of seismic recorders (Refraction Technology of Texas). The GeoNet nodes incorporate a new generation digital acquisition system (DAS) based on the CENS-developed LEAP (low-power energy aware processing) system and a newly developed low-power A/D converter from Texas Instruments (TI). They will have application over a wide range of field applications involving wide area networks and low power processing and delivery of event data. They will run on small batteries with a laptop sized solar panel. Preliminary tests in the Mojave desert have been carried out in which the boxes recorded explosions from a seismic refraction experiment, detected them as events (Figure 6), aggregated the data using WiFi and used cell phone modems to email the data and plots. As was seen in recent damaging earthquakes in Japan and New Zealand, rapid knowledge of seismic activity is of critical importance to emergency planners.

Figure6. March 2011 test of Geonet boxes Salton Sea, California. Data between boxes was transmitted by WiFi. Events (lower left) were automatically detected and transmitted via cell phone modems.


SEI 01 Surface wave analysis of MASE data and integration with other experiments

Team Members• Caroline Beghein, Faculty• Paul Davis, Faculty*• Igor Stubailo, Graduate Student, PI*

* Primary Contact

OverviewWe present azimuthally anisotropic fundamental mode Rayleigh wave phase velocity maps as well as the phase velocity and anisotropy inversions for the Mexico area. This region is especially interesting to study because it contains both steep and flat subduction and a volcanic arc that is oblique to the trench. Isotropic and anisotropic 3D velocity structure is needed to infer strain associated with the development of the tectonics. Most shear wave splitting analysis performed in subduction zones display a fast direction of propagation for seismic waves oriented parallel to the trench near the trench, and trench-perpendicular in the backarc. The Mexico subduction zone, however, is an exception to this rule: shear wave splitting analysis performed in that region shows fast directions that are perpendicular to the trench with no significant difference between measurements made above the slab and in the backarc.

ApproachIn this study, we analyzed data recorded at 165 temporary and permanent broadband stations installed in Mexico and Southern USA over a period of one and a half years for teleseismic events of magnitude 6.0 and above. We employed a two-station method to measure phase velocity dispersion curves between periods of 16 s to 170 s, using events located within 3 degrees of the great circle path between each pair of stations. We then inverted the measurements to obtain azimuthally anisotropic phase velocity maps and from them got velocity and anisotropy model of the upper mantle.

AccomplishmentsOur results revealed lateral variations in phase velocities at all periods consistent with the presence of a flat subduction.We found phase velocities larger than average in the forearc, and lower than average near the Trans Mexican Volcanic Belt. At short periods (16-33 s), we found strong lateral variations in phase velocities and anisotropy. These are rather complex, and may reflect changes in crustal structure and deformation history. However, at periods of 38 s and higher, the phase velocity anomalies are smoother, with a transition between faster and slower than average phase velocities that follows the shape of the slab depth lines, both in the flat and in the steeper portions of the slab. This enabled us to further constrain the three-dimensional shape of the slab by inverting the dispersion curves and discover a complex structure of the subducting slab (Fig.1).Our results also revealed variations in Vs velocities (Fig.2) and the anisotropy obtained by inversion for a layering structure separating the crust and the lithosphere.The study of anisotropy detects different fast directions at different periods in the north part of the study area. This complex pattern could be the signature of a locally modified flow field around the edges of the slab. To the west of the MASE transect, the fast directions are perpendicular to the trench. This may be explained by the alignment of dry olivine due to plate motion. The fast propagation directions east of MASE are parallel to the trench, which may


!

Fig.1. Phase velocity perturbation and anisotropy calculated from surface waves

correspond to trench-parallel flow induced by toroidal flow around the slab. Such directions could also be the signature of wet olivine near the slab. The fast direction interpretation strongly depends on the water and partial melt content which is currently uncertain.

Future DirectionsOur future research will be directed towards interpreting the results we have obtained so far and combining them with the SKS splitting results to better explain the mantle dynamics near the slab. We also would like to include the tectonic history and the current tectonic velocities which may especially help explaining the anisotropic results.


Fig.2 Phase velocity inversion results and the layering. It easy to identify the slow crust (top 40 km depth) and the separate slab signatures in the lithosphere at 30-90 km depth (blue color). The faster velocity between the slabs may be interpreted as a mantle flow induced by the slabʼs roll-back.

!

!

SEI 02 Influence of the Pacific trench of Mexico on the wave propagation. A new phase detected using the CENS seismic network.

Team Members• Paul Davis, Faculty*• Luis Dominguez Ramirez, Graduate Student*• Francisco Sancez-Sesma, Faculty

* Primary Contact

OverviewThe Pacific trench of Mexico marks the lateral discontinuity between the continental and oceanic crust. On one side of the trench, the Cocos plate subducts underneath the North American plate at a rate of ~5cm/year while the other side is composed of granitic crust with lower density and higher velocity of propagation. Unlike a typical subduction zone, the angle of subduction is almost zero. In other words, the Cocos plate subducts parallel to the surface up to a distance of approximately ~250km from the trench, and then suddenly dips into the mantle at an angle of ~60°. These special features of this tectonic environment forms a waveguide-like structure. Propagation of seismic waves from teleseismic events drastically change for events coming at either side of the trench. Figure 1 shows a dramatic example of this effect. In this project, we explored the mechanism that causes these differences.

Figure 1 Differences in wave propagation for two events coming from opposite sides of the trench. (Left) Interpolated records from the seismic array showing the SS phase, the red circle shows a trapped phase originated at the trench. (Right) Seismic record from different event arriving from the continental side of the trench, horizontal axes is inverted for comparison. Y-axis is the time from the event.

ApproachThe scattered signal is readily seen under specific circumstances: an appropriate distance, incident angle of the scattered field, and magnitude of the event. Nevertheless, the origin and nature of the scattered field required a meticolous analysis. We carried several signal and sesmic processing techniques to the data to explore the propagation features of the field. First, we applied a low pass butterword filter with corner frequency at 1Hz to the data to remove high frequency noise. Identification, timing of the scattered signal, and its relationship with the incident field was achieved by means of the tau-p transform. This transformation consists in slant-stacking the records along different combinations of slowness and time arrival intercepts, this permits to quantify the relationship between these two signals. In addition, analysis of the particle motion of the scattered field revealed that the scattered wave travels as a surface wave with a domind period of ~20s. This enhances our initial assumption that the scattered field travels as a trapped wave into the crust. Finally, we compared our results with a syntetic seismograms obtain using a boundary integral method.


System Description and/or ExperimentsThe Middle America Seismic experiment consisted of an array of one hundred seismic station installed in central Mexico from 2005-2007. This project had the collaboration of the three institutions: CENS/UCLA, the Institute of Technology of California (Caltech) and the National Autonomous University of Mexico (UNAM). Unlike other seismic experiments, this array was deployed with a wireless communication system developed at CENS. Fifty out of one hundred stations were equipped with a wireless radio that broadcasts data from one station to an adjacent station that allowed near real-time acquisition of the data. Another unique feature that highly benefited the interpretation of our results was its geometry and spatial density of the stations. The array extended linearly for almost ~550 km from the Pacific Coast of Mexico perpendicular to the trench with spacing between stations of ~5-6km. This network resulted in one of the most advanced portable arrays of the world that provided a unique set of data for exploring the seismicity and structure of this area.

AccomplishmentsThis project provided a successful identification of the factors that give rise to the partition of energy due to the lateral transition between the oceanic and continetal crust, and the flat subducting geometry of the slab. Table 1 summarizes the major results obtained in this project. We detected a striong scattered field for seven events magnitude Mw=7.0 occured in the Southern Hemisphere at teleseimic distances between ~7500-12500 km. The phases that provided the best conversion rates are SP, PS, and SS whose incident angles ranges between 30-47 degrees. The apparent velocity along the array ranges between 3.27-4.17 km/s that lies in the range of the velocities for surface waves. The propagation of the scattered field also shown interesting features. For example, it does not show significat dispersion of the scattered field, and the propagation occurs with very little attenuation. This suggest that the trench acts as a quasi-linear source when is excited by a distant field. Similar effects are poorly document in the literature despite of the fact that the ratio between the scattered and incident fields can reach up to 68% percent of the incident field.

Event – Inc. PhaseIncident

angleat 50 km

App. Vel.Inc. Field

[km/s]

Slowness Scat. Field

[s/km]

App. Vel.Scat. Field

[km/s]Ratio [%]

Tonga – SP 47 8.83 0.27 3.57 44Tonga – SS 31 8.83 0.26 3.81 21Fiji – SP 44 15.87 0.27 3.72 23Fiji – SS 30 19.98 0.26 3.86 60Kermadec – PS 23 11.00 0.29 3.49 28Kermadec – SS 30 7.63 0.27 3.66 68Scotia Sea – PS 23 14.20 0.28 3.62 11Scotia Sea – SS 30 7.77 0.29 3.43 66Vanuatu – PS 22 19.04 0.27 3.65 22Vanuatu – SS 30 9.80 0.26 3.81 47Sandwich – PS 22 11.00 0.30 3.28 38Sandwich – SS 29 11.00 0.28 3.51 37Solomon – PS 21 11.00 0.30 3.28 53Solomon – SS 28 11.00 0.24 4.17 44

Table 1 Summary of results

Future DirectionsThis year, we will analyze the internal scattering in the crust caused by hetereogeneities in the medium. This time, we wil process information from local earthquakes to determine statistical properties of the crust. We will also perform measurements of attenuation, and addrress several questions about what are the main contributors to attenuation in this area. For example, there is an intense debate whether the intrinsic attenuation (engery converted into heat), and scattering attenuation (redistributionof energy) can be separated, or not. Another question, we will explore is what is the role of the geometrical spreding and whether it can be misinterpred as attenuation. Our goal for this year, it is to provide new evidence for the better understanding of attenuation for future studies.


SEI 03 GeoNet: A Platform for Rapid Distributed Geophysical Sensing

Team Members• Paul Davis, Faculty, PI*• Deborah Estrin, Faculty, PI• William Kaiser, Faculty• Dustin McIntire, Graduate Student• Igor Stubailo, Graduate Student• John Wallace, Faculty, PI

* Primary Contact

OverviewThe science objectives are to use a rapidly installable wirelessly linked seismic network to measure earthquake or volcano sources in the near field to understand the underlying physics, or in buildings to understand earthquake damage. To accomplish these objectives, we collaborated with Reftek to construct a new generation digital acquisition system (DAS) based on the CENS-developed LEAP (low-power energy aware processing) system and a newly developed low-power A/D converter from Texas Instruments (TI). In preparation for GeoNet, we have used the Peru networks to test the software including improving Disruption Tolerant Shell (DTS), measurement of radio link quality (ETT), network logging, an embedded web interface based on Emstar for deployment and maintenance, network timing, a new routing protocol that caches the routes across sleep cycles for a fast startup. The Peru network has already been installed and we can add to it GeoNet nodes for further testing and debugging.The objective is to understand the basic physics of earthquakes and their effects. Presently we can not predict earthquakes. With the information we seek we hope to predict aftershocks and use that information to eventually predict main shocks. Because main shocks at a given location happen about every 200 years, it has been impossible to build up statistics on possible precursors. Modern understanding of the earthquake source is that an aftershock sequence is a time-compressed version of long term seismicity. If there are repeatable predictive behaviors before an event, an aftershock zone is where we are most likely to measure them. By knowing the time and location of a large aftershock, as well as giving warning, we can instrument buildings and infrastructure to measure structural failure during strong shaking.We propose to develop and install an intelligent wireless seismic network, a ʻSeismoScope (inverted telescope)ʼ to make FlexiRAMP recordings of earthquakes and their aftershocks. Our SeismoScope must be flexible to be installed quickly in an aftershock zone after a large event (for example in the west Americas) or, in the interim, in regions with high probabilities of main shocks. It must be made of lightweight components, low power, and radio linked so that the data can be analyzed immediately and (real time) warnings issued. This type of rapid response is termed FlexiRAMP, Flexible Rapid Array Mobilization Procedure. At present permanent network stations are too coarsely spaced to understand the earthquake process. We are missing coherent energy that can be measured if the stations are much closer, for example 1 km apartThe recently developed ETAS (Epidemic Type Aftershock Sequence) model describes earthquakes as a stochastic branching sequence of triggered events. It was used successfully this year, after the April 6, 2009, Friuli earthquake in Italy, to predict where damaging aftershocks would occur, but stations were too few and too widely spaced, to make detailed measurements. Given this new paradigm, after a large earthquake, for example in the western US, Mexico, or Peru (where we have installed networks) we will install the SeismoScope to record the thousands of aftershocks that occur. Very often the largest aftershock in a sequence is 1 magnitude smaller than the main shock, e.g., a magnitude 8 is followed by a magnitude 7. We will use our network to detect how the shocks leading up to the M=7, are distributed in space and time, with the objective of developing ETAS prediction algorithms based on seismicity. Through CENS, in a multidisciplinary project involving four UCLA Departments, the PIs have designed a novel FlexiRAMP node and have had two built by the leading manufacturer of seismic recorders (Refraction Technology of


Fig.1. Schematics of possible field instalations.

Texas) (Fig 2). As well as installing the network across country, once time and locations of aftershock shaking has been predicted, a portion of the equipment will also be installed in buildings and infrastructure. This will provide engineering data on design factors, necessary for buildings to remain operational after earthquakes, that are critical

for saving lives and minimizing economic impact.

ApproachWe used the Peru networks as testbeds for GeoNet systems software research. The field objective is to have only two separate parts: DAS and a seismic sensor. The DAS would have a solar panel attached on top, battery inside (with external power plug), internal GPS antenna (with a possibility attaching an external one), external N-type connector to attach an antenna. Field installation would involve attaching the box to a post and bury the seismometer (Fig 1). It would then be a node in a wireless network of neighbors, e.g. along a dirt road that could bring event data

out in real time, or, in the low power mode, on a duty cycle, e.g., 5 min every hour. The radios would also deliver network time, a backup to GPS.A prototype of a GeoNet box with modular approach inside. The software design for the seismological community would include CENS networking auto routing component and the low power configuration ability of the hardware. To conserve power the processor will duty cycle between sleep and wake states. The node is based on the Linux operating system, with 802.11 capabilities for wireless networking. Software development is required for array-event detect, network and GPS timing, multi-hopping the data, and near real-time data analysis and modeling. The software system will be based on the infrastructure used in the Peru network. However, the existing software system is not directly applicable because of a number of factors introduced by the application requirements. These include duty


Figure 2. Seismic node for FlexiRAMP Wireless seismic network delivered by Reftek. We propose a 100 station radio-linked network to monitor earthquakes. The digital system and software is of UCLA (Earth and Space Sciences, Civil Engineering, Electrical Engineering, and Computer Sciences) design Reftek president Paul Passmore and engineer Phil Davidson).

Figure 3. Test at Palmdale of new GeoNet boxes and Quanterra devices used in Peru. The upper figure shows the field layout. The lower figure shows the waveforms match very well.

cycling, non linear network topologies, array event detect, and the time span of the deployments. We are developing software to deal with these new factors, in particular: sleep schedule aware data delivery, disk space based routing, deployment tools, and integration with Wavescope to enable streaming, network processing, and cooperative event detection.

System(s) Description and/or ExperimentsThe best experiment for GeoNet is an afterschock deployment, but it could also cover any seismic experiment where rapid deployment is necessary (volcanoes, explosions of opportunity). A pool of identical instruments would be available to either install in buildings or in the field and could be standalone or where appropriate nodes in a wireless network. The objective is to measure strong shaking from the largest aftershock for both science and engineering objectives. The GeoNet/SHM requiremenets have much in common, but there are differences. Building installations do not have access to GPS time, may or may not have power. Field installations, being remote and widepread, need power. Building sensors tend to be active (Episensor). Field installations can have passive sensors, requiring significantly less power.We are going to test 2 GeoNet stations with passive sensors (L4-C) in the Salton Sea area (California) for a week in Maarch during a gephysics field trip. This will allow us to further improve the case design and user interface. We also will test some units in our Peru network. They will be mixed with high quality broadband Guralp 3T sensors and nodes running DTS software. We will be comparing data quality and the behaviour in a larger network while having the real time access to the sites.

AccomplishmentsA number of posters were presented about different features of system (software and hardware) at the AGU-2010 meeting as well as papers published. A contract for Reftek construction was signed and two prototypes were delivered. A successful field test of the GeoNet boxes was conducted at Palmdale for comparison with the Geometics digitizers (Figure 3). The LEAP board has been interfaced in the GeoNet boxes. The GeoNet boxes have been interfaced to cell phone modems. Event detect software has been written. A test is currently underway at the time of a series of explosions in the Salton Trough, in southern California. Dustin McIntire and Igor Stubailo will install the two prototypes linked by Wifi. The data from each will be aggregated and the event detect software will detect events and will send the event data and the catalogue to the CENS computing cloud as well as sending emails with plots of the events and 60 sec data to mobile phones that are on a list.

Future DirectionsWe will continue working on integrating duty cycling with existing software based used in Mexico and Peru. This includes further developing a robust routing algorithm that can deal with nodes with limited storage space with a complete a centralized method to dynamically schedule duty cycling and transmissions. We will develop software to deal with sleep schedule aware data delivery, disk space based routing, as well as tools for network processing and cooperative event detection. The GeoNet instruments will have application over a wide range of field applications involving wide area networks and low power processing and delivery of event data. They will run on small batteries with a laptop sized solar panel.


2.6 Seismic Applications (SEI) - pdfs.semanticscholar.org€¦ · 2.6 Seismic Applications (SEI) The Seismic research area continued analysis of data captured by the Middle America

Documents