G23A-01 INVITED
Integrating GPS and Seismic Data in Earthquake Source Inversions
Earthquake source inversions make use of recorded ground motion to image the seismic rupture. In general, two different types of data are used for this purpose: 1) records of the static field and 2) records of seismic waves. The static field data record the difference in ground positions before and after the earthquake; it can be used to infer the cumulative slip that occurs during an earthquake. However, it tells nothing about the temporal evolution of slip on the fault. In order to image the full space-time evolution of slip on the fault one must utilize data that record the radiated wavefield. The wavefield is normally captured by seismographs. High-rate GPS sensors also can record the radiated wavefield. However, these high-rate (1-Hz) GPS measurements are useful as seismograms in seismic source inversions only in a very limited frequency band. On the other hand, high-rate GPS can be very useful in seismic source inversions as it allows the determination of the truly co-seismic static field. In other words, the static field inferred from high-rate GPS does not include significant post-seismic deformation. Thus, the truly co-seismic static field provided by high-rate GPS data can be inverted to determine the cumulative slip that took place during the earthquake. In turn, the resulting cumulative slip distribution can then be used to constrain an inversion of the radiated wavefield for the temporal slip. Given the strong trade-offs between temporal and spatial parameters in inversions of the wavefield, it is extremely valuable to use a priori knowledge (in this case the cumulative slip). Accordingly, we developed a two-step procedure to combine GPS and seismic data in earthquake source inversions. In the first step we use the truly co-seismic static field (inferred from high-rate GPS) to infer the spatial distribution of cumulative slip on the fault. This inversion is performed on an irregular grid that takes into account the spatially heterogeneous resolving power of the GPS data. In the second step, we use records of ground acceleration to infer the space-time history of slip on the fault, while constraining the final slip distribution to match that inferred from the static field. This approach explores the resolving ability of each dataset (static and dynamic), thus using each dataset to determine appropriate parameters and leaving out unwanted artifacts.
G23A-02
Regional rumble: a seismological study of glacial earthquakes in Greenland
Glacial earthquakes occur primarily in Greenland, in connection with major outlet glaciers. They were first discovered by Ekstrom et al in 2003 using teleseismic data and a global detection algorithm based on surface waves. Seismograms recorded in Greenland reveal a much richer frequency content in the glacial earthquakes than observed in the teleseimic data. Thus the regional data have the potential to provide more accurate locations of the events as well as a better understanding of this peculiar phenomenon. Since 2000 a total of 30 broadband (BB) seismographs have been deployed in Greenland for periods ranging from 3 months to currently 7 years. While the seismographs were installed for structural studies, the data are equally useful for analyzing glacial earthquakes. One of the most productive glaciers in Greenland with respect to glacial earthquakes is Helheim Glacier in East Greenland. A BB seismograph has operated continuously in the nearby town of Tasiilaq since January 2000. In July and August 2007 two additional BB seismographs were installed in the vicinity of Helheim Glacier, where they will operate for at least a year. The first data have already been retrieved and will be presented. This study of glacial earthquakes is part of a large multidisciplinary effort involving seismology, geodesy, glaciology and climatology. Major field work has been carried out on Helheim Glacier during the summers of 2006 and 2007. In this presentation we will focus on the seismological results based on recordings from Greenland.
G23A-03
Analysis of time-variable crustal velocity using Continuous GPS Time-Series
The amount of continuous GPS data available both through EarthScope and from around the world is growing dramatically. Routine processing of these data is generating time-series of daily site positions and secular site velocities with incredible success. However, analysis techniques that explore the time dependence of crustal velocity are lacking. We are developing a method to characterize the four-dimensional crustal velocity field using continuous GPS in a systematic manner appropriate for routine processing. Time dependent crustal velocities can be used to study time dependent strain fields associated with viscous relaxation processes, aseismic slip events, and volcano deformation, and many other applications. Our method models crustal velocity using smooth continuous splines. The splines are fit to continuous position time series. This spline approach has several advantages 1) it does not assume a functional form for deformation a priori, 2) observations need not be regularly spaced in time, 3) it provides an easy decomposition of white noise and time-variable signal, and 4) it provides a continuous representation of time-variable site motion. Our time series model presently consists of an initial position, a secular velocity term, a time dependent velocity function, and white noise. The variability of the time dependent velocity function and the variance of the residual white noise are controlled by an empirically determined damping factor. The algorithm is highly configurable and appropriate for studies of both long-term and short-term signals. We demonstrate the technique using example time series from long-running continuous GPS stations in the western US and elsewhere.
G23A-04
Current Status and Future Directives of the Nicoya Peninsula Continuous GPS Network, Costa Rica, In Regard to Slip Style and Distribution
The close proximity of the Nicoya Peninsula land mass to the Cocos-Caribbean subduction zone plate boundary makes it a prime location to use GPS to study such plate boundary processes as locking zone dynamics and episodic tremor and slip events. Nicoya Peninsula currently has 12 continuous GPS (CGPS) stations installed each with varying degrees of equipment and communication. Seven of these sites are currently installed with Trimble NetRS receivers and five with Trimble 5700 receivers. Of these 12 sites, four are equipped with SIM cards and modems for direct download capabilities, one of which is connected to a router for direct internet access. This site, LIMO, located in Limonal, is being monitored on the PBO network by UNAVCO with all of the quality and systems checks that this implies. The other eight sites currently require manual download. All sites with NetRS receivers are currently partitioned to record both 5Hz and 30 second position data. We will present the first years time series from these stations. The advantage to recording at a 5 Hz interval is that the receivers may be used to record long period dynamic events during large earthquakes along the Cocos-Caribbean plate boundary. One goal of the network is to investigate episodic tremor and slip events, as have been observed in the Cascadia subduction boundary. Such events have been speculated upon in this region, however due to the sparse network at that time, we were unable to resolve the event. With the expanded network of CGPS stations, we hope to better record and understand such events. We will also present an anlaysis of synthetic data to assess the spacial resolution of detectable slip patterns with the current network spacing. There is a future plan of three more installations on the interior of the peninsula to further optimize the network for seismic and locking pattern studies.
G23A-05
Implementation of the Center-East Broadband Seismic Network in Colombia
The seismic phenomenon is one of the most important natural hazards for Colombia (e.g. Armenia, 1999, M=6.1; Calima, 2004, M=6.7). In spite of big efforts implemented by different institutions in order to reduce the seismic vulnerability in Colombia, the understanding of the nature of the seismic phenomena and its spatial and temporal distribution and evolution is insufficient. In order to estimate the seismic hazard it is necessary to know the geological conditions of those regions where the cities are developing, it is necessary to study their structure, dynamics and behavior, looking for information which allows understand the processes which generate the seismic events and to obtain an adequate model of ruptures of the system. In this paper we show the implementation of an instrumental seismological network for the Colombia's capital city: Bogota. This project: "The Center-East Broadband Seismic Network" consists on fifteen seismological three components, broadband stations, located near the coordinates: 4.5 N latitude and 73.5 W longitude joined to a similar distribution of geodetic stations (GEORED). This network is expected to be a complementary, high sensitive addition to the National Seismological Network of Colombia. This network has to bring the possibility of improving the understanding of the seismic hazard in the Bogota's region, considering the better understanding of the spatial and temporal microseismic activity in this region, where live more than 15 millions people (30% of Colombia's population), and more than 40% of the GIP is concentrated. Based on the obtained information it will be possible to incorporate structural elements to the adequate development of this region considering the seismic hazard and supply the guidelines to designs and constructions.
G23A-06
The combined EarthScope data set at the IRIS DMC
The IRIS Data Management Center (DMC) is the perpetual archive and access point for an ever-increasing variety of geophysical data in terms of volume, geographic distribution and scientific value. A particular highlight is the combined data set produced by the EarthScope project. The DMC archives data from each of the primary components: USArray, the Plate Boundary Observatory (PBO) & the San Andreas Fault Observatory at Depth (SAFOD). Growing at over 4.6 gigabytes per day, the USArray data set currently totals approximately 5 terabytes. Composed of four separate sub-components: the Permanent, Transportable, Flexible and Magnetotelluric Arrays, the USArray data set provides a multi-scale view of the western United States at present and the conterminous United States when it is completed. The primary data from USArray are in the form of broadband and short-period seismic recordings and magnetotelluric measurements. Complementing the data from USArray are the short- period, borehole seismic data and borehole and laser strain data from PBO. The DMC also archives the high- resolution seismic data from instruments in the SAFOD main and pilot drill holes. The SAFOD seismic data is available in two forms: lower-rate monitoring channels sampled at 250 hertz and full resolution channels varying between 1 and 4 kilohertz. Beyond data collection and archive management the DMC performs value-added functions. All data arriving at the DMC as real-time data streams are processed by QUACK, an automated Quality Control (QC) system. All the measurements made by this system are stored in a database and made available to data contributors and users via a web interface including customized report generation. In addition to the automated QC measurements, quality control is performed on USArray data at the DMC by a team of analysts. The primary functions of the analysts are to routinely report data quality assessment to the respective network operators and log serious, unfixable data issues for reference by data users. All of these data are managed in a unified SEED format archive and are seamlessly available to data users via the DMC's&pstandard data access methods along with all the other data managed by the DMC. The only exception is high resolution, special case SAFOD seismic data that is retained in its original SEG-2 format as an assembled data set. A data user can choose between a handful of data access methods ranging from simple email requests to technologically advanced CORBA-based access, streamlining the "information into application" philosophy. Currently totally over 8.5 terabytes and growing, the combined EarthScope data at the DMC provides an unparalleled, multi-measurement record of geophysical information ideal for determining Earth structure and processes in the United States and beyond. A website is maintained to provide current information regarding EarthScope data at the DMC: http://www.iris.edu/earthscope/.
G23A-07
The Plate Boundary Observatory: Data Management Progress and Highlights
The Plate Boundary Observatory (PBO), part of the NSF-funded EarthScope project, is designed to study the three- dimensional strain field resulting from deformation across the active boundary zone between the Pacific and North American plates in the western United States. To meet these goals, UNAVCO will install 880 continuous GPS stations, 103 borehole strainmeter stations, 28 tiltmeters, and five laser strainmeters by October 2008, as well as manage data for 209 previously existing continuous GPS stations and one laser strainmeter through the PBO Nucleus project and 11 GPS stations installed by the USArray segment of EarthScope. As of 1 September 2007, UNAVCO had completed 680 PBO GPS stations and had upgraded 89% of the planned PBO Nucleus stations. Most of these stations return data to the UNAVCO Boulder Network Operations Center (NOC) on a daily basis, with about 40 stations returning data on an hourly basis. Overall, the combined PBO and Nucleus network has now provided almost 350 GB of raw standard rate data, which are routinely processed by the PBO GPS Analysis Centers, at Central Washington University and the New Mexico Institute of Mining and Technology, and the PBO GPS Analysis Center Coordinator at MIT. These groups create a range of GPS products, including station position time series, GPS velocity vectors, and related information. As of September 2007, these centers processed data on a daily basis from about 920 stations; typical position uncertainties are under 1.5 mm horizontally and 4 mm vertically. All PBO GPS data products are archived at and available from the UNAVCO Facility, with a second archive at the IRIS Data Management Center (DMC). All these products may be accessed via the PBO web page at http://pboweb.unavco.org/gps_data. As part of PBO, UNAVCO will also install and operate the largest borehole seismic and strainmeter networks in North America, as well as tiltmeters and laser strainmeters. As of September 2007, 41 PBO borehole stations had been installed and three laser strainmeter stations were operating. Seismic data flow in real time to the Boulder NOC for initial quality checks, and then to the IRIS DMC for final quality checks, archiving, and distribution; all PBO seismic data flow is via the Antelope software suite. Strainmeter data flow hourly and daily to the Boulder NOC and then to the Borehole Strainmeter Analysis Center in Socorro, New Mexico, and the Laser Strainmeter Analysis Center at the University of California, San Diego. These groups transform the raw strainmeter observations into cleaned individual strain gauge components; time series of shear, areal, and linear strain; and related products. All strainmeter data products are archived at and available from the Northern California Earthquake Data Center and the IRIS DMC, in both the native raw formats and SEED format; all seismic data products are archived at and available from the IRIS DMC, in SEED format. By September 2007, the PBO seismic network had provided more than 200 GB of raw data, and the PBO strainmeter network had provided almost 100 GB of raw data. Please visit http://pboweb.unavco.org/strain_data for more information on data products from the PBO strainmeter and seismic networks. http://pboweb.unavco.org
G23A-08
PBO Borehole Strainmeters: Bridging the Gap Between Seismology and GPS
A key objective of the Plate Boundary Observatory, the geodetic component of Earthscope, is to capture the spatial and temporal deformation field across the western-US plate boundary. The observatory, being installed by UNAVCO, will consist of 103 borehole strainmeter/seismic sites, 880 GPS stations, 28 tiltmeters, and 5 laser strainmeters. Borehole strainmeters are designed to record deformation that lies between the spectral coverage of seismometers and GPS and are ideal for capturing strain transients that occur in periods of hours to months. Hence, they are being installed in arrays in targeted regions where they may capture short-term strain transients. As of 1st September 2007, 16 strainmeters have been installed in the Pacific Northwest, where strain transients associated with the 2005 and 2007 Cascadia ETS events were successfully recorded, 7 in Parkfield, in the transition zone between the creeping and locked section of the San Andreas Fault, 7 in Anza, in the aseismic section of the San Jacinto Fault, 4 in San Juan Bautista and 4 in central and southern Oregon. Three strainmeters of the first volcanic array were installed on Mt. St. Helens in July 2007. The combination of strainmeter, seismic, GPS and tilt measurements in these regions provides an unprecedented three-dimensional continuous dataset that spans the spectrum of plate boundary deformation. PBO borehole strainmeters record at 20-sps, 1-sps and 10-minute interval and the data are available in SEED format from the NCEDC and the IRIS DMC within minutes of download. Processed strainmeter data, produced by UNAVCO's Borehole Strainmeter Analysis Center are updated every 10 days and are available from the NCEDC and IRIS DMC in XML format. The processed data sets include tidal, atmospheric and borehole corrections plus data quality tags that flag non-tectonic signals. Users may also download raw strainmeter data via interactive plotting tools on the PBO web site or use SQUID, a GUI tool developed by PBO, to generate processed strainmeter data. A complete list of all strainmeter data products including metadata, borehole drawings and maintenance information is available at http://pboweb.unavco.org/strain_data.