G21A-0101 0800h
A Comparison of Geodetic Strain Rates With Earthquake Moment Tensors
In this paper we compare the global model from interpolation of GPS data with the global model inferred from earthquake moment tensors. We use the Harvard CMT catalog to calculate moment rates based on 3 assumptions: a. we assume earthquakes are self-similar; b. we assume a uniform Beta value of the Gutenberg-Richter distribution; c. we assume that all of the long-term strain is accommodated seismically. If these assumptions are correct then the seismicity rate is proportional to the tectonic moment rate. We then inferred a long-term moment rate tensor field estimate for all plate boundary zones from which we inferred a long-term seismic strain rate estimate. Using this estimate we solved for a self-consistent kinematic global solution (motions of rigid spherical caps and motions within plate boundary zones) using bi-cubic spline interpolation of the inferred strain rates. We tested the above assumptions by comparing the global kinematic model obtained from earthquake data with a global model inferred from interpolation of space geodetic data [Kreemer et al., 2003]. A comparison between the two models shows good agreement for motion directions of the North American, and Eurasian plates and for the plate boundary zones within these regions (e.g., Tibet). Problems arise, and our assumptions break down, for plates adjacent to fast spreading ridges where divergence of plates appears to be accommodated aseismically. We next investigated the correlation of strain rate tensor inferred from the interpolation of GPS observations within deforming Asia with the earthquake moment tensors, using both elastic and viscous rheologies. Our solutions satisfy the force balance equations for a given rheology. Our goal for this exercise is to investigate whether the interseismic signal, inferred from GPS, correlates better with moment tensor style for an elastic rheology as opposed to a viscous rheology. Results to date suggest that the viscous models only provide a better agreement with observed styles of faulting if the relationship between stress and strain rate is anisotropic.
G21A-0102 0800h
SPATIAL AND TEMPORAL VARIATIONS OF INTERSEISMIC STRAIN FROM GPS MONITORING (CGPS AND CAMPAIGN DATA) AND SEISMIC MONITORING IN THE NEPAL HIMALAYA
We investigate interseismic processes in the Himalaya from geodetic and seismic monitoring. Geological investigations have shown that the shortening rate across the Nepal Himalaya is taken up by a major thrust fault, the Main Himalayan Thrust fault, which emerges along the Himalayan piedmont. GPS campaign data are reasonably well adjusted from a model in which the fault is assumed to be fully locked over a width of the order of 100km with only minor lateral variations. However there is no necessity that interseismic strain be stationary, due to possible variation of the width the Locked Fault Zone (LFZ) or variations of ductile shear rates downdip of the LFZ in the postseismic period or throughout the interseismic period. In addiation we cannot exclude the possibility of slow events. Indeed, if the locked fault Zone were to slip only during seismic events, it would require a seismic release rate, on the long term average, well in excess of what is estimated from historical catalogues. This implies that the LFZ might slip during transient seismic events on some occasion. So geodetic strain rates might vary with time, and differ from that predicted form the simple kinematic model above. Since background seismicity in the Himalaya is driven by stress build up along the downdip end of the LFZ, seismicity rate might also show meaningful temporal variations. To investigate this problem, we have deployed a network of permanent GPS stations that complements the existing seismic network and GPS campaign data. Four stations (including one from IDYLHIM Team) have been in operation since 1997 and have been used to evaluate the best processing strategy. These stations are situated along a transect across the Himalaya of Central Nepal at the latitude of Kathmandu. All data were processed with Bernese in the International Terrestrial Reference Frame 2000 (ITRF2000). We have tested various processing strategy and found that strain within our network was best resolved if only one IGS station is jointly processed, probably because common mode in seasonal variations then cancel out. We used LHASA as the reference IGS station. The position of our stations in ITR2000 was next estimated from the position of LHASA relative to ITRF2000, with account for observed seasonal effects. We find that to the data are consistent with a present rate of shortening across the range of 17.5mm.yr$^{-1}$ slightly less than the Holocene slip rate of 21$\pm$1.5mm.yr$^{-1}$ along the MHT. In addition, significant temporal variations are observed at GUMBA, and DAMAN. These stations lies in the area that is most sensitive to strain accumulation near the downdip end of the locked fault zone, where interseismic strain induced a well clustered seismicity. This suggests that strain accumulation in the interseismic period might not be a stationary process.
G21A-0103 0800h
Systematic Analysis of Historic Seismicity and Geodetic Measurements
Large, extended fault systems are known to demonstrate complex space-time seismicity patterns which include, but are not limited to, repetitive events, precursory activity and quiescence, and aftershock sequences. Here we compare the results of an in-depth study of the southern California fault system using a pattern informatics (PI) analysis of seismicity, GPS measurements, and sequential InSAR images of the same region. The PI methodology (Tiampo et al., 2002) has been employed in the past to isolate emergent regions of coherent, correlated seismicity prior to their occurrence. Here it is associated with regions of anomalous stress and strain that can be quantified with local and regional geodetic measurements.
G21A-0104 0800h
Dynamics of Deformation in Asia from a Combined GPS Solution and Finite Element Models
The ability of plate tectonics concepts to describe deformation of large continental areas like Asia or the Western United States is subject to much debate. End-member models assume that continents deform either by rigid motion of lithospheric blocks along narrow fault zones or by viscous flow as continuously deforming solids. Discriminating these end-member models requires spatially dense measurements of surface strain rates covering the whole deforming area, including areas outside the obvious active faults and plate boundary zones. We present new results on present-day crustal deformation of Asia based on a combination of GPS solutions covering the entire deforming domain, including southeast Asia, China, Mongolia, and the Baikal rift zone. We discuss whether present-day surface displacements in Asia are best described by the motion of a limited number of rigid blocks or by diffuse deformation. We use the GPS results to validate dynamic deformation models aimed at better understanding the force balance that drives continental deformation in Asia. Major features of the best-fit models include (1) the importance of buoyancy forces in driving continental deformation, (2) localized deformation on major active faults, and (3) a good match of predicted fault slip rates with observed Holocene rates, except for the Altyn Tagh and Karakorum faults.
G21A-0105 0800h
Crustal Deformation in Jiashi Region,China Inferred from GPS
The Jiashi region lies in the north of collision zone between the Indian and Eurasian plates and it is a complex region with distributed seismicity and many active faults. A dozens of moderate and strong earthquakes have occurred there since 1900. Seven events greater than M6 took place consecutively during 3 months in 1997. In order to monitor crustal deformation, we started GPS observations in this region since 1994. A total of 9 stations, which are part of GPS moitoring network in southern Tienshan, were constructed within the area about 800 km long and 640 km wide. All of these stations were observed in campaign mode at least 2 times in 1994-2000. In each survey, 4-5 dual-frequency geodetic GPS receivers were utilized to collect the data at different stations simultaneously. All the stations were observed for 3 to 7 consecutive days with continuous recording for almost 24 hours a day. Data from three continuous IGS stations (SELE, POL2, KIT3) are also included in this study. All GPS data acquired either from campaign mode or continuous observations were processed with GAMIT/GLOBK software packages. Firstly we used the GAMIT software to get daily solutions with loose constraints to regional stations and tight constraint to the three IGS stations. Then the daily solutions are utilized to estimate the velocities of GPS stations. In this study we adopt ITRF97 reference frame. We also estimate strain field from the velocity of all GPS stations in this region. GPS monitoring in the Jiashi region provides us with a description of present-day curstual deformation. The main conclusions of this study are summarized as the following: All stations, as a whole in the Jiashi region move northward in the nearly north-south direction relative to northern Tianshan. Tarim basin push Tianshan in the direction perpendicular to Tianshan range. We estimate that the northward motion of GPS stations between $39.5\deg$N and $41.5\deg$N decreases from west to east, with the rate approximating 1.0mm/yr per degree. The GPS stations in the northwest of Kashi and Wushi move in different direction from those stations in the southeast, implying that a counter-clockwise rotation exists along the stations Kashi and Wushi, consitent with geological investigation. Given the current spacing of GPS stations, we cannot find that there are significant deformation apparently concentrating on any local area or active faulting. The strain estimation demonstrates that the principal axis is nearly in north-south direction. Compressure force dominates this region,which are consistent with the results inferred from the analysis of initial p phase of the events greater than M5 in the soutern Tienshan region during 1968-1982, which showed that compressure axis in Pamir and Karokunlun-west Tianshen was mostly in N-S direction, with plunge angle being horizontal. Our result demonstrates that there exists strong compressure in the directions of N-S and NNE-SSW. This might imply that northward movement of India plate is the main source of pushing force and the crustal deformation regime in the Jiashi region is related to the ongoing northward movement of India plate.
G21A-0106 0800h
Strike-slip fault evolution at intermediate (10 ka - 1 Ma) timescales: an example from the Aksai restraining stepover along the Altyn Tagh fault, NW China
Many studies of active, strike-slip fault systems have focused on both short-term (0-10 ka) and long-term ($>$5 Ma) strain, but less attention has focused on how such systems evolve at intermediate (10 ka to 1 Ma) timescales. Sophisticated models of evolving fault geometry at such intermediate timescales are particularly needed. To address this problem, we are investigating how fault-perpendicular shortening is accommodated across the 100 km long (E-W) and 20 km wide (N-S) Aksai stepover (93E-94E), one of four restraining bends along the left-slip, Altyn Tagh fault in NW China. The Aksai bend is defined by northern and southern ranges which flank the restraining bend in its NE and SW quadrants and are $\sim$70 km long (E-W) and $\sim$25 km wide (N-S). Between these ranges, the central Mt. Altyn uplift is cut by a complex network of fault traces which are part of the Altyn Tagh fault system. While these flanking ranges mirror each other in shape and morphology, sharing common features like sharp SE-NW trending range fronts and interior valleys that separate them from the central range, our 1:20,000-scale neotectonic and 1:100,000-scale structural mapping reveals that the northern range has a greater number of active thrust faults and folds that record larger magnitudes of active shortening than the southern range. The northern range is bordered to the north and south by the south-dipping Da Long thrust system and the north-dipping Sai Ma right-reverse fault, respectively. Evidence for active, stepover-perpendicular shortening along these structures includes fresh fault scarps with vertical separations of 2-35 m, offset drainages, $>$2000m of local topographic relief, and sets of late-Quaternary (?) to Holocene (?) fluvial terraces that are both folded and faulted. In contrast to the distributed deformation along the north side of the stepover, only a single active structure (the Tuo La Yi fault) accommodates stepover-perpendicular shortening within the southern flanking range. This fault defines the NE margin of the range. The southern range has fewer fault scarps than the north range, more muted topography, and no deformed terraces. These results suggest that active shortening is concentrated in the north range. We speculate that the short-term (10 ka to 1 Ma) strain pattern revealed by our mapping may help to explain the long-term ($>$5 Ma) evolution of the Aksai stepover. In particular, the Mt. Altyn uplift and the complex network of fault traces at the center of the bend may have resulted from earlier bend-perpendicular shortening, which has more recently been transferred to the younger structures within the flanking ranges.
G21A-0107 0800h
Revalidation Of Indian Plate Kinematics With Recent Plate Velocities From GPS-Geodesy
Already estimated Indian Plate motion at the rate of 37 \pm 0.2 mmyr$^{-1}$ towards NNE direction with respect to Eurasian Plate has been revalidated with a new global network spreading the geographical and azimuthal coverage, which almost includes all the plates surrounding India. 8 years of GPS data from September 1995 to mid 2003 from the Hyderabad IGS GPS Permanent Station (HYDE) have been processed in the global network solution along with the data from other 8 stations. The baseline lengths from Hyderabad to other chosen sites and the rate of changes were also estimated. The angular velocity of Indian plate motion with respect to ITRF-97 reference frame and Indo-Eurasia plate pair have also been estimated. The global network solution has resulted in the estimation of the pole of the angular velocity vector of India with respect to Eurasia to be about a pole of rotation at 29.44 \pm 1.2\deg N, 13.2 \pm 7.3\deg E with an angular velocity of 0.356 \pm 0.035\deg Myr $^{-1}$. Our results mostly conform to the REVEL-2000 Plate motion model but differs considerably from NUVEL-1A and other earlier studies. This departure could be attributed to the difference in geologic and geodetic estimations. The longer time span of GPS data from the central part of India yields more accurate estimations and the analysis is in the global network solution, which doesn't take into account the plate- interior site velocities if at all there are any differences from the plate velocities. The estimated strain accumulation in the south of Indian peninsula shows an increase in the elastic strain accumulation at the rate of 1.27x10$^{-8}$yr$^{-1}$.
G21A-0108 0800h
Evaluation Of Elastic Strain Accumulation In The Southern Indian Peninsula By GPS-Geodesy
The computed elastic strain accumulation in the southern Indian peninsula from the GPS derived velocity fields of the global network of GPS stations, in and around the Indian plate which includes Maitri, Indian Antarctic Station, show a significant departure from rigid plate behaviour in a manner consistent with the mapped intra plate stress field, observations of deformations and seismicity in the region. Our results of intraplate strain accumulation within Antarctica Plate covering three sites MAIT, CAS1 and DAV1 are 1.8x10$^{-9}$yr$^{-1}$, 1.6x10$^{-9}$yr$^{-1}$ and 1.1x10$^{-9}$yr$^{-1}$, respectively. Similarly, the estimates of interplate strain accumulation between Antarctica and other plates Somalia (SEY1), Africa (HARO), Australia (YAR1), and diffuse plate boundary between India and Australia (COCO) are found to be 1.1x10$^{-9}$yr$^{-1}$, 1.0x10$^{-10}$yr$^{-1}$, 1.27x10$^{-8}$yr$^{-1}$ and 1.18x10$^{-8}$yr$^{-1}$, respectively. These estimates are in good agreement with the earlier studies on estimation of global strain rate. The combined GPS and seismic analysis confirm the emergence of diffuse plate boundary between India and Australia and relates to the late Miocene Himalayan uplift. The calculated stress field in the West of the Indian Peninsula has a roughly N-S directed tensional and E-W oriented compressional character and the velocity vectors of all other sites throw a significant insight into the plausible causes of the strain accumulation processes in the Indian Ocean and the northward movement of Indian plate.
G21A-0109 0800h
Application of DInSAR in Monitoring the Metropolitan Land-Surface Deformation: Jungli Industry Park as an Example
Measuring the crustal deformation caused by human activity is difficult, because it's local and unsteady. Moreover, human activity is generally concentrated in urban areas, where cover and shadow from buildings raise errors in GPS observation. In this study, we try to use the InSAR technique to overcome this task. Ten ERS1/2-SAR images recorded in the last ten years are used for determining the land-surface deformation in the Jungli Industry Park in northern Taiwan. The interferometric results reveal a circular-shape subsidence, which is situated within center of this industry park. Quantitative analysis shows that this subsidence event occurred from 1996, reached its maximum rate around 1999, and ceased after 2000. The rapid change of subsidence rate and the circular-shape distribution indicate that the artificial influence, for example, groundwater pumping may be important factor for the land subsidence of the Jungli Industry Park. In fact, our field investigation shows that a steel factory located within the park center had largely pumped groundwater for its cooling system since 1996. This factory has moved out in 2000. Therefore, we can believe that the land-subsidence of this area occurs as dewatering of the acquifer below reduces the buoyant support of the sediments. A numerical simulation of land subsidence in terms of discharge/recharge of groundwater has also been carried out in this study. This simulation can substantiate in advance the relationship between the land deformation and the groundwater influence for the Jungli area.
G21A-0110 0800h
Seasonal Variation Of GPS Height Determination In Taiwan
The positioning accuracy of the Global Positioning System (GPS) has been considerably improved during the past twenty years. The sources of the major error such as ionospheric refraction, orbital accuracy, antenna phase center variation, signal multipath, and tropospheric delay have been reduced substantially if not eliminated completely. With the present methodology of GPS data analysis in geodesy, positioning accuracy is on the level of 1-V2 mm in horizontal coordinates and 5-V10 mm in the vertical coordinate. There are two major reasons for the poor accuracy in the GPS height. The first one is associated with a theoretical limit due to the satellite geometric distribution in the sky since observations are made within a minimum elevation angle (typically about 15 >X). The other one is due to tropospheric path delay, especially associated with inhomogeneity of water vapor (or wet path delay). Taiwan is located in a geographically unique zone with complex topography and abound of water vapor in the air, but spatially and seasonally dependent. In the study, seasonal effect on the GPS height accuracy is investigated. The GPS data were collected from continuously operating reference stations by International GPS Service (IGS), Ministry of the Interior (MOI), Central Weather Bureau (CWB) and Industrial Technology Research Institute (ITRI) of Taiwan. The research of the relationship between variance of the GPS height and change of climate is carried out by computing the GPS data collected in July and December 2003. The comparison in results by using the Saastamoinen model and Hopfield model for correcting the influence of the atmospheric path delay is analyzed. The difference between results from two models is about 4 mm, while the maximum difference in the monthly average of GPS height between July and December 2003 is about 20 mm. The corresponding daily maximum difference is 60 mm.
G21A-0111 0800h
Dense GPS Array Observations Across the Nankai Subduction Zone, Southwest Japan
Interseismic deformation in an oblique subduction zone is a mixture of short-term crustal shortening in the direction of plate convergence and permanent margin-parallel movement of a forearc block. We have deployed two dense GPS traverse arrays across the southwest Japan arc to better illustrate strain partitioning in the Nankai subduction zone. In 1998 we constructed the first array that composed of 22 stations along a 200km-long arc-normal line. The second array with 15 stations was constructed in 2002 nearly parallel to and 120km west of the first one. Both arrays cross the Median Tectonic Line (MTL), the arc-parallel strike-slip fault system dividing the forearc block from the rest of the overriding plate. More than 100 crustal velocities from these arrays and the nationwide continuous GPS arrays are used for inversion analysis to estimate back slip vectors on plate interface, a rate of margin-parallel forearc movement, and slip deficits on the upper fault zone of the MTL. Plate interface and MTL fault plane are reproduced by multi-rectangular segments and inversions are conducted for various dip-angles of the MTL. The optimum model shows that strong plate coupling causes a margin-parallel forearc movement at a rate of 3mm/yr together with a crustal shortening of 0.3-0.4 micro strain/yr in the direction of plate convergence. Slip deficits on the MTL are nearly equivalent to the rate of margin-parallel forearc movement in the opposite direction, showing a full locking of the upper fault zone of the MTL. Moreover inversion result favors northward dipping fault segments of the MTL throughout from east to west than the vertical. The rate of margin-parallel forearc movement is consistent with the geological slip rate of the MTL in the late Quaternary.
G21A-0112 0800h
Accuracy of absolute gravity measurements when measuring crustal deformation
To observe secular land movements of the order of a few millimetres per year, a very precise instrument with long-term stability is required. This can be achieved using absolute gravimeters, which do not depend on a reference frame. Vertical land movements would modify the gravity at a rate of about -10 nms-2 (1 $\mu$Gal) for 5 mm of uplift. Repeated absolute gravity (AG) measurements have now been performed at several sites for five or more years to constrain, for example, tectonic deformation and post-glacial rebound. It is often assumed that these time series contain only white noise. However, many geodetic data sets have now provided evidence for error sources that introduce large temporal correlations into the data. One common statistical model for many types of geophysical signal (which may contribute to the noise) maybe described as a power-law process. Accounting for the type of noise is very important when estimating gravity variations and their uncertainties. Moreover this can contribute to identify the noise source(s) affecting AG measurements We present an analysis of the error model using 10 years of AG data taken at the Proudman Oceanographic Laboratory (POL) and using 96 AG gravity measurements recorded over a period of 8 years at the Membach station (Belgium). A superconducting gravimeter (SG) is also continuously monitoring gravity variations at this site. The AG set-up white noise is estimated by comparison with the SG series. The coloured environmental noise is estimated using the Maximum Likelihood Estimation (MLE) technique to fit two types of stochastic model to the SG time series, power-law noise and first order Gauss Markov (FOGM) noise. The gravity rate of change and the associated uncertainties as a function of the noise structure are computed. Then we investigate the noise of AG values at frequencies higher than 1 cpd, where a white noise component usually dominates. Finally the POL and Membach experiments are applied to estimate the uncertainties of AG campaigns repeated once or twice a year to monitor crustal deformation. The results from repeated AG campaigns along a profile across the Ardenne and along the UK coastline are presented. Such repeated AG measurements should allow one to constrain gravity rate of change with an uncertainty of 1 nms-2 (or 0.5 mm) after 8 to 15 years. The conclusion is that long-term measurements using absolute gravimeters are appropriate for monitoring slow vertical tectonic deformation.
G21A-0113 0800h
FIRST INGV CNT GPS PERMANENT STATIONS FOR MONITORING CRUSTAL DEFORMATION IN ITALY
The National Institute of Geophysics and Volcanology in 1999 started to set up a first set of new continuosly operating GPS stations. The aims are to detect the geodinamic pattern at regional scale in the mediterranean area and densify the existing CGPS network belonging to ASI and to support the not permanent networks. Six stations (MAOO INGP RSTO INGR VVLO MTTM) are located in central appennine belt and one (GIBI) in northern Sicily. Geodetic pillars are located in the outcropping geological units and across tha main seismogenic sources of historical and recent eathquakes. Different monuments was planned, depending on local lithology, with 3D self centering system for precise antenna set up designed at INGV Daily 24 hours DAT and RINEX data at 30 sec sampling rate are locally collected by Trimble Reference Station (www.trimble.com) and transferred by telephone line to a server located at INGV. Data are stored in specific directory and compressed by Hatanaka procedures. Data transfer, storage and compression are controlled by automated tasks implemented on Windows operative system. Data are daily quality checked by TEQC software (UNAVCO) and processed with Bernese 4.0 software, producing daily and weekly coordinate solution in sinex format and velocity and strain field solution in sinex format.
G21A-0114 0800h
Preliminary Results From the COMET CGPS Network in Greece
The Hellenic subduction zone accommodates the convergence between the oceanic lithosphere of the African plate and the extending continental lithosphere of the Aegean. Former studies (e.g. Jackson & McKenzie, 1988) concluded that earthquakes can only account for 10% of the relative motion, suggesting that the subduction is in a stable sliding state. In contrast, some authors (e.g. Pirazzoli et al., 1982), using the Holocene geological record, suggested that large earthquakes (M$>$8) may have occurred in historical times, indicating that the zone may be locked. In order to discriminate between these two views, a network of permanent GPS stations is being developed throughout the south-west Aegean region to monitor the strain accumulation and release along the interface and over the surrounding region, and potentially detect aseismic slip events as seen in others subduction zone (e.g. Dragert et al., 2001). We analyse the CGPS network together with a subset of IGS and EUREF sites in Europe on a routine basis to derive the daily solutions and time series. We present the analysis of the time series and results for sites having about 18 months of data. Preliminary estimates of velocities suggest that the current convergence rate accommodated across the Hellenic trench is $\sim$35 mm/yr, in agreement with McClusky et al. (2000, 2003). No shortening perpendicular to the trench is found between sites located in the southern Peloponnese and Crete relative to sites located further landward. The results are compared to the seismicity of the region in order to reassess the ratio of seismic and aseismic deformation. Moreover, the geodetic results confirm the presence of strike-parallel spreading along the arc at $\sim$3.5 mm/yr between south-western Peloponnese and western Crete, which correlates with several extensional structures described by Armijo et al. (1992) and earthquake focal mechanisms.
G21A-0115 0800h
New constraints on the present-day kinematics of the East African Rift from GPS and earthquake slip vector data
The East African Rift (EAR), a major 5,000 km long and up to 1,000 km wide tectonic structure that marks the extensional boundary between the Nubian and Somalian plate, is interpreted either as a wide zone of uniformly distributed, diffuse deformation, or as a mosaic of microplates. Testing these models and quantifying the present-day kinematics of the EAR has so far resited investigation because of a critical lack of geodetic data within the EAR as well as on the surrounding Nubian and Somalian plates. Here, we present an updated GPS velocity field covering the Nubian and Somalian plates and combine it with earthquake slip vectors along the EAR in a joint inversion. Our objectives are to better constrain the Somalia/Nubia plate motion and to try to resolve block motions within the plate boundary zone. We find a Somalia/Nubia angular velocity similar to the one proposed by Fernandes et al. (EPSL, 222, 2004). We show that Tanzanian craton, nested between the western and eastern branches of the EAR and underlained by an upper mantle plume, can be modeled as an independent block, rotating counterclockwise w.r.t. Nubia. We discuss the implications of this kinematic model on the tectonics of the EAR.
G21A-0116 0800h
InSAR Measurements of Ground Deformation Related to the 2003, May 21 Mw = 6.8 Zemmouri, Algeria Earthquake
The Mw=6.8 Zemmouri earthquake was the largest earthquake felt in the region around Algiers since 1716. The quake occurred on a thrust fault within a system of folds and thrusts in northern Africa resulting from convergence between the African and Eurasian plates. Modern seismicity in the Atlas Mountains of northern Africa indicates that this fold and thrust belt is active, although the partitioning of strain within the belt is poorly understood. Because of the high population density in northern Algeria and tragic loss of life resulting from the Zemmouri earthquake, it is important to use all seismological, geodetic, and stress analysis tools available to study this earthquake and analyze its implications for future seismic hazards. We have processed and are analyzing the InSAR data from the Canadian Space Agency RADARSAT-1 satellite. We gained access to the data through the Alaska Satellite Facility with support from NASA and the Office of U.S. Foreign Disaster Assistance. The data reveal the two-patch nature of the rupture associated with the Zemmouri earthquake. This is in general agreement with field work by Meghraoui et al. (GRL, in press) that used shoreline uplift, GPS, and leveling measurements to show that the Zemmouri rupture occurred on two patches, one east and one west of the epicenter. Although InSAR coherence in the area around the epicenter is poor, as much as 0.45 m of InSAR measured uplift can be seen in the area of Boumerdes in the patch west of the epicenter. This is the same area where Meghraoui et al. measured shoreline uplift that was generally between 0.4 and 0.6 m, but as much as ~0.8 m. We are incorporating the disparate data sets and investigating how the addition of InSAR data might offer improvement over models derived from field data alone. We have also calculated patterns of stress transfer caused by the 2003 Zemmouri earthquake. Preliminary results reveal that the Zemmouri quake increased stresses on the thrust fault system in the Mitidja Basin south east of Algiers, especially on the thrust faults immediately southwest of the 2003 rupture zone.
G21A-0117 0800h
Elastic Versus Non-Elastic Co-Seismic Strain Response at the Trizonia Borehole Sacks Evertson Dilatometer in the Rift of Corinth, Greece, for a Local, Moderate Earthquake.
The western part of the rift of Corinth, in Greece, is the site of continuous geophysical monitoring of its seismic and aseismic activity, in the frame of the Corinth Rift Laboratory project (CRL). The objectives of CRL is to contribute to a better understanding of the relationship between faults, fluids, and earthquakes. Among the various sensor arrays in operation (seismometers, accelerometers, strainmeters, tiltmeters, pore pressure in deep boreholes,.) , a Sacks-Evertson borehole dilatometer has been installed in the Trizonia island, 10 km above the roots of the major, north-dipping active normal faults. We present here the analysis of the dilatometer record of a local magnitude Ml=4.6 earthquake, which produced a clear coseismic static strain step. The simple modeling of the coseismic static deformation suggests a dominant non-linear response caused by the strong S-wave shaking. However, it is still possible to recover information on the source through the analysis and modeling of the intermediate and near-source terms recorded on the dilatometer before the S arrivals. The results are compared to the records of several sensors installed at less than 1 km away (an STS2 seismometer, two long base hydrostatic tiltmeters, and several short period tiltmeters).
G21A-0118 0800h
Comparison of Estimates of Coseismic Displacement From the 2003 M 6.5 San Simeon Earthquake
Estimation of crustal deformation requires trading off solution precision and latency. Between large earthquakes the relative station velocities of southern California stations are small, on the order of a few mm/yr, and the deformation across the entire region is about 50 mm/yr. Estimation of such interseismic crustal deformation rates requires the use of the best GPS software, precise orbits which are available with a delay of one week, and reference frame stabilization; such processing strategies yield precisions of about 1 to 2 mm in the horizontal and about 3 to 4 mm in the vertical. Coseismic offsets for large earthquakes, however, are tens or hundred of mm, or more, for stations near the epicenter, and it is possible to sacrifice a few mm of precision to obtain a rapid solution. Such rapid estimates are useful to both scientists and emergency response agencies. The M 6.5 San Simeon, California, earthquake occurred on December 22, 2003, at the northern edge of the Southern California Integration GPS Network (SCIGN). The nearest station (CRBT) was 34 km from the epicenter, and its horizontal displacement was about 43 mm to the south and 35 mm to the west. The U.S. Geological Survey (USGS), the Scripps Institution of Oceanography, the Massachusetts Institute of Technology (MIT), and the Jet Propulsion Laboratory (JPL) estimated the coseismic displacements for the SCIGN stations. These groups used different software, stations, orbits, processing strategies and parameters. After obtaining rapid solutions, the groups re-analyzed the data using better orbits and more time-consuming processing strategies. Mean differences between pairs of solutions range from sub-mm to 4.5, 6.1, and 8.3 mm in the north, east, and vertical, respectively. Estimates of the coseismic displacement of station CRBT differed by up to 7.3, 5.2, and 10.1 mm in north, east, and vertical components, respectively. Otherwise identical processing with the IGS ultrarapid and precise orbits yielded differences of 0.3, 0.1, and 4.9 mm in the north, east, and vertical; these differences are minor compared to the coseismic signal.
G21A-0119 0800h
InSAR and the Hector Mine Earthquake
A series of SAR interferograms of the southwestern Mojave Desert record the apparent development and collapse of a topographic anomaly near the epicenter of the Oct. 16, 1999 Mw7.1 Hector Mine Earthquake. Interferograms generated from Feb to Sept 1999 ERS-1 data show an elliptical uplift, centered about 20 km north-northeast of the epicenter. The uplift apparently developed between Aug 11 and Sept 15, 1999, as it is not observed in an interferogram generated from Feb 17, 1999 and Aug 11, 1999 data. The uplift covers approximately 2700 km2 and a regional amplitude of 2.8 cm; peak amplitude is nearly 7 cm. The uplift is partially obliterated on a Sept 15 and Oct 20, 1999 interferogram (the coseismic interferogram), and completely absent on three post-event interferograms. NOAA data record no rainfall in the area of the anomaly for October 1999, suggesting that the deformation may not be related to atmospheric anomalies. If this pattern of deformation can be documented for other seismic events, InSAR would be validated as a powerful new tool in earthquake prediction
G21A-0120 0800h
Coseismic Deformation From the 2002 Mw 7.9 Denali Fault Earthquake: Insights From SAR Speckle Tracking
The 3 November 2002 Mw 7.9 Denali fault earthquake in central Alaska presents both unique opportunities and challenges for crustal deformation studies using synthetic aperture radar. While synthetic aperture radar interferometry (InSAR) can produce accuracies similar to those of GPS data, it has an inherent limitation: if the deformation gradient is too large, image coherence is destroyed. The displacements of the Denali fault earthquake greatly exceeded the limit. In addition, the fault passes beneath glaciers and rugged topography exists to the south. These factors combined to create many areas of incoherence around the fault. By exploiting offsets in correlation between images rather than phase differences, the technique of speckle tracking can overcome coherence problems. While not as accurate, speckle tracking can provide data where InSAR cannot. Here we present the results of a speckle tracking study of the Denali fault earthquake for an area surrounding the intersection of the fault with the Trans-Alaska Pipeline. We were able to use data as close as 1.2 km from the rupture. A fault-normal profile near the junction of the fault and pipeline shows a maximum of about 4 m of displacement across the fault in the radar^s line-of-sight (LOS). Values along the profile agree well with those predicted by a GPS-derived model except in an area north of the fault where no GPS data was available. A series of along-strike profiles reveals an increase in displacement across the fault from about 3.5 m in LOS at the western end of the image to about 5 m in LOS at the eastern end. These values show general agreement with the GPS model except for a region to the north of the fault and in the eastern portion of the image where GPS data did not exist. Image displacements and the measured geologic offsets are similar until the eastern end of the image, where the image values are larger. This area corresponds to the section of the fault covered by the Canwell and Gakona Glaciers.
G21A-0121 0800h
The 2004 Cascadia Slow Earthquake
Three years ago, continuous Global Positioning System observations revealed the stunning periodicity of recurring silent earthquakes along the Cascadia subduction zone, observed at stations that surround Puget Sound and the Straits of Georgia in the region near Seattle, Wa. and Victoria, B.C. These events are now recognized to be more widespread and occur at different frequencies in different regions along the length of the subduction zone. This results in sporadic or partial histories of creep events in intervening areas, some of which are only sparsely monitored. The frequency and timing of events in the Puget Lowlands supported a prediction that the 2004 event would occur during the late spring or early summer. Displacements in southern Washington were observed first, and propagated to the northern Puget Lowlands by mid-year. Alternatively, southern Washington sometimes appears to act in concert with events that are evident in Oregon and out of phase with the Puget Lowlands events; thus this could also be interpreted as a separate event on the southern flank of the arch in the down going slab. The displacements of individual GPS stations yield a coherent displacement field that shows episodic extension across the fore-arc region. The motion is sub-parallel to and in the opposite direction from the plate convergence motion and consistent with elastic recovery of strain caused by a discrete creep event along the plate interface.
http://www.geodesy.cwu.edu
G21A-0122 0800h
Cycle-Up of Multiple Rifting Event Models: How Long Does it Take to Reach A Steady State Stress?
Many numerical models are initiated with a background stress state of zero. Often these numerical results are compared directly to geodetic data. Recent work has shown that modeled deformation rates can change as the model is `cycled-up' following repeated earthquakes or rifting events. In this study, we investigate model cycle-up in the context of time-dependent deformation following rifting during the 1975-1984 Krafla eruption in Iceland. We consider the number of rifting cycles required for complete cycle-up, variations in cycle-up time at different locations in the model, background stress magnitudes in fully cycled-up models, and errors incurred when the models are not properly cycled-up. The modeling is done using the commercial software ABAQUS. An ABAQUS user-subroutine is used to apply repeated rifting events within the finite element model. We have generated various 3D models with different fault/rift geometries. The models include (1) a straight rift oriented perpendicular to the far-field velocity boundary conditions, (2) a rift oriented at an angle to the far-field velocities, (3) a model containing two intersecting rifts, one perpendicular to the far-field velocities and the other rift intersecting the first at an angle, and (4) overlapping rift segments in which the overlapped region is bounded by strike-slip faults. We find that different locations in the model have different cycle-up times and steady-state stresses. There are different factors that contribute to model cycle-up. Changes in rheology, far-field boundary conditions, and rifting pattern cause variations in the cycle-up time at different locations in the model. For example, for points in the viscoelastic half-space of the straight rift model, the cycle-up time varies from 4 cycles to 16 cycles as we descend along a vertical line from the base of the rift to the bottom of the model. For a viscosity of ~1.2*1019 Pa-s, steady-state stress values for the same points varies from 0.1 to 0.8 MPa. Similar variations (1.2-3.0 MPa) can be seen for different points in the elastic crust. American Geophysical Union 2004 Fall Meeting San Francisco, California
G21A-0123 0800h
Kinematics of the Sierra Nevada and Oregon Crustal Blocks
Northern California contains the intersection of the NW moving Sierra Nevada block and the clockwise rotating Oregon block. Adding to the tectonic complexity of the region are the interactions of the Pacific plate, Basin and Range, California Coast Ranges, Mendocino fracture zone, San Andreas fault and the Juan de Fuca plate. Our research focuses on how the interactions of these features influence the motions of the Oregon and Sierra Nevada blocks. We processed Global Positioning System (GPS) data collected during the 1998 and 1999 National Geodetic Survey High Accuracy Reference Network (HARN) surveys and by us in 2003. Our preliminary analysis of the velocity field indicates that the Oregon crustal block is rotating clockwise relative to the Sierra Nevada block around a pole approximately west of the Mendocino Triple Junction. Our future work involves finding a more exact location of the relative pole rotation between the blocks, the degree of rotation, and how this motion is being taken up.
G21A-0124 0800h
Spatial Variations in Strain Inferred by GPS Across the Yucca Mountain Region, Southern Nevada
We present the results of processing 4.5 years of data from 28 BARGEN GPS stations in the region of Yucca Mountain. The GPS network includes 16 stations densely spaced around Yucca Mountain. Far-field stations are located in the southern Basin and Range and also across the Eastern California Shear Zone (ECSZ) at the latitude of $\sim$36$\deg$N. Velocity estimates for the local Yucca Mountain stations, relative to a station (LITT) $\sim$15 km to the SE of Yucca Mountain, generally trend NW (after North American plate rotation has been removed). Velocities increase in magnitude from east to west. The total change in velocity across the local network is 1.0 $\pm$ 0.1 mm/yr (from TIVA to BULL). There is also a sudden change from higher to lower velocities $\sim$10 km to the east of Yucca Mountain. Shear strain magnitude for the 16 stations $<$55 km from Yucca Mountain is estimated to be 16.7 $\pm$ 0.7 ns/yr, with right-lateral simple shear at $\sim$N30$\deg$W. This strain rate assumes a uniform strain field, however, which is not strictly the case. For the 9 stations to the west of the sudden change in velocities (to the east of Yucca Mountain) the shear strain magnitude is 13.4 $\pm$ 1.2 ns/yr. For the 7 stations to the east this is 22.0 $\pm$ 0.9 ns/yr. We estimate a total slip budget of 12.3 $\pm$ 0.2 mm/yr across the width of the GPS network, in a profile trending NE from the southern Sierra Nevada, through Yucca Mountain to the Nevada - Utah stateline (station LIND to station ECHO). There is little evidence for extension across the central Basin and Range stations, with a dilitation rate of only -0.3 $\pm$ 0.6 ns/yr, but the Basin and Range stations exhibit a right lateral shear strain of 8.7 $\pm$ 0.4 ns/yr. We estimate a total slip budget of 11.6 $\pm$ 0.2 mm/yr across a profile through the ECSZ stations (at latitude $\sim$36$\deg$N), from the Sierra Nevada to Las Vegas (station LIND to station APEX). We attempt to establish the distribution of slip across the ECSZ faults using elastic displacement models to create model profiles that are compared with the measured velocity profiles. Across the stations in our network at $\sim$36$\deg$N it is possible to produce a reasonable profile fit by allocating $\sim$3.9 mm/yr slip rates to the Owens Valley, Panamint Valley-Hunter Mountain and Death Valley - Furnace Creek fault systems, with locking depths of 8, 12 and 16 km respectively. Across the central part of our network, at the latitude of Yucca Mountain, it is difficult to produce an adequate model fit to the GPS results. This difficulty is a result of the sharp decrease in velocity estimates at stations to the east of Yucca Mountain. A possible cause of this change in velocities could be a local fault at Yucca Mountain or Bare Mountain. The addition of such a fault, with a slip rate of $\sim$0.5 to 1.0 mm/yr, produces a modeled profile that better fits the GPS results. The addition of such a fault also increases the modeled strain rates to levels closer to the measured strain rates, although not enough to completely explain the discrepancy. We therefore suspect that the models are not sophisticated enough to explain the deformation profiles, or that postseismic deformation (from the 1999 Hector Mine earthquake or 1992/2002 Little Skull Mountain earthquakes) is affecting the strain field at Yucca Mountain. Postseismic deformation from the 1999 Hector Mine earthquake would not be unlikely, since we observed a coseismic offset in the timeseries for this earthquake, although there is no clear evidence for postseismic deformation in the timeseries. Interseismic deformation on the faults of the Mojave Desert may also be affecting results.
G21A-0125 0800h
Ground motion measurement in the lake Mead (Nevada, USA) area by temporal analysis of multiple interferograms.
SAR interferometry has proven to be a reliable method for detecting small displacements due to ground subsidence. In this study, we propose to measure ground motion around the lake Mead (Nevada, USA) using InSAR. This artificial lake has been filled with water in 1935. An earlier studie, based on levelling measurements, has shown that the lake impoundement has induced a subsidence of 17 centimeters (Kaufmann et al., 2000). This relaxation process is analogous to the postglacial rebound, but at a smaller scale. To quantify the deformation and constrain the crust and mantle rheological parameters in the lake area, we have analysed multiple interferograms (245) based on 45 ERS images between 1992 and 2001. The interferometric phase contains information about deformation occurring between two satellite passes, as well as satellite orbits errors, topographic, and atmospheric artefacts. The topographic signature is removed using the 3-arc seconds SRTM data. To correct for orbital errors, we remove a best fitting linear ramp. Atmospheric artefact, in our interferograms, are mainly due to the variation of water vapor vertical stratification between the two passes. This results in a interferometric phase correlation with altitude which we remove by minimization. These corrections are then refined through an iterative procedure and validated using data from global atmospheric models. Corrected interferograms are then inverted to solve for deformation using a method based on the large spatial coverage of coherent pixels, allowing to strengthen the signal to noise ratio (Schmidt and Burgmann, 2003). This data inversion provides a time series of the expected deformation in the lake Mead area. The analysis of the deformation evolution during the period covered by the ERS satellites (1992-2001) shows a correlation between the vertical motion and the water level changes. So, we observe a subsidence of up to 1.5 cm between 1996 and 1998, followed by an uplift due to the drop of the water level after 2000. Apart from this deformation, we observe a local anticorrelated movement in a 20*10 km area located North of the lake with a significative uplift of 1.2 cm during the period of the lake area subsidence. Finally, the west side of our ERS scene shows a small subsidence located in Las Vegas. This ground motion is independant of the lake water load variation.
G21A-0126 0800h
Estimation of tectonic stress rates from NeoKinema models in southern California
We applied 2-D kinematic F-E program NeoKinema to estimate long-term-average velocity, fault slip rates, and strain rate field in southern California. We use weighted least-squares to fit the input data (geological fault slip rates, geodetic benchmark velocities, and horizontal principal stress directions) and invert the velocity field. The grid is composed of mostly 7-km spherical-triangle finite element and 4-km fault bands. Geological fault slip rates, geodetic velocities, and stress directions are from California Geological Survey 2002, SCEC Community Motion Map3.0, and World Stress Map 2003, respectively. We have calculated $\sim$60 models to explore two tuning parameters to find the optimal model. Currently the best model has RMS discrepancies of $\sim$1.6 sigma for geodesy, $\sim$0.9 sigma for fault data, and $\sim$0.35 sigma for stress direction. The long-term-average velocity field is continuously/self-consistently corrected for temporary fault locking by summing the contributions of the faults that move freely at a constant rate below a locking depth with the slip rate determined in the optimized NeoKinema model. At present a constant locking depth is used for all faults and regular dislocation patches with constant dip angles are used in the correction. We do not consider the locking contributions from seismic slip in non-faulted elements as they are likely small. We computed strain rate from the "corrected" velocity field. The tectonic stressing rate is computed from strain rate assuming constant elastic modulus. Current results show that maximum tectonic shear stress accumulation concentrates around the major fast-moving faults such as San Andreas, San Jacinto etc. and that stress rate decays away from the faults. The newly derived tectonic stressing rate provides a better estimate of stress from plate tectonics since it utilizes the information from both geodetic and geologic data without preassuming any block dynamics. The comparison with tectonic stress rates estimated from rigid block models would expect to reveal new insights provided by this kinematic approach. The new estimation of tectonic stress accumulation, along with incremental stress release from each earthquake in the catalog, will allow us to test various earthquake interaction and triggering hypotheses.
G21A-0127 0800h
Use of a Principal Component Analysis to Identify Precise Tectonic Rates From GPS Time-Series in the Ventura Basin
The region in the Ventura basin is one of the fastest converging areas in Southern California. Convergence rates in the basin and the Santa Barbara channel from geodetic measurements are ~6-8mm/yr, with uplift rates estimated at 1-2mm/yr in places measured by leveling. These low uplift rates can easily be masked over the short time periods of GPS measurements by strong seasonal signals that may differ from station to station and result from differing processes. Similar to the Los Angeles basin, the seasonal signals in this area are almost certainly caused by fluid injection/removal, and are not related to fault motion. By identifying and removing these signals we can better estimate the interseismic strain rates and aid in estimating fault locking depths. A principal component analysis (PCA) is used to isolate the tectonic signals in the GPS time-series from Southern California Integrated GPS Network (SCIGN) stations within the Ventura basin, with measurements that span a two year period. Using data products available from the regional processing centers, the PCA is applied to the regionally filtered time-series to break down the signal and identify the different components. Our method of post processing requires that all of the data be simultaneous, and have equal time steps. Since zero or null values are not allowed in the time-series data, the data gaps are filed via interpolation. This approach stacks the data within the network and breaks it down into the dominant patterns common to all of the time-series data within the region. As a control we test the approach against a set of synthetic time-series that contain variations in linear rates, annual signals and random noise. Our analysis of the Ventura Basin, as well as the synthetic data shows that this process is most effective at identifying and removing strong isolated signals present at 1 or 2 stations within the network, as opposed to signals spatially and temporally correlated across the network, and is useful in removing scatter in the data, resulting in a better determined rate.
G21A-0128 0800h
Improved GPS time series in Southern California through the combination of SOPAC and JPL solutions
The Southern California Integrated GPS Network (SCIGN) is now in full operation and provides valuable information on deformation due to tectonic, geomorphic and hydrologic processes. We combine the daily loosely constrained solutions from SOPAC and JPL analysis centers to derive integrated time series to enhance the signal to noise ratio and to provide versatile services to tectonic researchers. We investigate the process model and strategy of the two analyses centers first and try to correct the systematic differences between the solutions of the two centers before the formal combination. Our combination methodology, the definition of the reference frame, the quality control and the final products will be addressed. Our preliminary times series analyses reveals secular tectonic movement, seasonal mass loading caused deformation, and some short term transient variations. In particular, the vertical components of the site position time series also provide useful constraints on tectonic movement and other geophysical processes.
G21A-0129 0800h
Progress towards a multiple-use GPS array in the Puget Sound region
The cooperative Puget Reference Station Network (PRSN) array of continuous GPS stations (CORS or CGPS) in western Washington provides targeted densification of the Pacific Northwest Geodetic Array (PANGA) around the Seattle fault and above an areas of the Cascadia subduction zone known to experience recurring silent earthquakes. Public agencies equip and operate PRSN stations, while the university coordinates siting and provides stable monuments. To date, 8 deep drill-braced, 1 shallow drill-braced, and 3 H-pile monuments enhance this network to geodetic standards for studies of solid earth deformation. Full operation of the network is anticipated for summer 2005, with approximately 18 stations. The network provides both real-time corrections for surveying applications and files for post processing at a nominal 1-second sampling interval. The main scientific goal of the project is to increase the density and quality of CORS in the Puget Sound region, in order to measure strain accumulation across the major structures (e.g., the Seattle fault, regional creep events). This will require monitoring over at least the next few years, yet new stations have already contributed to better understanding subduction-related slow earthquakes. Combined with close seismic monitoring, the new GPS stations indicate that the most recent episode of tremor and slip (ETS) occurred in at least two phases. The first occurred in late March-early April of 2004 and affected southern Puget Sound. This was followed somewhat tardily by a larger, predicted event in northern Puget Sound in July. The two areas appear to correspond to the two limbs of the arch in the subducting Juan de Fuca plate; some past events have shown similar progression but were not as well documented. Advances in real-time horizontal positioning no longer require particularly dense networks, but the PRSN provides denser (~30 km) spacing that resolves subtler deformation. In addition to bracketing structures with better density than the Plate Boundary Observatory (PBO), the network provides redundancy and autonomy for individual agencies, and should facilitate increasing use of GPS for leveling. Some strengths of the cooperative approach have been efficient use of limited resources by surveyors and scientists, as well as fostering communication between public, private and academic GPS users. The main drawback is that the time required to establish a station is approximately proportional to the fourth power of the number of agencies involved. However the end result should be an unusually stable, robust, and flexible network that is not strongly dependent on any one source of funding.
http://www.geodesy.cwu.edu
G21A-0130 0800h
The Plate Boundary Observatory: Operational Status and Data Plans
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. The science goals of PBO require that plate boundary deformation be adequately characterized over the wide range of spatial and temporal scales common to active continental tectonic processes. PBO will meet these needs using 875 continuous GPS sites, 143 borehole strainmeter stations, and five laser strainmeters, all installed over the next five years. In addition, there will be a pool of 100 portable GPS receivers available for survey-mode observations, and in 2005, we anticipate incorporating 225 existing continuous GPS sites into PBO with funding under the PBO Nucleus proposal. These stations will provide raw observations from which PBO Analysis Centers will create a wide range of derived data products, including time series of strain and GPS station position, GPS velocity vectors, and strainmeter and GPS processing auxiliary information. All PBO data and data products will be made available to the community as rapidly and freely as possible through the PBO Archives and ultimately through the EarthScope Data Access System, part of the EarthScope Portal. PBO began operations in September 2003 and the first five new PBO continuous GPS (CGPS) stations were installed in January 2004. Currently, there are 39 stations installed and collecting data, of which 27 are returning data automatically to PBO Headquarters in Boulder on a daily basis. Data from 14 of these stations are available through the PBO GPS Archives at the UNAVCO Facility in Boulder and the Scripps Orbit and Permanent Array Center in San Diego. We anticipate having a total of 50 CGPS stations installed by the end of September 2004 and an additional 200 CGPS stations installed by the end of September 2005. The first PBO borehole strainmeter installation is expected by early 2005, with a total of 16 stations to be installed by the end of September 2005. The first PBO laser strainmeter will be installed in 2005. We will present an update on the current status of, and future plans for, PBO operations, data collection and analysis, and distribution of PBO data products.
http://pbo.unavco.org
G21A-0131 0800h
Isla Guadalupe, Mexico (GUAX, SCIGN/PBO) a Relative Constraint for California Borderland and Northern Gulf of California Motions.
Using ITRF2000 as a common reference frame link, I analyzed survey mode and permanent GPS published results, together with SOPAC public data and results (http://sopac.ucsd.edu), in order to evaluate relative present day crustal deformation in California and northern Mexico. The crustal velocity field of Mexico (Marquez-Azua and DeMets, 2003) obtained from continuous GPS measurements conducted by Instituto Nacional de Geografia e Informatica (INEGI) for 1993-2001, was partially used. The preferred model for an instantaneous rigid motion between North-America and Pacific plates (NAPA), is obtained using results of Isla Guadalupe GPS surveys (1991-2002) giving a new constraint for Pacific plate (PA) motion (Gonzalez-Garcia et al., 2003). It produces an apparent reduction of 1 mm/yr in the absolute motion in the border zone between PA and North-America (NA) plates in this region, as compared with other GPS models (v.g. Prawirodirdjo and Bock, 2004); and it is 3 mm/yr higher than NNRNUVEL-1A. In the PA reference frame, westernmost islands from San Francisco (FARB), Los Angeles (MIG1), and Ensenada (GUAX); give current residuals of 1.8, 1.7 and 0.9 mm/yr and azimuths that are consistent with local tectonic setting, respectively. In the NA reference frame, besides the confirmation of 2 mm/yr E-W extension for the southern Basin and Range province in northern Mexico; a present day deformation rate of 40.5 mm/yr between San Felipe, Baja California (SFBC) and Hermosillo, Sonora, is obtained. This rate agrees with a 6.3 to 6.7 Ma for the "initiation of a full sea-floor spreading" in the northern Gulf of California. SFBC has a 7 mm/yr motion in the PA reference frame, giving then, a full NAPA theoretical absolute motion of 47.5 mm/yr. For Puerto Penasco, Sonora (PENA) there is a NAPA motion of 46.2 mm/yr and a residual of 1.2 mm/yr in the NA reference frame, this site is located only 75 km to the northeast from the Wagner basin center. For southern Isla Guadalupe (GUAX) there is 51.8 mm/yr in the NAPA reference frame. Finally full present day NAPA motion at the Alarcon Rise must be only 50.1 ±0.2 mm/yr in agreement with the lower limit of the NAPA "geological" model obtained by DeMets and Dixon (1999).
G21A-0132 0800h
PBO Strainmeters: Distribution, Design and Data Products
PBO will install and operate up to 143 three-component borehole strainmeters and five long-baseline laser strainmeters along the Pacific-North American plate boundary over the next four years. Drilling for the first borehole strainmeter, located in the Cascadia region of Washington, will start in October/November 2004 with instrumentation being installed in early 2005. Installing the first borehole station in Cascadia at that time should allow us to capture the next anticipated silent earthquake on the Cascadia subduction zone. In the second year of PBO, four new borehole strainmeters will be installed at Parkfield, 11 more in the Cascadia region, and up to four on Vancouver Island. The remaining strainmeters will be distributed so as to improve our understanding of tectonic and volcanic processes along the entire plate boundary. Three-component borehole strainmeters (BSM) measure change in diameter of the borehole in 3 directions oriented 120 degrees apart. The 3 measurements are then combined to determine the change in areal and shear strains. The BSMs will be installed at depths of between 150 m and 240 m. Long-baseline laser strainmeters (LSM) measure change in the relative position of end monuments hundreds of meters apart using an unequal-arm Michelson interferometer. The LSMs will be installed on the surface. Each BSM station will have a 2 Hz, 3 component, passive borehole seismometer installed above the strainmeter making the PBO BSM network the second largest borehole seismic network in the world. Each BSM station will monitor pore and atmospheric pressures, and some will monitor rainfall, temperature, wind speed, and relative humidity. Some BSM stations will also be colocated with PBO GPS stations. The BSM data, collected at 20 samples per second (sps), will be buffered on-site and downloaded in near real time to PBO via direct Internet connections, along with environmental data. BSM data will be transferred to a central quality-checking system and then passed to the Strainmeter Analysis and Archive Centers. LSM data, collected at 1 sps, will be buffered on-site and downloaded at least daily to PBO. It will then be sent to Scripps Institution of Oceanography for analysis and to the archives for storage. Strainmeter data analysis will include: instrument calibration, removal of spikes and offsets, identification of strain induced by changes in atmospheric and pore pressure, removal of borehole relaxation trends in BSM data, and production of an earth tide model. Borehole seismic data will likely be sent to the Array Network Facility (ANF) located at Scripps for analysis, quality-checking, distribution and arching. The ANF also process and analyses data for the USArray arm of EarthScope. PBO strain data products will be divided into 3 levels. The raw data will be considered the Level 0 product. Level 1 data will consist of clean, scaled, gauge data. The Level 2 data set will contain derived products: areal and shear strains, tidal and pressure corrections. Level 2 data will be further divided into 3 sub-levels ranging from a rapid solution produced every 24 hours to a final verified data set updated every 3 months. The data will be available in XML and SEED format from the EarthScope Data Access System. Metadata such as site information and scale factors will be retrievable from the PBO Operational Database.
http://pbo.unavco.org
G21A-0133 0800h
The PBO Nucleus: Integration of the Existing Continuous GPS Networks in the Western U.S.
Tectonic and earthquake research in the US has experienced a quiet revolution over the last decade precipitated by the recognition that slow-motion faulting events can both trigger and be triggered by regular earthquakes. Transient motion has now been found in essentially all tectonic environments, and the detection and analysis of such events is the first-order science target of the EarthScope Project. Because of this and a host of other fundamental tectonics questions that can be answered only with long-duration geodetic time series, the incipient 1400-station EarthScope Plate Boundary Observatory (PBO) network has been designed to leverage 432 existing continuous GPS stations whose measurements extend back over a decade. The irreplaceable recording history of these stations will accelerate EarthScope scientific return by providing the highest possible resolution. This resolution will be used to detect and understand transients, to determine the three-dimensional velocity field (particularly vertical motion), and to improve measurement precision by understanding the complex noise sources inherent in GPS. The PBO Nucleus Project is designed operate, maintain and upgrade a subset of six western U.S. geodetic networks: the Alaska Deformation Array (AKDA), Bay Area Regional Deformation network (BARD), the Basin and Range Geodetic Network (BARGEN), the Eastern Basin and Range/Yellowstone network (EBRY), the Pacific Northwest Geodetic Array (PANGA), and the Southern California Integrated Geodetic Network (SCIGN), until they are subsumed by PBO in 2008. Uninterrupted data flow from these stations will effectively double the time-series length of PBO over the expected life of EarthScope, and create, for the first time, a single GPS-based geodetic network in the US. Other existing sites will remain in operation under support from non-NSF sources (e.g. the USGS), and EarthScope will benefit from their continued operation. On the grounds of relevance to EarthScope science goals, geographic distribution and data quality, 209 of the 432 existing stations have been selected as the nucleus upon which to build PBO. We have begun converting these stations to a PBO-compatible mode of operation; data now flow directly to PBO archives and processing centers while maintenance, operations, and meta-data requirements are currently under upgrade to PBO standards.
http://www.unavco.org/exnet/exnet.html
G21A-0134 0800h
Tectonic Implications of Recent Campaign GPS Measurements Along the Central Region of the Lesser Antilles arc: Results from Dominica 2001-2004
The volcanic island of Dominica is located in the central region of the Lesser Antilles arc, an obliquely convergent boundary between the Caribbean and North American plates. An initial GPS campaign was conducted in 2001 to expand our regional GPS field for the eastern Caribbean and to provide baseline geodetic data for examining volcanic unrest in Dominica. In 2001, nine sites were established, the majority near the southern volcanic region, where a recent shallow seismic swarm had occurred. A second GPS campaign was conducted in 2003, following another seismic swarm in the north. This campaign re-occupied the original nine sites and established three more. The density of GPS sites on the island was improved and all the existing sites were reoccupied in 2004. Today there are eighteen high precision GPS sites on the island. All GPS observations were made with dual-frequency, code-phase receivers and choke ring antenna. At least 2.5 days of continuous observations were obtained on each site for each epoch in 2001, 2003, and 2004. Daily site positions were calculated with an absolute point positioning strategy using GIPSY-OASIS-II and final precise orbit and clock corrections from JPL. The measured surface deformation field on Dominica potentially contains components of motion from both shallow volcanic sources as well as elastic strain accumulation from the plate interface. Because volcanic deformation may be cyclical and superimposed on the background tectonic deformation field, we have chosen to examine on the observations from the far eastern and southern sites on the island, located well away from the region of shallow seismicity and any potentially active volcanic system. Caribbean-fixed residual velocity magnitudes range from 2$\pm$2 to 8$\pm$2 mm$/$yr directed approximately west. Although these results must be regarded as preliminary, they are consistent with simple models of elastic strain accumulation along the plate interface.
G21A-0135 0800h
Deformation along The Cordillera Blanca Fault System inferred from GPS observations
The Cordillera Blanca (CB) is a mountain range in the central Andes of Peru between 8.4°S and 10°S, composed mainly of a 5-12 Ma granitic batholith. West of the range the following geological provinces can be distinguished: the Callejon de Huaylas basin, the Cordillera Negra and a coastal batholith. Bounding the Cordillera Blanca and the Callejon de Huaylas basin, a series of high scarps delineate a normal fault system extending SE-NW for about 200 km. Recent geological studies recognize the system as an active detachment fault with ongoing extensional deformation. We present preliminary results of GPS observations between 2001 and 2004 on 5 stations distributed along the aforementioned geological provinces. Horizontal velocities relative to stable South America range from 6mm/yr on the CB batholith and the coastal block to 1 mm/yr or less on the Cordillera Negra. The pattern of velocities suggest a complex deformation field consistent with extensional motion and possible a block rotation component.
G21A-0136 0800h
The Network Strain Filter - A new tool for monitoring and detecting transient deformation signals in GPS arrays
Data from large-scale continuous GPS arrays have revealed transient signals caused by aseismic fault slip and magmatic intrusion. It is almost certain that more subtle signals due to smaller magnitude events exist, but have gone undetected. The huge amount of data from large GPS arrays makes it difficult to search for time varying processes by eye, and automated methods to detect transients are urgently required. We have developed a time-domain filtering method to detect spatially and temporally coherent signals in data from large GPS arrays, which we refer to as a Network Strain Filter (NSF). The position time series are expressed as a sum of tectonic motions, local benchmark motions, reference frame errors, and unmodeled errors. The tectonic displacements are expanded in a wavelet basis with time varying coefficients. The coefficients are modeled as integrated random walks, such that the velocities (and strain rates) follow a random walk. The wavelet coefficients, as well as the benchmark motion and reference frame errors, are estimated with Kalman filtering techniques. The variance of the integrated random walk, which controls temporal smoothing of the displacements, is also estimated simultaneously using an "extended" Kalman filtering algorithm. Orthogonal wavelets are used to form the basis, which enables us to carry out a multiresolution analysis (MRA). MRA expands a signal into a sum of translations and dilations of a unit "mother" wavelet and a sum of translations of one "father" wavelet whose scale corresponds to the spatial extent of the network. The smallest scale wavelet is chosen so that the wavelet covers some minimum number of stations, which ensures that time-varying deformation is separable from local benchmark motions. We determine the minimum scale such that the residual variance is consistent with the a priori errors of the data. We carried out simulations to test the performance of the filter using a variety of wavelets. Hypothetical time-varying deformation fields were constructed for GPS arrays in Japan and California. Preliminary results show that with appropriate initialization of the temporal smoothness hyperparameter, the filtered displacement field recovers the pattern of the input transient. This demonstrates the capability of the NSF to separate transient from steady state deformation. Analysis of GPS array data from Japan will be discussed.
G21A-0137 0800h
Measuring Short Term Transient Regional Deformation With High-Frequency GPS Network Positioning
Advances in GPS receiver technology now permit geodetic data sampling at very high rates. A time series analysis of instantaneously (epoch-by-epoch) computed components of baselines recently measured at 10 and 20 Hz shows near white noise characteristics for frequencies above 0.5 Hz. It suggests that the same spatial accuracies obtained for 1 Hz sampling of strong motion surface displacement could be obtained with a much finer temporal resolution. Furthermore, temporospatial resolution of short-term transient signals (such as coseismic motion and slow earthquakes) and their detection threshold could be significantly improved by suitably filtered position fixes from high-rate sampled GPS geodetic networks. Zero-baseline tests indicate that receiver noise significantly contributes to the high frequency spectrum. Below 0.5 Hz, spectral energy increases exponentially with decreasing frequency until becoming again frequency independent at periods of several hours. Upgrading dense GPS geodetic networks in seismogenic areas to high rate sampling and data streaming will thus provide a tool to constrain source parameters, examine propagation characteristics, and assess local site response of strong motion events.
G21A-0138 0800h
Seafloor Geodetic Measurements Reveal Locking Within 5 km of the Peru-Chile Trench at 12S
A combination of GPS and acoustic travel time measurements were taken from the R/V Roger Revelle in 2001 and 2003 from two seafloor precision transponder arrays at 12°S off the coast of Peru. The objective was to locate the updip limit of locking between the downgoing Nazca plate and the South America plate; a critical measurement in understanding the earthquake cycle and the potential hazard of onshore tsunamis. High pressure between the plates creates a coupled, or locked zone, where strain increases and deforms the overriding plate. Shipboard GPS, seafloor transponders and an underwater survey package were used to measure the horizontal and vertical deformation above the thrust fault. 3D modeling of horizontal deformation has shown that locking occurs within 5 km of the trench. The initiation of locking at 12°S is thought to be dominated by a lack of sediment and pore water, compared to more southern latitudes of the Peru-Chile trench where locking begins deeper due to increased sediment and fluid volume. This is the first study to measure seafloor deformation above a subduction thrust fault and create a GPS network covering the width of the convergence zone. Land GPS sensors are not sensitive to this limit and seismic observations may need to cover a complete earthquake cycle to illuminate the seismogenic zone.
G21A-0139 0800h
An Experiment on GPS/A Seafloor Positioning in the Central Part of Kumano-nada, Central Japan
Kumano-nada, northeast of Kii Peninsula in the central part of Japan, is a seismogenic zone of the M-8 class Tonankai earthquakes that occurred repeatedly at an interval of about 100 years. The MEXT (Ministry of Education, Culture, Sports, Science and Technology, Japan) initiated a 5-year program in 2003 for seafloor observations in and around Kumano-nada. Nagoya and Tohoku Universities initiated experiments on GPS/A seafloor positioning to monitor crustal deformation in the subduction zone with a focus on investigation of the effect of sound velocity structure in the ocean on seafloor positioning (e.g., Tadokoro et al., this meeting). Tohoku group deployed five precision acoustic transponders (PXPs) jointly developed with Scripps Institution of Oceanography in the central part of Kumano-nada at depths of about 2,000m. The deployed PXPs A, B, C, and D form a diamond on the seafloor, and PXPs C, D, and E form a triangle. Although three typhoons were in the way of our 12-day cruise in August this year, we carried out GPS/A observation for several days. After an observation for locating the precise position of each PXP, we tried to keep the buoy near the center of the diamond or the triangle. The vessel held the position within 20-30m from the center, and the buoy_fs position was kept with 100m from the center. Kinematic GPS positioning is now under processing with GEONET data observed in Kii Peninsula. We also tried monitoring the sound velocity structure with 3 sets of inverted echo sounders (IESs) deployed near the PXPs C, D, and E. The IESs can monitor temporal and spatial variation in the sound velocity structure in the triangle array of PXPs. We plan to carry out the second GPS/A observation in November. The result is worthy of notice. Strange earthquakes of magnitude 6.9, 7.4, and 6.4 occurred on the nearest Nankai Trough axis on September 5-7, 2004. Co-seismic crustal deformation observed by the GEONET was about 4 cm near the coast of Kii Peninsula. The seismogenic zone is in the list ocean drilling under the IODP (Integrated Ocean Drilling Program).
G21A-0140 0800h
GPS/Acoustic seafloor positioning off the northeastern Japan with a new acoustic system
We developed a seafloor positioning system combining kinematic GPS (KGPS) and precise underwater acoustic ranging to investigate dynamics in the subduction of the Japan Trench. The acoustic system for medium water depths is simple enough to use a signal with a single frequency of 10kHz. We deployed 3 precision acoustic transponders (PXPs) off Miyagi Prefecture, where the next earthquake is suspected. In this talk, we present results of GPS/A experiments carried out using the PXPs in August and October, 2003, and in August 2004. We used a towed buoy with 3 or 4 GPS receivers an acoustic ranging system for GPS/A observation in August 2003 and 2004. In October, we put the transducer on a post in the stern of a vessel and determined its location by using a GPS antenna and a motion sensor on the top of the post. Obtained dataset consists of three elements: position of a seasurface transducer, two-way travel times between the transducer and the PXPs, and the sound velocity structure. We calculate the two-way travel time of an acoustic signal of 18.6ms between the transducer and a PXP with a resolution of 0.002ms (= 1.5mm in ranging), by taking correlation between the transmitted and the received signals. During the experiments, we deployed CTD one hour before the buoy operation to measure salinity, temperatures, and pressure. We also performed XBT casts every 1-3 hours to measure temperature. Sound velocity structure is then estimated by using the empirical law by Del Grosso (1974) from these data. There are 2 steps to estimate the location of the center of the PXP array. At first, we estimate the positions of each PXP using data collected while the buoy shifted around each PXP. In the second step, we determine the center of the PXP array using the data collected near the pre-defined center. In order to remove the influence of a ocean tide, we use the tide prediction program called NAOTIDE-J (Matsumoto et al., 2000). The number of shots which can be used in the analysis in the center was only 14 for shortage of the observation period. However, analyzing with this small number of shots, the accuracy of the central position of the array are about 5-6cm in standard deviation. We could carry out GPS/A observation for 8 hours on a vessel in October 2003 and 73 hours on a buoy in August 2004. These data are under processing.
G21A-0141 0800h
A Synthetic Test of Precision in GPS/Acoustic Measurements of Seafloor Positioning
In this decade, much efforts have been made in detecting seafloor movement using GPS/Acoustic ranging method. GPS/Acoustic measurement consists of two parts, one is the positioning of a transducer at sea surface using K-GPS and the other is measuring travel-time or slant-range between the transducer and transponders put on ocean bottom using acoustic ranging. Currently precision of the seafloor positioning reaches less than 10cm. Several possible factors arise, which still decrease the precision, such as GPS positioning itself in 1 Hz, interpolating of exact position of the transducer at an arbitrary timing of acoustic measurement, uncertainty of the sound velocity variation both in time and space as well as depth variation due mainly to ocean tide. A linear or spline interpolation can be applied to undulation of the transducer position when GPS antennas and the transducer are equipped on a ship. On the contrary when using a buoy system, one need additional physical sensors to detect flutter of the buoy in higher frequency. For the sound velocity, one can estimate it based on CTD or XCTD/XBT data. However, such data cannot cover entire period and space of observation. In general, sound velocity thought to varies smaller in lateral than in time. Accounting for this nature, Scripps group developed a strategy that set transponders like triangle and make acoustic measurement with the transducer kept laterally equidistant to the transponders. In this method most of laterally stratified component of sound velocity variation in time will be cancelled to determine the center of the triangle. Prior to this measurement, position of individual transponder must be estimated. Error in this estimation will biases the ``absolute'' final position of the center, which is not important to detect the crustal ``relative'' movement, as long as one take the same estimations in the next survey. However this error increase the misfit of travel-time, which results in the final position to be sensitive to other possible error source. Therefore position of the individual transponder should be solved simultaneously with the center position. In addition, we have to account for the randomness of the error to consider the ``mean distribution'' of the center position as the precision of survey. We will present theoretical estimate of the precision with an application to actual data set.
G21A-0142 0800h
Introduction of new Acoustic Transducer to the Seafloor Geodetic Monitoring System
Institute of Industrial Science, University of Tokyo and Hydrographic and Oceanographic Department, Japan Coast Guard have been constructing the geodetic observation network on the seafloor around Japan. The observation network, which consists of sixteen seafloor geodetic reference stations, has been built along the ocean trench regions by the end of 2003. Two more stations are planned to be added to cover whole area of the Nanakai trough and to focus on the area off Miyagi in 2004. The observation results were accumulated as the observations have been routinely repeated. We are on the stage that we can discuss and improve the system and the observation based on the results. Issues and knowledge that we have through the observations should be fed back to the system and further observation for improvement. One of such issues is one inherent in a cylindrical acoustic transducer that is employed both on the ship-board unit and on the seafloor transponder. Wave length ( 15cm ) of the ranging signal that we use is compatible to the dimension of the cylindrical transducer. This implies that possible ranging error is caused depending on the incident angle of the signal to the transducer. We are planning to take measures against this issue by introducing new acoustic transducer which is designed so that same response is secured within some range of the incident angle. Tank tests were conducted to evaluate the response of the newly developed acoustic transducer in advance of introducing it to the observation system. Ranging procedure has repeated by receiving a signal transmitted from the transducer with hydrophone at points keeping constant distance (150.0cm) from the transducer. Evaluation of the ranging response of the current transducer was also performed under the same conditions in order to compare these responses directly. Some corrections to retain continuity in the seafloor geodetic observation could be deduced from these comparisons. Results of the tests show that the new transducer provided 3-4 cm shorter ranges within the signal incident angles of 0 to 50 degrees than by the current transducer. Ranging with the current transducer indicate that there is a possibility of providing about 15 cm longer range in case of around 0 degree than it really is. We are planning to have further examinations on these transducers to understand their characteristic features fully and improve the seafloor geodetic observation system based on these data.
G21A-0143 0800h
A System for Observing Sea-floor Deformation: System Configuration and the Monitoring at the Nankai Margin, Japan
\ \ \ The Japanese Islands are located close to the plate boundaries, and large subduction earthquakes repeatedly occur at the plate boundaries. The source regions of the earthquakes are beneath the see bottom. It is, therefore, necessary to monitor the crustal activities, such as seismicity and crustal deformation, for the sake of earthquake prediction and disaster prevention. We have developed a system for observing sea-floor crustal deformation using the acoustic ranging technique. In this presentation, we present the general concept of our system, observation sites for the long-time monitoring of the sea-floor deformation, and preliminary results of the monitoring. \ \ \ The observation system is composed of 1) acoustic measurement between a ship transducer and sea-bottom transponders, and 2) kinematic GPS positioning of the observation vessel when the acoustic signal is transmitted. The sea-bottom transponders are set in a 13-inchs glass sphere, and are equipped with the batteries for five-years-measurements. The acoustic measurement is performed by the following procedure. An m-sequence signal with a length of 14.322 ms is transmitted from the ship transducer to sea-bottom transponders. The frequency of the carrier wave is 12.987 kHz. The sea-bottom transponder returns the same signal as it received after a delay time of 1048.576 ms to diminish the reverberation between the sea-floor and the sea-surface. We record the waveform of the returned signal at the ship transducer. The two-way travel time is measured by means of the cross-correlation computation. The travel time is converted to the path length using the sound speed on the basis of the CTD measurements. Combining the travel time data, ship positions, and attitude of the observation vessel, we determine the position of the sea-bottom transponders. \ \ \ We have installed the transponder networks at the Nankai margin and the adjacent region, Suruga bay; large earthquakes are expected to occur in the regions. The water depths at the regions are 800-2,100 m. Each network is composed of two to four transponder arrays. We have repeatedly measured the sea-floor deformation twice a year. We report the preliminary results of the measurements. We also performed experiments for estimating the error factors in sea-bottom positioning, that is, kinematic GPS and spatial changes in sound speed structure. The error in the kinematic GPS positioning behaves quadric increase with baseline length. The spatial variations in sound speed are conspicuous at a depth range from 0 to 500 m, and are 7 m/s at the two positions with a separation of 2 nautical miles.
G21A-0144 0800h
Accuracy Evaluation of Kinematic GPS with a Moving Object for Observing Sea-floor Crustal Deformation
In order to monitor seafloor crustal deformation, we have developed a system with kinematic GPS positioning and acoustic ranging. The major error factors of the present system related to locate seafloor transponders are: 1) estimations of seawater acoustic velocities, 2) detection of acoustic signals, and 3) positioning by kinematic GPS. In our previous study, an accuracy evaluation experiment of kinematic GPS was conducted with baseline lengths of 15km, 30km, and 90km. As a result, the error of kinematic GPS is minimized by 1-2cm when baseline length is 10-30km especially with small satellite arrangement error. In this meeting, we will show the result of our recent experiments with extended baseline lengths. Moreover, this experiment also investigated for horizontal as well as vertical accuracy by moving a rover antenna along a lifting device. The result of this experiment is important and would benefit the marine observation to reevaluate accuracy classified by baseline length of kinematic GPS. This experiment was conducted as follows: we constructed five baselines with varying lengths (30, 60, 90, 110 and 150km). ROVER, an antenna moving on a rail way, was located at Nagoya Univ. In order to know the error of an internal clock of receiver, the external frequency clock was attached in the receiver of each point. The data were logged at 1 second interval with the minimum elevation of GPS satellites as 15 degree. Moreover, in consideration of satellite arrangement, we experimented only during the time with PDOP<=2. A software RTD was used for analysis of kinematic GPS and data of precise orbits provided by IGS was also used. Bernese version 4.2 was used for the position determination of base stations and the rail with one week data. In accuracy evaluation, the position of an antenna was estimated on the rail in reference with the base station installed in Nagoya University (near ROVER). The position of the antenna was also set to a true coordinate value wherein the distance between the true value and coordinate values were calculated with each baseline solution.
G21A-0145 0800h
Ocean Floor Positioning With Temporal Variation in Acoustic Velocity
We are developing an ocean floor positioning system to monitor the crustal deformation on the ocean floor using GPS and acoustic ranging. We determine the location of a vessel by Kinematic GPS positioning to measure the distance between the vessel and a benchmark unit on the floor using acoustic ranging. We have installed benchmark units to carry out a continuous monitoring at Suruga Bay in 2002. In this study, we report the results of the observation using the benchmark units at two sites situated in the northwestern and northeastern part of the Suruga Bay on October and November 2002, respectively. Each site is composed of three units within radius of about 200m. The first measurement was performed during their initial installation and the subsequent measurements were undertaken twice a year for each site. The acoustic ranging for each unit was repeatedly done from 200 to 400 times during each measurement in a one-day observation period. Throughout the observation period, we let the vessel moved to a certain distance adrift with the wind and current for a several hours while sending and receiving acoustic signals. Each drift made an observation line and was repeated which covered suitable area. In parallel with the acoustic ranging measurement, we also measured the acoustic velocity of ocean water using a CTD profiler. It showed that the velocity averaged over the ray path sometimes varies up to 3m/s (0.2%) during one-day measurement. According to a theoretical research, the positioning error might be up to 1 m if the position of the units were determined using homogeneous acoustic velocity. However, it should be noted that the actual velocity varies with time. Thus, analysis must be done considering the temporal variation in velocity structure. We first tried to determine the position of the units and velocity by grid-search method on assumption that velocity should be homogeneous in time and space. Residuals of the travel time were up to 1 ms. At that time, the actual velocity variation measured with CTD profiler. The measured velocity gave a good explanation for the residual of travel time. Furthermore, we tried to determine the benchmark position assuming temporal variations in the velocity structure, in which velocity varied with time but was constant for respective drift lines. Residual of the travel time reduced to about 0.1 ms. In this case the positions and the velocity were estimated for each unit independently. The temporal variation for estimated velocities were similar in pattern for all tree units in each site although they were different in absolute value of up to 2m/s (0.13%). The temporal variation patterns of velocity were similar even though the geometrical relationship between the wake of the vessel and the locations of the units were different. This may suggests plausibility of the model of the temporal variation in acoustic velocity. In contrast, the difference in absolute velocities between the units suggests that a trade-off between acoustic velocities and distance has not been solved yet. The measurements were done four times for northwestern site and three times for northeastern site from 2002 to 2003. Based on these observations, we estimated the location of the weight centers of the three units in the each site through the above method. As a result, repeatability of positioning was about 50 cm for both horizontal and vertical directions for both sites. Lastly, we will also attempt to address and discuss the result of analysis in constraining three units simultaneously using a varying velocity.
G21A-0146 0800h
Effects of Temporal Variations in Sound Speed Structure at Suruga Bay, Central Japan, on the Observations of Seafloor Crustal Deformation
Observations of seafloor crustal deformation is very important to understand the dynamics of plate boundary that include the strain accumulation processes, great interplate earthquakes mechanisms, and submarine volcanoes activities. Since most of the plate boundaries and seismogenic zones are located under the sea, we have been developing an observation system for monitoring of seafloor crustal deformation. This system consists of the following two main components: (1) kinematic GPS positioning of an observation vessel and (2) accurate acoustic measurements of distance between a transducer equipped to the vessel and a transponder installed on the seafloor. We have repeatedly tested for the repeatability of transponder positioning for two months in 2002 at Suruga Bay, central Japan. Suruga Bay is situated in the northeast part of Nankai Trough, a place where the subduction plate boundary between the Philippine Sea and Eurasian plates. The horizontal and vertical errors in the repeated measurements (bias) are 20 cm and 44 cm, respectively [Sugimoto et al., 2003 AGU Fall Meeting]. We started the repeated observation from October 2002 in Suruga Bay and five-times measurements have been performed until May 2004. During the above measurements, temporal and spatial variations of sound speed structure in seawater are the error sources. Using the CTD (conductivity, temperature and depth) profiler, we repeatedly measured sound speed profiles during each acoustic ranging observation. In this study, we will show that the residuals of the observed acoustic travel-times can be explained by the temporal variations of sound speed structure. To achieve our purpose, using actual undersea velocity profiles measured with CTD profile and observed acoustic travel-time data, we have been trying to calculate acoustic travel-times after estimation of the temporal model of horizontal multi-layer velocity structure, and evaluate the effects of temporal variation of sound speed structure for seafloor positioning. The results of the present evaluation study may provide important information on the seafloor positioning including the temporal variations of sound speed.
G21A-0147 0800h
Spatial variations in Acoustic velocity at Kuroshio region for the accurate ocean-bottom positioning
We have developed a system for observing seafloor crustal deformation. This system combines GPS and acoustic techniques. One of the main errors in this system is heterogeneity of acoustic velocity in sea water. In this talk we discuss effects of lateral heterogeneity of the velocity in sea water on the location of a seafloor station. We evaluated lateral heterogeneity of the velocity in sea water for accurate determination of the ocean-bottom position. If we carry out CTD (Conductivity Temperature Depth profiler) measurements with a single ship, we cannot distinguish spatial and temporal variations from the measurement results. We therefore made simultaneous CTD measurements with two ships at the Kuroshio region (Kumano Basin). The distances between two ships are 2, 1, 0.6, 0.3, and 0.15 nautical miles (nm). The two ships were aligned parallel or orthogonal to the direction of ocean current. The CTD profilers used are SBE-19, SBE-25, and SBE-911plus of Sea-Bird Co., Ltd. The spatial variation is large at depths above 600 meters. It increases with the ship separation; and it is negligible at 0.15 nm separation. The spatial variation is up to 2.3 m/s when the two ships were aligned parallel to the direction of ocean current, and up to 6.6 m/s for the orthogonal direction at outside of Kuroshio region. At Kuroshio region, the variation is up to 2.8 m/s for parallel direction, and up to 7.0 m/s for the orthogonal direction. In addition, the spatial variation is long-period variation at Kuroshio region. We estimated the location error in ocean-bottom positioning caused by the spatial variation of the velocity structure using the CTD data. The data used in this calculation are five simultaneous CTD measurements inside of the Kuroshio region and eight measurements at outside of Kuroshio region. We modeled a 3-D velocity structure from the above dataset. Then we obtained the location of transmission from the travel time data. We searched the location of seafloor station which explains the travel time data assuming a homogeneous velocity structure. The difference is defined as lateral error between the true location and the estimated location. Results as follows: (1) Inside of the Kuroshio region The lateral error was negligible for parallel direction, and up to 50 cm for the orthogonal direction. (2) Outside of the Kuroshio region The lateral error was 20 cm for the parallel and orthogonal directions. On the basis of the results, we evaluate the effects of the spatial variations on the location error of a sea-floor position.