Geoffrey Blewitt

Crustal displacements due to continental water loading

The effects of long-wavelength (> 100 km), seasonal variability in continental water storage on vertical crustal motions are assessed. The modeled vertical displace- ments (ARM) have root-mean-square (RMS) values for 1994- 1998 as large as 8 mm, with ranges up to 30 mm, and are predominantly annual in character. Regional strains are on the order of 20 nanostrain for tilt and 5 nanostrain for hori- zontal deformation. We compare ArM with observed Global Positioning System (GPS) heights (Aro) (which include ad- justments to remove estimated effects of atmospheric pres- sure and annual tidal and non-tidal ocean loading) for 147 globally distributed sites. When the Aro time series are ad- justed by ArM, their variances are reduced, on average, by an amount equal to the variance of the ArM. Of the Aro time series exhibiting a strong annual signal, more than half

A geodetic plate motion and Global Strain Rate Model

We present a new global model of plate motions and strain rates in plate boundary zones constrained by horizontal geodetic velocities. This Global Strain Rate Model (GSRM v.2.1) is a vast improvement over its predecessor both in terms of amount of data input as in an increase in spatial model resolution by factor of ∼2.5 in areas with dense data coverage. We determined 6739 velocities from time series of (mostly) continuous GPS measurements; i.e., by far the largest global velocity solution to date. We transformed 15,772 velocities from 233 (mostly) published studies onto our core solution to obtain 22,511 velocities in the same reference frame. Care is taken to not use velocities from stations (or time periods) that are affected by transient phenomena; i.e., this data set consists of velocities best representing the interseismic plate velocity. About 14% of the Earth is allowed to deform in 145,086 deforming grid cells (0.25° longitude by 0.2° latitude in dimension). The remainder of the Earths surface is modeled as rigid spherical caps representing 50 tectonic plates. For 36 plates we present new GPS-derived angular velocities. For all the plates that can be compared with the most recent geologic plate motion model, we find that the difference in angular velocity is significant. The rigid-body rotations are used as boundary conditions in the strain rate calculations. The strain rate field is modeled using the Haines and Holt method, which uses splines to obtain an self-consistent interpolated velocity gradient tensor field, from which strain rates, vorticity rates, and expected velocities are derived. We also present expected faulting orientations in areas with significant vorticity, and update the no-net rotation reference frame associated with our global velocity gradient field. Finally, we present a global map of recurrence times for Mw=7.5 characteristic earthquakes.

Uplift and seismicity driven by groundwater depletion in central California

Groundwater use in California’s San Joaquin Valley exceeds replenishment of the aquifer, leading to substantial diminution of this resource and rapid subsidence of the valley floor. The volume of groundwater lost over the past century and a half also represents a substantial reduction in mass and a large-scale unburdening of the lithosphere, with significant but unexplored potential impacts on crustal deformation and seismicity. Here we use vertical global positioning system measurements to show that a broad zone of rock uplift of up to 1–3 mm per year surrounds the southern San Joaquin Valley. The observed uplift matches well with predicted flexure from a simple elastic model of current rates of water-storage loss, most of which is caused by groundwater depletion. The height of the adjacent central Coast Ranges and the Sierra Nevada is strongly seasonal and peaks during the dry late summer and autumn, out of phase with uplift of the valley floor during wetter months. Our results suggest that long-term and late-summer flexural uplift of the Coast Ranges reduce the effective normal stress resolved on the San Andreas Fault. This process brings the fault closer to failure, thereby providing a viable mechanism for observed seasonality in microseismicity at Parkfield and potentially affecting long-term seismicity rates for fault systems adjacent to the valley. We also infer that the observed contemporary uplift of the southern Sierra Nevada previously attributed to tectonic or mantle-derived forces is partly a consequence of human-caused groundwater depletion.

Road Reduction Filtering for GPS‐GIS Navigation

A novel method of map matching using the Global Positioning System (GPS) has been developed for civilian use, which uses digital mapping data to infer the systematic position errors of less than 100m which result largely from ‘selective availability’ (S/A) imposed by the US military. Selective availability was switched off on the 2nd of May 2000, and is to be replaced with ‘regional denial capabilities in lieu of global degradation’. The system tracks a vehicle on all possible roads (road centre-lines) in a computed error region, then uses a method of rapidly detecting inappropriate road centre-lines from the set of all those possible. This is called the Road Reduction Filter (RRF) algorithm. Point positioning is computed using C/A code pseudorange measurements direct from a GPS receiver. The least squares estimation is performed in the software developed for the experiment described in this paper. Virtual differential GPS (VDGPS) corrections are computed and used from a vehicles previous positions, thus providing an autonomous alternative to DGPS for in-car navigation and fleet management. Height aiding is used to augment the solution and reduce the number of satellites required for a position solution. Ordnance Survey (OS) digital map data was used for the experiment, i.e. OSCAR 1 m resolution road centre-line geometry and Land Form PANORAMA 1:50,000, 50 m-grid digital terrain model (DTM). Testing of the algorithm is reported and results are analysed. Vehicle positions provided by RRF are compared with the ‘true’ position determined using high precision (cm) GPS carrier phase techniques. It is shown that height aiding using a DTM and the RRF significantly improve the accuracy of position provided by inexpensive single frequency GPS receivers.

Methodology for global geodetic time series estimation: A new tool for geodynamics

A method of automatically combining geodetic network solutions to produce station coordinate time series with realistic computed errors has been developed and tested and is being applied on a weekly basis to Global Positioning System (GPS) global and regional networks of the International GPS Service. Our techniques include modified Helmert blocking, stochastic modeling to minimize frame bias, Monte Carlo simulation, variance component estimation, and multiparameter data snooping. An 18-month time series evaluation of 150 globally distributed stations demonstrates that our combined weekly solution is more complete, precise, and reliable than any contributing solution. Our method of attaching regional networks without perturbing the global network solution, rather than combining normal equations, improves the quality measures. The median RMS of station position residuals with respect to a constant velocity model is 2.4 mm in latitude, 3.0 mm in longitude, and 7.2 mm in height. Our solution has since been incorporated into the reference frame ITRF96 (International Terrestrial Reference Frame 1996), showing a RMS coordinate difference of 5.4 mm, the lowest of all contributing solutions. As an independent test, the RMS difference with the ITRF94 is 4.5 mm in horizontal and 8.1 mm in height. As a second external test, the station velocity solution was used to estimate plate tectonic Euler vectors, which were then compared with the NUVEL-1A model and found to differ at a level consistent with the computed errors. Given a few more years of data, our error model predicts solutions that will be sufficiently precise to rigorously test NUVEL-1A or its successors.

On the determination of a global strain rate model

The objective of this paper is to outline the fundamental concepts underlying the estimation of a global strain rate model. We use a variant of the method first introduced by Haines and Holt (1993) to estimate the strain rate tensor field within all of the Earth’s deforming regions. Currently the observables used are ~1650 geodetic velocities, seismic moment tensors from the Harvard CMT catalog, and Quaternary fault slip rate data. A model strain rate field and velocity field are obtained in a least-squares fit to both the geodetic velocities and the observed strain rates inferred from fault slip rates. Seismic moment tensors are used to provide a priori constraints on the style and direction (not magnitude) of the model strain rate field for regions where no fault slip rate data are available. The model will soon be expanded to include spreading rates, ocean transform azimuths, and more fault slip rate data. We present a first estimate of the second invariant of the global model strain rate field. We also present Euler poles obtained by fitting geodetic vectors located on defined rigid plates. We find that 17% of the total model moment rate is accommodated in zones of (diffuse) continental deformation.

Degree‐2 harmonics of the Earth's mass load estimated from GPS and Earth rotation data

A fluid, mobile atmosphere and oceans surrounds the solid Earth and upon its land surface lays a continually changing distribution of ice, snow, and ground water. The changing distribution of mass associated with the motion of these surficial fluids changes the Earths rotation by changing its inertia tensor and changes the Earths shape by changing the load on the solid Earth. It has recently been demonstrated that large-scale changes of the Earths shape, and hence of the mass load causing the Earths shape to change, can be measured using the global network of GPS receivers. Here, the degree-2 mass load coefficients determined from GPS data are compared with those obtained from Earth orientation observations from which the effects of tides, winds, and currents have been removed. Good agreement is found between these two estimates of the degree-2 mass load, particularly at seasonal frequencies.

Contemporary uplift of the Sierra Nevada, western United States, from GPS and InSAR measurements

Modern space geodesy has recently enabled the direct observation of slow geological processes that move and shape Earth’s surface, including plate tectonics and crustal strain accumulation that leads to earthquakes. More elusive has been the direct observation of active mountain growth, because geodetic measurements have larger uncertainties in the vertical direction, while mountain growth is typically very slow. For the Sierra Nevada of California and Nevada, western United States, the history of elevation is complex, exhibiting features of both ancient (40–60 Ma) and relatively young (<3 Ma) elevation. Here we exploit the complementary strengths of high-precision three-component point positions from the GPS and blanket coverage line-of-sight measurements from interferometric synthetic aperture radar (InSAR) to show that contemporary vertical motion of the Sierra Nevada is between 1 and 2 mm/yr. The motion is upward with respect to Earth’s center of mass and with respect to a relatively stable eastern Nevada, indicating generation of relief and uplift against gravity. Uplift is distributed along the entire length of the range, between latitude 35°N and 40°N, and is not focused near localized, seismically imaged mantle downwellings. These results indicate that the modern episode of Sierra Nevada uplift is still active and could have generated the entire modern range in <3 m.y.

MIDAS robust trend estimator for accurate GPS station velocities without step detection

Abstract Automatic estimation of velocities from GPS coordinate time series is becoming required to cope with the exponentially increasing flood of available data, but problems detectable to the human eye are often overlooked. This motivates us to find an automatic and accurate estimator of trend that is resistant to common problems such as step discontinuities, outliers, seasonality, skewness, and heteroscedasticity. Developed here, Median Interannual Difference Adjusted for Skewness (MIDAS) is a variant of the Theil‐Sen median trend estimator, for which the ordinary version is the median of slopes vij = (xj–xi)/(tj–ti) computed between all data pairs i > j. For normally distributed data, Theil‐Sen and least squares trend estimates are statistically identical, but unlike least squares, Theil‐Sen is resistant to undetected data problems. To mitigate both seasonality and step discontinuities, MIDAS selects data pairs separated by 1 year. This condition is relaxed for time series with gaps so that all data are used. Slopes from data pairs spanning a step function produce one‐sided outliers that can bias the median. To reduce bias, MIDAS removes outliers and recomputes the median. MIDAS also computes a robust and realistic estimate of trend uncertainty. Statistical tests using GPS data in the rigid North American plate interior show ±0.23 mm/yr root‐mean‐square (RMS) accuracy in horizontal velocity. In blind tests using synthetic data, MIDAS velocities have an RMS accuracy of ±0.33 mm/yr horizontal, ±1.1 mm/yr up, with a 5th percentile range smaller than all 20 automatic estimators tested. Considering its general nature, MIDAS has the potential for broader application in the geosciences.

Global deformation from the great 2004 Sumatra-Andaman Earthquake observed by GPS: Implications for rupture process and global reference frame

Static coseismic offsets > 1 mm are observed up to 7800 km away from the great Sumatra-Andaman earthquake of 26 Dec. 2004 using global GPS network data. We investigate the rupture process based on far-field continuous GPS data. To reduce error in the coseismic offset estimates due to post-seismic deformation in the days following the main shock, we simultaneously fit a model of co- and postseismic offsets for nearby stations SAMP (500 km) and NTUS (900 km). The 3-month cumulative postseismic displacement for station SAMP amounts to 20% of the coseismic displacement, and can be well modeled by velocity-strengthening afterslip. We find that coseismic slip on the northern rupture segment is ∼3 m, which is consistent with seismic estimates. Our best estimate of the moment magnitude is Mw = 9.13 if we take into account the expected increase of the shear modulus with depth (for uniform μ = 30 GPa, the moment-magnitude would only be 8.97). Our geodetic results, and thus our inferred rupture model, are different from a similar study using far-field data of Banerjee et al. (2005). These differences highlight the challenge in earthquake studies on a global scale in terms of the sensitivity of far-field offset estimates to the analysis strategy and reference frame treatment. Our predicted coseismic offsets from this event are at least 1 mm across almost the entire globe. This warrants a reconsideration of how to maintain the global terrestrial reference frame affected by earthquakes of Mw > 9.0.