Abstract.Measurements of surface displacements using

Two years after the Great Sumatra‐Andaman earthquake the 3.1 m WSW coseismic displacement at Port Blair, Andaman Islands, had increased by 32 cm. Postseismic uplift initially exceeded 1 cm per week and decreased to <1 mm/week. By 2007 points near Port Blair had risen more than 20 cm, a 24% reversal of coseismic subsidence. Uplift at eight GPS sites suggests a gradual eastward shift of the coseismic neutral axis separating subsidence from uplift. Simulations of the GPS postseismic displacements as viscoelastic relaxation of coseismic stress change and as slip on the plate interface indicate that slip down‐dip of the seismic rupture dominates near‐field deformation during the first two years. Postseismic slip beneath the Andaman Islands released moment equivalent to a magnitude Mw ≥ 7.5 earthquake, and the distribution suggests deep slip in the stable frictional regime accelerated to catch up to the coseismic rupture.

In the winter of 1811-1812, near the town of New Madrid in the central United States and more than 2,000 km from the nearest plate boundary, three earthquakes within three months shook the entire eastern half of the country and liquefied the ground over distances far greater than any historic earthquake in North America. The origin and modern significance of these earthquakes, however, is highly contentious. Geological evidence demonstrates that liquefaction due to strong ground shaking, similar in scale to that generated by the New Madrid earthquakes, has occurred at least three and possibly four times in the past 2,000 years (refs 4-6), consistent with recurrence statistics derived from regional seismicity. Here we show direct evidence for rapid strain rates in the area determined from a continuously operated global positioning system (GPS) network. Rates of strain are of the order of 10(-7) per year, comparable in magnitude to those across active plate boundaries, and are consistent with known active faults within the region. These results have significant implications for the definition of seismic hazard and for processes that drive intraplate seismicity.

Position time series from Global Positioning System (GPS) stations in the New Madrid region were differenced to determine the relative motions between stations. Uncertainties in rates were estimated using a three-component noise model consisting of white, flicker, and random walk noise, following the methodology of Langbein, 2004. Significant motions of 0:37 0:07 (one standard error) mm/yr were found between sites PTGV and STLE, for which the baseline crosses the inferred deep portion of the Reelfoot fault. Baselines between STLE and three other sites also show significant motion. Site MCTY (adjacent to STLE) also exhibits significant motion with respect to PTGV. These motions are consistent with a model of interseismic slip of about 4 mm=yr on the Reelfoot fault at depths between 12 and 20 km. If constant over time, this rate of slip produces sufficient slip for an M 7.3 earthquake on the shallow portion of the Reelfoot fault, using the geologically derived recurrence time of 500 years. This model assumes that the shallow portion of the fault has been previously loaded by the intraplate stress. A GPS site near Little Rock, Arkansas, shows significant southward motion of 0:3-0:4 mm=yr (0:08 mm=yr) relative to three sites to the north, indicating strain consistent with focal mechanisms of earthquake swarms in northern Arkansas.

More than six years after the great (M w 9.2) Sumatra-Andaman earthquake, postevent processes responsible for relaxation of the coseismic stress change remain controversial. Modeling of Andaman Islands Global Positioning System (GPS) displacements indicated early near-field motions were dominated by slip down-dip of the rupture, but various researchers ascribe elements of relaxation to dominantly poroelastic, dominantly viscoelastic and dominantly fault slip processes, depending primarily on their measurement sampling and modeling tools used. After subtracting a pre-2004 interseismic velocity, significant transient motion during the 2008.5-2010.5 epoch confirms that postseismic relaxation processes continue in Andaman. Modeling three-component velocities as viscoelastic flow yields a weighted root-mean-square (WRMS) misfit that always exceeds the WRMS 26.3 mm/yr of the measured signal. The bestfitting models are those that yield negligible deformation, indicating the model parameters have no real physical meaning. GPS velocities are well-fit (WRMS 4.0 mm/yr) by combining a viscoelastic flow model that best-fits the horizontal velocities with ~50 cm/yr thrust slip downdip of the coseismic rupture. Both deep slip and flow respond to stress changes, and each can significantly change stress in the realm of the other, so it is reasonable to expect that both transient deep slip and viscoelastic flow will influence surface deformation long after a great earthquake.

We use Global Positioning System (GPS) measurements acquired from 1991 to 1995 to constrain the motion of sites in Bangalore, in southern India, and Kathmandu, Nepal, relative to a global GPS network. These measurements permit estimates of the northward motion of the Indian plate and convergence between the southern Himalaya and the Indian subcontinent. The velocities of Bangalore and Kathmandu in the ITRF92 reference frame agrees with that predicted by the NNR‐NUVEL1A plate motion model for Indian plate motion, and differ from that predicted for the Australian plate, confirming the independent motion of the Indian and Australian plate fragments. No significant motion was detected between Bangalore and Kathmandu during the three years from 1991–1994, even though Kathmandu is located in the hanging wall of the active Himalayan thrust system. The Himalayan thrust system is thought to accommodate 18±7 mm/yr of convergence and has been the source of several historic M ∼ 8 earthquakes. The absence of motion of Kathmandu relative to the Indian plate can be explained if the thrust system is presently locked south of the Greater Himalaya. Our preferred model has no steady slip on the detachment south of the Greater Himalaya, and steady slip at a rate greater than 6 mm/yr (1/3 of the long‐term convergence rate) can be ruled out at 95% confidence level.

No abstract

We report here on the campaign GPS data from the Andaman Islands just previous to the great 2004 Sumatra-Andaman earthquake. The campaign-mode acquisitions at Port Blair showed that the site started to subside between 2003 and 2004. In addition, during this period, the horizontal displacement of Port Blair with respect to Indian plate, deduced from 1996-2000 GPS data, changed its orientation to that obtained during the 26 th Dec 2004 co-seismic. This short-term subsidence can be modeled as slip in the up-dip portion of the fault, a slip that is equivalent to an earthquake with moment magnitude of 6.3. Previously, slow slip was thought to appear at intermediate depths roughly 35-55 km but simple models of the deformation at this single site suggest slow slip at much shallower depth than this. This observation of subsidence obtained by GPS methods is in rough agreement with subsidence observed from tide gauge data. Campaignmode GPS data between 1996 and 2000 suggest uplift for Port Blair during the inter-seismic period and so does the reported field observations of interseismic micro-atoll emergence. Lack of exposed land with GPS stations along the southern part of the thrust fault deprive of arriving at any indication of this preseismic subsidence in those areas. Although GPS data is lacking the geological indices reported from some sites on the Alaskan Coast, for example, imply short-term subsidence just previous to the great 1964 earthquake. The pre-earthquake subsidence recorded in Port Blair, therefore, may have global implications as a precursor signal of great earthquakes at least along some of the subduction zones.