Continuous GPS Research Articles

Tectonic and non-tectonic crustal deformation in Kumaun Garhwal Himalaya

A detailed characterization of the crustal deformation in the Himalaya is important to understand the tectonic and non-tectonic deformation processes responsible for the strain build-up and transients in deformation. In the present work, we have characterised the present day tectonic and non-tectonic deformations in the Kumaun Garhwal Himalaya estimated from the continuous GPS measurements. We observed strong seasonal variations in the horizontal and vertical displacement components (i.e. North, East and Vertical). The transient surface deformation estimated from global hydrological models is found to explain ~40–60% of the observed seasonal variations at sites. Plate convergence rate, and variation of slip deficit on the Main Himalayan Thrust (MHT) are calculated using site velocity estimated after removal of seasonal variations from the GPS time series. We report an excessive compression with an average principal shortening rate of −0.09 μ strain/yr in the north-northeast direction. The derived plate convergence rate is 18 ± 1 mm/yr and the width of locked frontal part of MHT is 100 ± 15 km. The quiescence of major to great earthquakes in the region since at least past 200 years and high rate of strain build up make this region vulnerable for a potentially devastating earthquake of Mw ≥ 8 in near future.

Effect of antenna snow intrusion on vertical GPS position time series in Antarctica

Vertical surface displacements from continuous Global Positioning System (GPS) stations often show strong seasonal signals, which in some cases may be associated with surface mass loading, including hydrological, and non-tidal oceanic and atmospheric loading. In Antarctica, many GPS stations show vertical motions in phase with seasonal snow accumulation changes, but these variations cannot be fully explained with snow load variations between seasons. Instead we show, for many sites in Antarctica, that a significant component of the annual cycle in vertical GPS coordinates time series may be related to snow/ice inside antennas causing as an artefact apparent seasonal variation, with amplitudes of up to 4 mm. We present a method based on the Empirical Mode Decomposition (EMD) algorithm to remove this artefact signal. The corrected GPS time series show an improvement in the agreement with displacements predicted by elastic modelling using GRACE-derived surface mass loads.

Present‐Day Crustal Vertical Velocity Field for the Contiguous United States

AbstractThe study of vertical crustal motion in the contiguous United States (CONUS) has traditionally focused on the high‐amplitude deformation caused by glacial isostatic adjustment. To better understand more subtle vertical crustal motion resulting from other geophysical processes, we take advantage of the ongoing expansion of continuous Global Positioning System (GPS) networks, whose geodetic observations provide ever‐increasing accuracy and spatial resolution of surface deformation. Using position data for 2,782 GPS stations operating between 2007 and 2017, we produce a new vertical crustal velocity field for the CONUS region. We estimate our own station velocities to ensure consistent treatment of time series outliers, noise, and offsets, and we use adaptive smoothing and interpolation to account for spatially varying station density. Our velocity field reveals spatially coherent vertical features that are representative of regional tectonics, hydrologic, and anthropogenic processes. By removing the effects of modeled glacial isostatic adjustment and hydrologic loading estimated from Gravity Recovery and Climate Experiment (GRACE) data, we reduce the variance in our velocity field by36%and show residuals potentially due to geocenter motion and underlying tectonics.

Impacts of Local Effects and Surface Loads on the Common Mode Error Filtering in Continuous GPS Measurements in the Northwest of Yunnan Province, China.

While seasonal hydrological mass loading, derived from Gravity Recovery and Climate Experiment (GRACE) measurements, shows coherent spatial patterns and is an important source for the common mode error (CME) in continuous global positioning system (cGPS) measurements in Yunnan, it is a challenge to quantify local effects and detailed changes in daily GPS measurements by using GRACE data due to its low time and spatial resolutions. In this study, we computed and compared two groups of CMEs for nine cGPS sites in the northwest Yunnan province; rCMEs were computed with the residual cGPS time series having high inter-station correlations, while oCMEs were computed with all the GPS time series. The rCMEs-filtered time series had smaller variances and larger root mean square (RMS) reductions than those that were oCMEs-filtered, and when the stations local effects were not removed, spurious transient-like signals occurred. Compared with hydrological mass loading (HYDL), its combination with non-tidal atmosphere pressure and ocean mass reached a better agreement with the CME in the vertical component, with the Nash–Sutcliffe efficiency (NSE) increasing from 0.28 to 0.55 and the RMS reduction increasing from 15.19% to 33.4%, respectively. Our results suggest that it is necessary to evaluate the inter-station correlation and remove the possible noisy stations before conducting CME filtering, and that one should carefully choose surface loading models to correct the raw cGPS time series if CME filtering is not conducted.

Contemporary movement of the Earth's crust in the Northwestern Vietnam by continuous GPS data

The paper presents an estimation of the Earth’s crustal motion from the continuous GPS data at 6 stations (MTEV, MLAY, DBIV, TGIV, SMAV and SLAV) in the Northwestern and at PHUT (Hanoi) station using GAMIT/GLOBK software. The absolute displacements of the Earth’s crust at 7 stations in the IGS14 frame are respectively: 34.10±0.71 mm/yr (DBIV), 34.31±0.65 mm/yr (PHUT), 34.51±0.75 mm/yr (SMAV), 34.55±0.80 mm/yr (MLAY), 34.80±0.72 mm/yr (TGIV), 34.93±0.99 mm/yr (SLAV) and 35.59±0.73 mm/yr (MTEV), in the southeastward with the azimuth range 104-108o. The Son La fault is a right-lateral slip fault with a shear amplitude of ~1.5 mm/yr. The Lai Chau-Dien Bien fault is a left-lateral slip fault with a shear amplitude of ~1.9 mm/yr. Although the absolute velocities at the DBIV, SMAV, SLAV, TGIV and MLAY stations are evaluated with the error <1 mm/yr, the relative displacement on the Ma River fault is of ~0.5 mm/yr, and it seems that we still do not have a reliable assessment of the slip rate on the Ma River right-lateral slip fault.

Active deformation in the San Salvador extensional stepover, El Salvador from an analysis of the April–May 2017 earthquake sequence and GPS data

Modern and historic seismic observations indicate that earthquake swarms, often with similar characteristics, have occurred within a few tens of kilometers of the San Salvador volcano, which poses a major active hazard to the San Salvador metropolitan area. Here, we evaluate whether the April–May 2017 earthquake sequence near the San Salvador volcanic complex was tectonic or volcanic and more broadly consider the implications of up-to-date GPS observations and seismic data for earthquake and volcano hazards in the San Salvador metropolitan area. Based on magnitudes calculated for 532 earthquakes in the April–May sequence, we report a Gutenberg-Richter law b-value of 0.95 ± 0.12, consistent with b values for tectonic earthquakes, and based on temporal distribution of aftershocks we report an Omori law p-value of 1.5 ± 0.1. Focal mechanisms estimated for the mainshock and 14 largest foreshocks and aftershocks are all strike-slip with NNE-striking, left-lateral-slip and ESE-striking right-lateral-slip nodal planes. The best located foreshocks and aftershocks unequivocally indicate that the earthquake sequence accommodated left-lateral slip along a NNE-striking fault, compatible with the local tectonic setting. An absence of seismic and GPS evidence for any unrest of the San Salvador volcano during the years after the 2017 earthquake sequence is consistent with its tectonic origin. Continuous GPS measurements at a site 5 km south of the volcano reveal a previously unknown inflationary episode from mid-2010 to mid-2012, but no evidence for volcanic unrest in 2017 or later. Updated GPS velocities indicate that ~4–5 mm yr−1 of E-W-dominated stretching occurs between the San Salvador volcano and Ilopango caldera. That the 2017 earthquakes and Mw = 5.7 1986 San Salvador earthquake both accommodated left-lateral strike-slip along NNE-striking faults suggests that the GPS-measured E-W stretching is accommodated by bookshelf faulting between blocks that rotate clockwise in response to 10 ± 1.4 mm yr−1 (95%) of dextral motion across the volcanic arc at the location of San Salvador.

Resolving the Kinematics and Moment Release of Early Afterslip Within the First Hours Following the 2016 Mw 7.1 Kumamoto Earthquake: Implications for the Shallow Slip Deficit and Frictional Behavior of Aseismic Creep

AbstractAs stresses following rupture are dissipated continuous measurements of postseismic surface deformation provide insight into variations of the frictional strength of faults and the rheology of the lower crust and upper mantle. Due to the difficulty of capturing the earliest phase of afterslip, most analyses have focused on understanding postseismic processes over timescales of weeks to years. Here we investigate the kinematics, moment release, and frictional properties of the earliest phase of afterslip within the first hours following the 2016 Mw 7.1 Kumamoto earthquake using a network of 5‐minute sampled continuous Global Positioning System (GPS) stations. Using independent component analysis to filter the GPS data, we find that (1) early afterslip contributes only ~1% of total moment release within the first hour and 8% after 24 hr. This suggests that the lack of a coseismic slip deficit, which we estimate using standard geodetic data sets (e.g., InSAR, GPS, and pixel offsets) and which span the first 4 days of the postseismic period, is largely reflective of the dynamic rupture process and we can rule out contamination of moment release by early afterslip. (2) Early afterslip shows no evidence of a delayed nucleation or acceleration phase, where instead fault patches transition to immediate deceleration following rupture that is consistent with frictional relaxation under steady state conditions with dependence only on the sliding velocity. (3) There is a close correlation between the near‐field aftershocks and afterslip within the first hours following rupture, suggesting afterslip may still be an important possible triggering mechanism during the earliest postseismic period.

Postseismic Deformation Following the 2015 Mw7.8 Gorkha (Nepal) Earthquake: New GPS Data, Kinematic and Dynamic Models, and the Roles of Afterslip and Viscoelastic Relaxation

AbstractWe report Global Positioning System (GPS) measurements of postseismic deformation following the 2015 Mw7.8 Gorkha (Nepal) earthquake, including previously unpublished data from 13 continuous GPS stations installed in southern Tibet shortly after the earthquake. We use variational Bayesian Independent Component Analysis (vbICA) to extract the signal of postseismic deformation from the GPS time series, revealing a broad displacement field extending >150 km northward from the rupture. Kinematic inversions and dynamic forward models show that these displacements could have been produced solely by afterslip on the Main Himalayan Thrust (MHT) but would require a broad distribution of afterslip extending similarly far north. This would require the constitutive parameter (a − b)σ to decrease northward on the MHT to ≤0.05 MPa (an extreme sensitivity of creep rate to stress change) and seems unlikely in light of the low interseismic coupling and high midcrustal temperatures beneath southern Tibet. We conclude that the northward reach of postseismic deformation more likely results from distributed viscoelastic relaxation, possibly in a midcrustal shear zone extending northward from the seismogenic MHT. Assuming a shear zone 5–20 km thick, we estimate an effective shear‐zone viscosity of ~3·1016–3·1017 Pa·s over the first 1.12 postseismic years. Near‐field deformation can be more plausibly explained by afterslip itself and implies (a − b)σ ~ 0.5–1 MPa, consistent with other afterslip studies. This near‐field afterslip by itself would have re‐increased the Coulomb stress by ≥0.05 MPa over >30% of the Gorkha rupture zone in the first postseismic year, and deformation further north would have compounded this reloading.

Comparative analysis of strain parameter acquisition based on Cave strain and GPS observation

A set of three-component extensometers are installed on the Xibozi Seismic Station in Yanqing District, Beijing. The layout of the extensometers is an isosceles triangle, and the observation is the amount of expansion and contraction of each side, which is then converted into the strain value of each side. In this paper, the regional crustal strain parameters were calculated based on the observations of the extensometers and the continuous GPS observations around Xibozi Sesimic Staion. From two aspects of observation at the same site and different scales, the ground strain information reflected by the two observation methods is compared and analyzed. From the strain parameters calculated by a single point (extensometers) and the strain parameters calculated by the triangular surface observation (GPS) with different specifications, the overall Beijing area is in the context of tensile strain. However, there is a gap of several orders of magnitude between the two results, which may be related to the artificial operation of the pre-processes such as zero drift and step removal of the instrument’s strain observation data.

Automatic Detection of Volcanic Surface Deformation Using Deep Learning

AbstractInterferometric Synthetic Aperture Radar (InSAR) provides subcentimetric measurements of surface displacements, which are key for characterizing and monitoring magmatic processes in volcanic regions. The abundant measurements of surface displacements in multitemporal InSAR data routinely acquired by SAR satellites can facilitate near real‐time volcano monitoring on a global basis. However, the presence of atmospheric signals in interferograms complicates the interpretation of those InSAR measurements, which can even lead to a misinterpretation of InSAR signals and volcanic unrest. Given the vast quantities of SAR data available, an automatic InSAR data processing and denoising approach is required to separate volcanic signals that are cause of concern from atmospheric signals and noise. In this study, we employ a deep learning strategy that directly removes atmospheric and other noise signals from time‐consecutive unwrapped surface displacements obtained through an InSAR time series approach using an end‐to‐end convolutional neural network (CNN) with an encoder‐decoder architecture, modified U‐net. The CNN is trained with simulated synthetic unwrapped surface displacement maps and is then applied to real InSAR data. Our proposed architecture is capable of detecting dynamic spatio‐temporal patterns of volcanic surface displacements. We find that an ensemble‐average strategy is recommended to stabilize detected results for varying deformation rates and signal‐to‐noise ratios (SNRs). A case study is also presented where this method is applied to InSAR data covering Masaya volcano, Nicaragua and the results are validated using continuous GPS data. The results confirm that our network can indeed efficiently suppress atmospheric and other noise to reveal the noise‐free surface deformation.

The contemporary strain rate field in Uruguay and surrounding region and possible implications for seismic hazard

We utilize data from the SIRGAS continuous GPS network in and around Uruguay from 2006 to 2018 to model the surface strain rate regime over this period. During this time, we detected the effects from two subduction earthquakes that took place in Chile (the 27 Feb 2010 MW 8.8 Maule and the 16 Sep 2015 MW 8.3 Illapel earthquakes). Both affected surface velocities over a wide area and consequently the surface strain-rate regime. We modeled the strain rate regimes in three periods (pre-Maule, post-Maule and post-Illapel) and confirmed that there were significant changes over the region, in both sense – compression v. extension - and azimuth. We also examined the limited history and distribution of seimic events in this apparently stable continental region and found some indication that their timing may be have been affected by Nazca plate subduction earthquakes over 1500 km away.

Flank sliding: A valve and a sentinel for paroxysmal eruptions and magma ascent at Mount Etna, Italy

AbstractRising magma, dike intrusions, and flank collapse are observed at many volcanoes worldwide, but how they interact is still poorly documented. Extensive synthetic aperture radar interferometry and continuous global positioning system observations captured a sharp dike intrusion at Mount Etna, Italy, during the 2018 paroxysm that triggered a vigorous seaward sliding of the eastern flank connected with brittle failure and deep magmatic resourcing. We propose a feedback process between flank acceleration and magma intrusion that derives from the interaction between the long- and short-term deformation of the volcano. The flank sliding acts as a valve that modulates the emplacement and eruption of magma within the shallow system. Rapid flank acceleration could potentially evolve into sudden collapses and seismic release at shallow depth. In turn, flank slip events could act as a sentinel for changes in magma depth and paroxysmal eruptions at Mount Etna.

Spatiotemporal variation of vertical displacement driven by seasonal hydrological water storage changes in Kalimantan, Indonesia from GPS observation

The Global Positioning System (GPS) Indonesian Continuously Operating Reference Stations (InaCORS) network has been used for various applications ranging from land administration to geodynamics. Crustal dynamic process plays a significant role in almost all Indonesia regions except in Kalimantan Island since it's located on the stable continental plate. Therefore, the study of deformation in an area with less tectonic activities would be relatively simple to elucidate another terrestrial process. Here, we analyzed the deformation process in Kalimantan Island using continuous geodetic observation. We found that vertical displacement shows oscillation signals which are related to the hydrological cycle of water storage changes. Along with the GPS data, we utilized rain precipitation data to analyze both spatial and temporal variation of vertical displacement to support the hypothesis of elastic response to the load of water mass variations. Finally, we emphasized that continuous GPS observation in Kalimantan is reliable for detecting seasonal hydrological water storage changes.

Seismicity in Strike‐Slip Foreland Faults (Central Betic Cordillera Front): Evidence of Indentation Tectonics

AbstractUnexpected seismicity has been detected in the past few years along the northern Guadiana Menor River (Guadalquivir foreland basin) at the northern boundary of the Betic Cordillera. Earthquake focal mechanisms evidence the activity of N‐S to NNE‐SSW sinistral faults in the basement. Yet continuous GPS (CGPS) data show a westward movement in both the Prebetic Arc and the eastern Guadalquivir basin infill, which disagrees with the strike‐slip faults. To more precisely describe the structure and evolution of the area, new Bouguer anomaly data from the southeastern basin combined with seismicity data, electric resistivity tomography profiles (ERT), and surface studies are provided. Given the soft consistency of the sedimentary infill in the area, surface evidence of faulting is scarce, limited to elongated channel and minor vertical faults affecting Quaternary sediments, where ERT profiles suggested the presence of faults. These results suggest that the Guadalquivir basin infill, the Prebetic Arc and the central Betic Cordillera move towards the west, independent from the Iberian crust, but indicating to some extent an upward propagation of the basement deformation. However, the basement is coupled with the Alboran Domain and undergoes NNW‐SSE Eurasia‐Nubia convergence and orthogonal extension. This scenario is suggestive of the initial stages of indentation tectonics, better developed in the eastern Betic Cordillera and the central Alboran Sea.

Analisis Pengaruh Gempa Tektonik 10 Juli 2013 Terhadap Perubahan Koordinat Stasiun Pantau Segmen Mentawai

Mentawai Segment is located in Mentawai Islands, Sumatra, Indonesia. This segment is a subduction zone between Indo-Australian plate and Eurasian plate. This subduction zone can lead to high potential of tectonic earthquake in Mentawai Segment. The high potential of tectonic earthquake has negative impact for the community, so it is necessary to monitor the risk of tectonic earthquake in Mentawai Segment. This monitoring can be done by using GPS data of monitoring station that spread in Mentawai Segment. Therefore, this research aims to analyze the effect of tectonic earthquake on the coordinate change of Mentawai Segment, so that it can reduce the risk of negative impact of tectonic earthquake in Mentawai Segment. This research use observation data of 10 continuous GPS monitoring station (Sumatran GPS Data Array / SuGAr) in Mentawai Segment. Day of observation data was day of year (doy) at the time of tectonic earthquake occurence on July 10, 2013. Data processing used GAMIT / GLOBK software. The results of this research indicate that the tectonic earthquake (July 10, 2013) affected coordinates changes of the SuGAr station significantly two hours after the tectonic earthquake occurred.

Kinematics of Fault Slip Associated with the 4–6 July 2019 Ridgecrest, California, Earthquake Sequence

ABSTRACT The 2019 Ridgecrest, California, earthquake sequence produced observable crustal deformation over much of central and southern California, as well as surface rupture over several tens of kilometers. To obtain a detailed picture of the fault slip involved in the 4 July M 6.4 foreshock and 6 July M 7.1 mainshock, we combine strong-motion seismic waveforms with crustal deformation observations to obtain kinematic and static slip models of both events. We sample the regional seismic wavefield for both the foreshock and mainshock with three-component records from 31 stations of the California Integrated Seismic Network. The deformation observations include Global Positioning System (GPS), Interferometric Synthetic Aperture Radar (InSAR), and borehole strainmeter recordings of the dynamic strain field. These data collectively constrain the kinematic coseismic slip distributions of the events, with measurements variously observing coseismic slip from one event (e.g., seismic waveforms, kinematic solutions from continuous GPS, and strainmeter time series) or coseismic slip from both events combined (InSAR). We find that the foreshock ruptured two separate faults, one with left-lateral strike slip on a northeast–southwest-trending fault and the other with right-lateral strike slip on an orthogonal fault, with unilateral rupture propagation along both. The mainshock ruptured a series of northwest–southeast-trending faults with right-lateral strike slip concentrated in the uppermost 6 km with exceptionally low-rupture velocity averaging 1.0–1.5 km/s. A possible explanation for the low-rupture velocity is that the mainshock rupture expended relatively high energy, generating secondary fractures in off-fault deformation, which is consistent with field and seismic evidence of plastic deformation on small fault strands adjacent to the main rupture trace.

Impact of power-law rheology on the viscoelastic relaxation pattern and afterslip distribution following the 2010 Mw 8.8 Maule earthquake

After large earthquakes at subduction zones, the plate interface continues moving due to mostly frictional afterslip or simply afterslip processes. Below approximately 60 km depth, the seismic moment release at the plate interface is quite small indicating that the shear strength is low and stable sliding is the prevailing process. This agrees with the lack of significant interseismic locking at deeper segments (>60 km) resulting from the inversion of geodetic data and thus low afterslip can be expected. However, inversion models that employ linear viscoelastic mantle rheology and an elastic crust result in significant afterslip at depths >60 km. In this paper, we present a combination of a 3D forward geomechanical model with power-law rheology that simulates postseismic relaxation with dislocation creep processes in the crust and upper mantle and an afterslip inversion. We estimate the cumulative viscoelastic relaxation and the afterslip distribution for the first six years following the 2010 Mw 8.8 Maule earthquake in Chile. The cumulative afterslip distribution is obtained from the inversion of the residual surface displacements between the observed displacements from the continuous GPS (cGPS) and the ones from the forward modelling. We investigate five simulations, four with different dislocation creep parameters for the crust, slab, and upper mantle and one with elastic properties for the crust and slab, and a linear viscoelastic upper mantle for comparison. Our preferred simulation considers a weak crust since it shows the best fit to the cumulative cGPS postseismic displacements, a good fit to the time-series, and, in particular, a good spatial correlation between afterslip and aftershock activity. In this simulation, most of the viscoelastic relaxation occurs in the continental lower crust beneath the volcanic arc due to dislocation creep processes. The resulting afterslip pattern from the inversion is reduced at depths >60 km, which correlates to the low cumulative seismic moment that is released from aftershocks at these depths. Furthermore, the cumulative afterslip moment release from this simulation corresponds to 10% of the main shock in six years, which is approximately half of the moment release that results from models with an elastic crust and linear viscosity in the upper mantle. We conclude that an integrated analysis by considering power-rheology with dislocation creep processes in the continental crust and upper mantle along with aftershock activity may be used to constrain location and magnitude postseismic relaxation processes better.

Reduced ice mass loss and three-dimensional viscoelastic deformation in northern Antarctic Peninsula inferred from GPS

SUMMARY We consider the viscoelastic rheology of the solid Earth under the Antarctic Peninsula due to ice mass loss that commenced after the breakup of the Larsen-B ice shelf. We extend the previous analysis of nearby continuous GPS time-series to include five additional years and the additional consideration of the horizontal components of deformation. They show strong uplift from ∼2002 to 2011 followed by reduced uplift rates to 2018. Modelling the GPS-derived uplift as a viscoelastic response to ongoing regional ice unloading from a new ice model confirms earlier estimates of low upper-mantle viscosities of ∼0.3–3 × 1018 Pa s in this region but allows a wide range of elastic lithosphere thickness. The observed and modelled north coordinate component shows little nonlinear variation due to the location of ice mass change to the east of the GPS sites. However, comparison of the observed and modelled east coordinate component constrains the upper-mantle viscosity to be less than ∼9 × 1018 Pa s, consistent with the viscosity range suggested by the uplift rates alone and providing important, largely independent, confirmation of that result. Our horizontal analysis showed only marginal sensitivity to modelled lithospheric thickness. The results for the horizontal components are sensitive to the adopted plate rotation model, with the estimate based on ITRF2014 suggesting that the sum of residual plate motion and pre-2002 glacial isostatic adjustment is likely less than ∼±0.5 mm yr−1 in the east component.

Patterns of the Seismic Cycle in the Kuril Island Arc from GPS Observations

The 2006–2007 Simushir earthquakes started a new episode in the rupture history of the Kuril-Kamchatka subduction zone. This earthquake doublet terminated the seismic silence in the Central Kurils, which had lasted for almost a century. The analysis of seismologic, geologic and GPS data over the Kuril Islands region allows us to reveal the patterns of deformation of the lithosphere in the Kuril subduction zone, in particular, the peculiarities of accumulation and relaxation of elastic stresses during the seismic cycle. We analyzed > 9 years of continuous GPS observations on the Kuril Islands to determine the surface displacements caused by the geodynamic processes in the subduction zone. We used available GPS data to model the coseismic stress release and postseismic stress relaxation related to the 2006–2007 Simushir earthquakes. We showed that the duration of postseismic relaxation after the 2006 earthquake exceeds 10 years, preventing transition of the central Kurils to the interseismic stage and thus delaying the beginning of the new seismic cycle. The value of Maxwell viscosity of the asthenosphere found in this study (3 × 1017 Pa s) in the Kuril Island arc is substantially lower than the values determined for several other subduction zones. This contrast in viscosity agrees with Kogan et al. (J Geophys Res 118:3691–3706, 2013), who used shorter GPS time series than in our study. The results of our analysis show that multidirectional motions along the Kurils Islands are explained by the fact that segments of the island arc are at different stages of the seismic cycle.

Contemporary crustal deformation of Northeast Tibet from geodetic investigations and a comparison between the seismic and geodetic moment release rates

We processed the data recorded by 32 continuous GPS stations from the Crustal Movement Observation Network of China and combined them with published GPS observations in a common geodetic reference frame to better depict the ongoing crustal velocities and deformation field of Northeast Tibet. The derived velocity field indicates that the deformation of Northeast Tibet is characterized mainly by crustal shortening coupled with the eastward extrusion of crustal material and clockwise block rotation. We further used the GPS velocity field to evaluate the strain rate tensors by the method of Sandwell and Wessel (2016). The strain rate field shows an obvious transition of the contraction rate from NE-SW to NEE-SWW, and a decrease in the maximum shear strain rate from west to east was observed along the East Kunlun fault and Haiyuan fault, indicating that the block rotation and lateral extrusion therein may be partitioned by NE-SW-oriented compression to some extent. The inhomogeneity of the rotation strain rate implies that Northeast Tibet does not behave like a single rigid block but instead resembles a left-lateral shear zone, supporting the hypothesis that the northward transfer of motion along the East Kunlun fault may play a significant role in the deformation style of this area. The dilatation rate verified that most of Northeast Tibet experiences crustal thickening, which coincides with the average uplift rate of ~1 mm/y revealed by GPS velocities and channel steepness indices. We also estimated the geodetic moment rates and compared them with the seismic moment rates estimated using the truncated Gutenberg-Richter earthquake distribution. The ratio of the seismic strain rate to the geodetic strain rate in the regions near the Altyn Tagh fault, the Qaidam Basin adjoining the Qilian Mountains, the Haiyuan fault and the central part of the East Kunlun fault are near unity, indicating that the deformation in these areas is dominated by seismic release and may represent fully coupled seismogenic zones. The occurrence of large earthquakes in the past (estimated magnitudes M ≥ 6) coupled with a high geodetic deformation rate in these regions could indicate overdue M ≥ 6 earthquakes. However, other factors, including the occurrence of creep and the long recurrence time of Mw ≥ 8.0 earthquakes, should be taken into consideration along the Haiyuan fault. By contrast, in other regions, including the southern and eastern parts of the Haiyuan fault, along the Elashan fault, and the northern Qilian Mountains, the geodetic moment rate is higher than the seismic moment rate by a factor between 3 and 9. We propose that this phenomenon is related to the undersampling of long-term earthquake rates within the seismic catalogue, long-term aseismic crustal deformation, or a combination of these factors. Our study has implications for better comprehending the current seismotectonic pattern of Northeast Tibet.