Improving Detection of Unsteady Crustal Deformation by using GNSS Data: Common Mode Error Reduction, Target Area Expansion to Hyuga-Nada, and Transient Signal Monitoring with Postseismic Deformation Removal

<p>The Lisbon Metropolitan Area, in the southern sector of the Lusitanian basin, SW Portugal, has been affected by relevant seismicity. Known destructive earthquakes affecting the region range in time from 1344 to 1969, and include catastrophic occurrences in 1356, 1531, 1909 and 1755. Modern instrumental data are available for the M7.8 Cadiz Gulf earthquake of 1969 only, which reached EMS-98 intensity 5 to 6 in the study area. While several of these earthquakes nucleated offshore, the 1909 earthquake, with estimated magnitude in the range M6.0-M6.5, had a clear intraplate nature, and it is widely accepted that the M7 1531 earthquake also nucleated onshore, in the active structures of the Lower Tagus Valley. The relative importance of the contributions of onshore versus offshore sources to seismic hazard in Portugal is largely debated. On one hand, in view of the modest NW Africa – SW Iberia convergence rate (~4 mm/yr in a NW-SE direction), it has been argued that most of the cumulated crustal deformation is fully released by 1969-type offshore earthquakes of the Gulf of Cadiz, implying that intraplate faults account for very small slip-rates. It follows that destructive intraplate earthquakes are deemed very rare events with limited contribution to the probabilistic hazard. This view is supported by very low intraplate slip-rate estimates of 0.005 to 0.3-0.5 mm/yr derived from geological studies. However, seismic hazard disaggregation studies indicate that the dominant scenario is the rupture of an intraplate fault.</p><p>Using a dense GNSS dataset coupled with PSInSAR analysis, we characterize the style of crustal deformation in the Lisbon Metropolitan Area and estimate the associated fault slip rates. We provide evidence of sinistral simple shear driven by a NNE-SSW first-order tectonic lineament. PSInSAR vertical velocities corroborate qualitatively the GNSS strain-rate field, showing uplift/subsidence where the GNSS data indicate contraction/extension. We propose the presence of a small block to the W of Lisbon moving independently towards the SW with a relative velocity of 0.96±0.20 mm/yr. Comparison between geodetic and seismic moment-rates indicates a high seismic coupling. We conclude that the contribution of intraplate faults to the seismic hazard in the Lisbon Metropolitan Area is more important than currently assumed.</p>

The motion of a continuously operating reference station is usually dominated by the long-term crustal motions of the tectonic block on which the station is located. Monitoring changes in the coordinates of reference stations located at tectonic plate boundaries allows for the calculation of velocity fields that reflect the spatial and temporal characteristics of the region. This study analyzes the spatiotemporal relationships of regional reference frame points with GNSS data from 25 reference stations in Sichuan, China, from 2015 to 2021. The common mode errors are extracted and eliminated by principal component analysis. A time series function model is developed for the reference stations and their constituent baselines for calculating the velocity field. Subsequently, the spatiotemporal characteristics of the regional reference frame in Sichuan is analyzed by a stochastic model. The results show that the influences of the common mode error on the horizontal and vertical directions of the reference stations is 2.5 mm and 4.3 mm, respectively. Generally, the horizontal motion of the reference stations in the Sichuan region tends to be in the southeast direction and the vertical motion trend is mainly uplifting. The east–west and vertical components of the baseline tend to be shortened, and the random influence among the reference stations is larger in the north–south and east–west directions—0.39 mm and 0.54 mm, respectively. Polynomial functions are more appropriate for constructing the fitted random influence covariance model.

The Noto Peninsula has experienced seismic swarms accompanied by transient crustal deformation since November 2020, followed by two major earthquakes (M6.5 on May 5, 2023, and M7.6 on Jan. 1, 2024). Previous studies have suggested that fluids are involved in a series of activities. Most evidence on fluids constrains only their existence, and quantitative information on dynamic fluid migration remains scarce. Past precise gravity measurements in volcanic areas captured changes at the μGal scale (10–8 m/s2) due to magma movement. Here, we report the gravity difference caused by the M6.5 earthquake that was obtained via a similar method of measurement. Most of the observed gravity change can be explained by a fault slip model determined from the geodetic inversion of GNSS data. However, a significant change of approximately 10 μGal remains unexplainable in the northern coastal area of the northeastern tip of the Noto Peninsula. To explain this change, we estimate environmental effects, such as groundwater and sea-level variations. These environmental effects are too small to fully explain the change unless large local groundwater changes that are not represented in the groundwater model are considered. Instead, adding a fluid-fed fault that opens above the coseismic fault could reasonably explain both the GNSS and gravity data. The inferred volume of fluids is approximately 10% of the volume to have accumulated in a deeper fault by June 2022, as estimated from GNSS data. This result suggests that fluids migrating to shallower areas may have increased the risk of the M7.6 earthquake. The relatively shallow seismic velocity anomalies inferred by seismic tomography might indicate that such an upward migration process due to large earthquakes has been repeated in the past.Graphical

Since the 2011 Tohoku megathrust earthquake had been occurred, there were strong demands to monitor and mitigate the geological hazards such as an earthquake and volcano in South Korea. This study focused on the crustal deformation, which indicates the earth's dynamics on or below the surface, for the long periods or by the transient events using the GNSS data and the Bernese software v5.2. The Korea peninsula has moved to the south-east direction with 3.3cm per a year during 2004–2015 when the GNSS data have been available. Also, it was testified that the transient displacement had been appeared on the Korea peninsula by the 2011 Tohoku earthquake. Therefore, we conclude that there needs to accumulate GNSS observations over the longer duration and to periodically analyze the crustal deformation as a phenomenon related to the seismic activity.

Spatial filtering is an effective way to improve the precision of coordinate time series for regional GPS networks by reducing so‐called common mode errors, thereby providing better resolution for detecting weak or transient deformation signals. The commonly used approach to regional filtering assumes that the common mode error is spatially uniform, which is a good approximation for networks of hundreds of kilometers extent, but breaks down as the spatial extent increases. A more rigorous approach should remove the assumption of spatially uniform distribution and let the data themselves reveal the spatial distribution of the common mode error. The principal component analysis (PCA) and the Karhunen‐Loeve expansion (KLE) both decompose network time series into a set of temporally varying modes and their spatial responses. Therefore they provide a mathematical framework to perform spatiotemporal filtering. We apply the combination of PCA and KLE to daily station coordinate time series of the Southern California Integrated GPS Network (SCIGN) for the period 2000 to 2004. We demonstrate that spatially and temporally correlated common mode errors are the dominant error source in daily GPS solutions. The spatial characteristics of the common mode errors are close to uniform for all east, north, and vertical components, which implies a very long wavelength source for the common mode errors, compared to the spatial extent of the GPS network in southern California. Furthermore, the common mode errors exhibit temporally nonrandom patterns.

The effective extraction and elimination of the Global Positioning System's (GPS) common mode error (CME) is of great significance to improving the signal-to-noise ratio of the coordinate time series and accurately estimating the characteristics of the crustal deformation. In this study, the principal component analysis (PCA), Karhunen–Loeve expansion (KLE) and stacking filtering methods are used to extract the CMEs of Greenland GPS station coordinates in the East-North-Up (ENU) directions. Their influence on GPS coordinate residuals and the relationship amongst the observations are also analyzed. The results show that the PCA, KLE and stacking filtering methods can effectively eliminate the common mode noise in the GPS station coordinate time series and reduce their uncertainty. The PCA method is better than KLE and stacking methods, and the filtering effect in the up (U) direction is better than those in the east (E) and north (N) directions. Then, as the CME can be detected periodically in the U direction, its physical mechanism is studied. On the one hand, the period of CME maximum power is 1 year but is not constant. On the other hand, the correlation between the GPS CME and GRACE ice loading deformation is 0.5. When the ice loading deformation is deduced by the GRACE method, the CME of the GPS network can be eliminated by 25.6%, which indicates that the ice loading deformation is part of the source of GPS CME in Greenland. Finally, the influence of the U direction CME on the station noise characteristics is analyzed using maximum likelihood estimation (MLE). The results show that the CME noise type is mainly WN + FN. Moreover, the estimated coordinate velocities and most of the annual amplitudes of the stations have an increasing trend after CME elimination, whereas the uncertainty has a decreasing trend. These findings are consistent with the trend of noise component change. The above results indicate that the CME elimination of the GPS stations cannot be ignored, as they can help to improve the estimation of station velocity and annual amplitude and reduce the uncertainty.

<p>Large earthquakes are followed by a post-seismic period during which the stresses induced by the co-seismic phase are relaxed through different processes. This post-seismic phase participates to the redistributions of stresses in the earth, and understanding its mechanism is a key to understand interactions between earthquakes at different spatial and temporal scales. In subduction zones the most important terms are the afterslip and the visco-elastic relaxation. It is generally considered that the two mechanisms affect different spatial and temporal scales: the afterslip is prevalent the first months in the surrounding of the fault, while the visco-elastic relaxation process affects a larger area and lasts a longer time. The time-space pattern of the measured deformation can help to characterize the rheology of the underlying structure.</p><p>In this work we look at the processes involved after the Iquique earthquake.</p><p> </p><p>To explore the processes driving the post-seismic deformation, we use a finite element model (FEM) (2D model, using the FEM software Pylith) that is constrained with InSAR and GNSS data. The GPS time series (processed with GipsyX) include 83 stations located in North Chile, Peru and Bolivia. The post-seismic signal is isolated using a trajectory model. The InSAR data consist in two Sentinel-1 time series (ascending and descending tracks) processed with the NSBAS chain, they include 514 interferograms, starting 7 months after the earthquake up to the end of 2019. In the model we impose a co-seismic displacement on the plate interface and explore the influence of the structure and the rheology on the predicted surface displacement.</p><p> </p><p>Our tests reveal that the viscosity in the continental and the oceanic mantle both have an impact on the displacement produced at the surface. The difference between these viscosities controls the movements allowed at depth. The crust thickness and the presence of a cold nose have a clear impact on the wavelength and the location of the maximum of amplitude, respectively.</p><p>The afterslip is the major contribution at short time. At longer time, it affects weakly the near trench displacements. To fit the long-term data, we show that visco-elastic relaxation is needed. After 7 months, the InSAR data show a clear spatial wavelength with a strong signal 150 to 300 km from the trench which can be explained by the visco-elastic process.</p><p>We pointing out that these is a trade-off between the contribution of afterslip and visco-elastic relaxation. However, both processes affect different space and time, and the comparison with GNSS data and two InSAR tracks allows to strongly constrain the model and reduce the range of plausible models.</p>

Vulcano is a composite volcanic edifice representing the southernmost emerged island of the Aeolian archipelago (Tyrrhenian Sea, Italy). Grown at the convergence of the Africa and Eurasian plates, Vulcano is part of a complex volcano-tectonic system characterized by a NNW-SSE fault system which controls the volcanism evolution of the Aeolian central branch and its continuous long-term deformation. Vulcano experienced many eruption episodes in historical times, the most recent of which occurred in 1888-1890. Since then, it has undergone repeated phases of unrest characterized by shallow seismicity, increased fumarolic activity, and sometimes ground deformation. The most recent unrest episode occurred from the summer till the end of 2021 with intense fumarolic activity mainly concentrated at La Fossa cone and some sectors of the caldera.  Long-term tectonic movement and unrest phases cause measurable deformation which can provide important insights into the volcanic system behavior. In this work we leveraged two types of satellite-geodesy deformation data: GNSS and InSAR. We considered the time series of 6 continuous GNSS stations managed by the Osservatorio Etneo of INGV. The SAR dataset consists of Sentinel-1A images acquired from January 2016 to December 2023 along ascending and descending orbits, and processed through the Interferometric Point Target Analysis (IPTA) approach to retrieve ground deformation velocity and displacement time series.Focusing on the 2016-2023 interval, we first validated the InSAR results with GNSS data, obtaining a general good agreement. The GNSS time series clearly show different phases of deviation from the long-term linear trend, particularly in 2018 and 2021. The 2021 period is associated with up to about 2 cm uplift and 1 cm of nearly radial pattern around La Fossa caldera. InSAR data are noisier, but also show transient signals, with a clear signal in 2021, associated with an elliptical deformation area of up to about 3 and locally 5 cm in Line of Sight at La Fossa caldera.InSAR and GNSS data provide complementary information respectively about the near- and far-field deformation pattern associated with the 2021 unrest phase. We jointly inverted these data using a new modeling algorithm implementing elastic and inelastic (thermo-poroelastic) sources to retrieve the volcanic source of the ground deformation observed during the recent unrest phase. Results indicate as preferred model a spheroid/cylindrical source located below La Fossa cone, with cumulated volume and pressure variations in agreement with previous studies using only InSAR or GNSS data. We also analyzed the 2018 deformation through GNSS data, whose pattern reveal an additional unrest episode possibly located at the northern edge of La Fossa Caldera. We discuss the highlighted unrest episodes in the context of the more general volcano-tectonic deformation pattern affecting the island.  

<p>In order to investigate crustal deformation within the upper plate of the Ionian Subduction Zone (ISZ) at different time scales, we have (i) mapped and modelled sequence of Late Quaternary raised marine terraces tectonically deformed by the West Crati normal fault, in northern Calabria, and (ii) refined geodetic rates of crustal extension from continuous GNSS measurements. Indeed, this region experienced damaging earthquakes such as the “1184 Valle del Crati” (M 6.7) and the “1638 Crotonese” (M 6.7) events, possibly on the West Crati Fault; however, an in-depth evaluation of the deformation rates inferred from geologic and GNSS data has not yet been performed. Furthermore, fault slip-rates and earthquake recurrence intervals for the understudied West Crati Fault are still debated and poorly-constrained. Raised Late Quaternary marine terraces are preserved on the footwall of the West Crati Fault; however, it is still debated if the “local” effect of the footwall uplift is affecting the “regional” signal of uplift likely related to the deformation associated either with the subduction or mantle upwelling processes. Within the investigated region lying in the northern part of the uplifting Calabrian-Peloritani Arc there are 32 regionally distributed permanent GNSS stations, for 18 of which the coordinate time series are adequately long (at least 4.5 years) to allow the study of the crustal kinematics. The data of these 18 stations are used to geodetically estimate fault slip-rates and then earthquake recurrence intervals for the West Crati Fault, with the aim of at least partially solve the aforementioned problem of the poor constrains. In particular, velocity and strain across this fault, based on reasonable hypotheses about the fault dip and the mechanical properties of the involved material, are computed starting from GNSS data about the surface kinematics.</p><p>Our preliminary results show that GIS-based elevations of Middle to Late Pleistocene palaeoshorelines, as well as temporally constant uplift rates, vary along the strike of the West Crati Fault, mapped on its footwall. This suggests that the fault slip-rate governing seismic hazard has also been constant through time, over multiple earthquake cycles. We then suggest that our geodetically-derived fault slip-rate for the West Crati Fault may be a more than reasonable value to be used over longer time scales for an improved seismic hazard approach, allowing to derive new earthquake recurrence intervals. These results thus suggest a significant yet understudied seismic hazard for the investigated area also because the regional extension might be likely accommodated by a few more active faults across-strike in northern Calabria. These facts highlight the importance of mapping crustal deformation within the upper plate above subduction zones to avoid unreliable interpretations relating to the mechanism controlling regional uplift.</p>

This paper presents a method for objective detection of long-term slow slip events with durations on the order of years, on the plate boundary along the Nankai Trough, relying on global navigation satellite system daily coordinate data. The Chugoku region of Japan was held fixed to remove common mode errors, and a displacement component was calculated relative to the direction of plate subduction. Correlations were then calculated between this displacement component and a 3-year ramp function with a 1-year slope. Nearly all periods of strong correlation coincide with periods of previously reported long-term slow slip events. A period of strong correlation around the Kii Channel in 2000–2002 is attributed to a previously undocumented long-term slow slip event beneath the Kii Channel and the eastern part of Shikoku Island with an equivalent moment magnitude of 6.6. This detection method reveals variation among long-term slow slip events along the Nankai Trough.

Detecting infrastructure hazard potential change by SAR techniques on postseismic surface deformation: A case study of 2016 Meinong earthquake in southwestern Taiwan

Throughout the earthquake cycling at fault zones, Earth’s crust undergoes deformations. GNSS coordinate time series record linear tectonic motion, seismic displacements, postseismic decays, and periodic signatures such as non-tidal loading. Any additional motions can be classed as transient tectonic signals, i.e., unexpected accelerations with respect to the standard trajectory model. As the number of permanent stations increases and as time series grow, we are increasingly able to recognise transient tectonic signals. Since some of these suspected tectonic transients have subtle magnitudes or sometimes unusual spatiotemporal features, we need to develop methods for determining which transients are artifacts and which are of tectonic origin.Here, we investigate the impact of certain GNSS processing choices and how they affect the appearance of transients in the GNSS displacement time series solutions. In this study, we choose data from Cascadia, a region for which the occurrence of transient signals in the GNSS time series is well known. We processed data based from 189 selected stations in network mode for the time span 2015 to 2019. After producing coordinate time series, we then built a pipeline to isolate processing artifacts and tectonic transients, using the regression model-based algorithm known as GrAtSiD (Greedy Automatic Signal Decomposition).The residuals of GNSS observations show that most sites have a precision of 5 mm to 10 mm. Using the GrAtSiD algorithm, we detected transient signals with velocities exceeding 0.3 mm/day near the ALBH station.

The area located East to Cairo, the Capital of Egypt, represents an obvious example for the rapid urban expansion, as it contains many new housing cities. Beside its socio-economic importance, it’s located in the Cairo-Suez seismic zone. We utilized Persistent Scatterer Interferometry (PSI) of 2015–2021 Sentinel-1 SAR scenes along with two GNSS stations (KATA and PHLW) to assess the distribution and rates of crustal deformation of this region. The PSI analysis is applied to 140 Sentinel-1 SAR images collected from the ascending track number (58) and the descending track number (167). The Bernese software V. 5.0 is used for the processing of the GNSS data. A good agreement between the rates estimated from the PSI analysis and GNSS data is observed. Based on results, most of the new cities showing land subsidence with variable rates. The rates at Obour, New Cairo, Shorouk, Madinty, and Capital Gardens are 0.54 ± 0.30 mm/year, 0.58 ± 0.30 mm/year, 1.01 ± 0.30 mm/year, 0.58 ± 0.30 mm/year, and 0.99 ± 0.30 mm/year, respectively. The highest recorded subsidence rates are at Asher, Administrative Capital, and Badr with 2.18 ± 0.30 mm/year, 1.89 ± 0.30 mm/year, and 1.69 ± 0.30 mm/year, respectively. Nasr city is the only city with an uplift of 0.82 ± 0.30 mm/year. Our new findings introduce the probable use of integrated techniques such as GNSS and InSAR to evaluate the extent of crustal deformation connected to rapid urbanization in arid areas. Beside tectonic setting, it should be considered while executing mega-projects for sustainable development especially within Egypt’s Vision 2030.

<p>Nowadays, both, the number of observations and the accuracy of satellite-based geodesic measurements, like GNSS, have increased. Therefore, GNSS provides more data as displacement values and velocities. This paper demonstrates that GNSS data analysis is a powerful tool to study geodynamic processes.</p><p>In this study, the analyzed GNSS data correspond to continuously recorded GPS (CGPS) stations, what we call the SPINA network. These stations are located in a region called Ibero-Maghrebian which includes the southern areas of the Iberian Peninsula and northern Africa.</p><p>The CGPS stations are included in the following organizations: RENEP (National Network of Permanent Stations), RAP (Andalusian Positioning Network), the Murcia Region CGPS Networks, ERVA (Valencian Reference Stations Network), IGN (National Geographic Institute) and the network TOPOIBERIA. The velocity was obtained in two steps: (1) preprocessing position time-series data of daily GPS measurements and (2) applying a combined model using the weighted least-squares method.</p><p>The prior knowledge of the crustal strain rate tensor provides a description of geodynamic processes such as the fault strain accumulation.</p><p>Based on the distribution of the GNSS stations, several grid sizes were tested to identify the best resolution. A Python script was used to compute the full two-dimensional velocity gradient tensor by means of inverting the GNSS velocities. The tensorial analysis provides different aspects of deformation, such as the maximum shear strain rate, including its direction, and the dilatation strain rate. These parameters can be used to characterize the mechanism of the current deformation.</p><p>Based on the computations from the GNSS-data model of components of horizontal deformations, the rates of both principal, values and axes, of the Earth’s crust deformation were found. Deformations measured in the Ibero-Maghrebian region with GPS could be interpreted in terms of either elastic loading or ductile deformation.</p>

Present-day crustal deformation was an attempt to estimate earthquake potential, yet the presence of postseismic deformation should be carefully identified. Studying crustal deformation in West Sumatra has been important for this purpose since the series of Sumatran Great Earthquake from 2004-2010. This study utilized present-day GNSS data (2017-2021) and pre-2004 GNSS velocities to understand the present-day crustal deformation. Bernese 5.2 was used to process the GNSS data and linear regression was used to calculate present-day velocities. These velocities were transformed into an ITRF2000-based Sundaland plate reference frame and then the velocities were compared to pre-2004 velocities in the same reference frame. The present-day velocities were ranging from 28.4 mm/yr to 58.3 mm/yr in ITRF2014 and from 8.8 to 44.8 mm/yr in the Sundaland plate reference frame. This suggests West Sumatra was located on the Sumatra block of the Sundaland plate. The low velocity difference ( 11.7 mm/yr) with the random vector direction between present-day velocities and pre-2004 velocities shows that there is no postseismic deformation affecting West Sumatra. This proposes the utilization of present-day velocities for earthquake potential estimation in West Sumatra.