ALBERT

Description: The geocentre motion is the motion of the centre of mass of the entire Earth, considered an isolated system, in a terrestrial system of reference. We first derive a formula relating the harmonic degree-1 Lagrangian variation of the gravity at a station to both the harmonic degree-1 vertical displacement of the station and the displacement of the whole Earth's centre of mass. The relationship is independent of the nature of the Earth deformation and is valid for any source of deformation. We impose no constraint on the system of reference, except that its origin must initially coincide with the centre of mass of the spherically symmetric Earth model. Next, we consider the geocentre motion caused by surface loading. In a system of reference whose origin is the centre of mass of the solid Earth, we obtain a specific relationship between the gravity variation at the surface, the geocentre displacement and the load Love number $h^{\prime }_1$ , which demands the Earth's structure and rheological behaviour be known. For various networks of real or fictitious stations, we invert synthetic signals of surface gravity variations caused by atmospheric loading to retrieve the degree-1 variation of gravity. We then select six well-distributed stations of the Global Geodynamics Project, which is a world network of superconducting gravimeters, to invert actual gravity data for the degree-1 variations and determine the geocentre displacement between the end of 2004 and the beginning of 2012, assuming it to be due to surface loading. We find annual and semi-annual displacements with amplitude 0.5–2.3 mm.

Description: We review the theory of the Earth's elastic and gravitational response to a surface disk load. The solutions for displacement of the surface and the geoid are developed using expansions of Legendre polynomials, their derivatives and the load Love numbers. We provide a matlab  function called diskload that computes the solutions for both uncompensated and compensated disk loads. In order to numerically implement the Legendre expansions, it is necessary to choose a harmonic degree, n max , at which to truncate the series used to construct the solutions. We present a rule of thumb (ROT) for choosing an appropriate value of n max , describe the consequences of truncating the expansions prematurely and provide a means to judiciously violate the ROT when that becomes a practical necessity.

Description: Two types of signals are clearly visible in continuous GPS (cGPS) time-series in Iceland, in particular in the vertical component. The first one is a yearly seasonal cycle, usually sinusoid-like with a minimum in the spring and a maximum in the fall. The second one is a trend of uplift, with higher values the closer the cGPS stations are to the centre of Iceland and ice caps. Here, we study the seasonal cycle signal by deriving its average at 71 GPS sites in Iceland. We estimate the annual and semi-annual components of the cycle in their horizontal and vertical components using a least-squares adjustment. The peak-to-peak amplitude of the cycle of the vertical component at the studied sites ranges from 4 mm near the coastline up to 27 mm at the centre of the Vatnajökull, the largest ice cap in Iceland. The minimum of the seasonal cycle occurs earlier in low lying areas than in the central part of Iceland, consistent with snow load having a large influence on seasonal deformation. Modelling shows that the seasonal cycle is well explained by accounting for elastically induced surface displacements due to snow, atmosphere, reservoir lake and ocean variations. Model displacement fields are derived considering surface loads on a multilayered isotropic spherical Earth. Through forward and inverse modelling, we were able to reproduce a priori information on the average seasonal cycle of known loads (atmosphere, snow in non-glaciated areas and lake reservoir) and get an estimation of other loads (glacier mass balance and ocean). The seasonal glacier mass balance cycle in glaciated areas and snow load in non-glaciated areas are the main contributions to the seasonal deformation. For these loads, induced seasonal vertical displacements range from a few millimetres far from the loads in Iceland, to more than 20 mm at their centres. Lake reservoir load also has to be taken into account on local scale as it can generate up to 20 mm of vertical deformation. Atmosphere load and ocean load are observable and generate vertical displacements in the order of a few millimetres. Inversion results also shows that the Iceland crust is less rigid than the world average. Interannual deviation from the GPS seasonal cycle can occur and are caused by unusual weather conditions over extended period of time.

Description: Seismic waves produced by fault ruptures give rise to gravity perturbations. So far, these perturbations have either been modelled as permanent coseismic gravity change in a half-space or spherical Earth model, or as full time-domain model in infinite space. In this paper, we present the explicit solution of gravity perturbations in time domain produced by a double-couple buried in a homogeneous half-space. This result is especially suited to study gravity perturbations up to a few hundreds of kilometres from the epicentre. It facilitates detailed parametric studies of gravity perturbations from fault rupture, and predicts gravity perturbations of real earthquakes with greatly improved accuracy. The results may serve to develop first designs of gravity-assisted earthquake early-warning systems, made possible by a new generation of ultrasensitive gravity gradiometers, which is currently under development.

Description: Continuous gravimetric observations have been made with three successive generations of superconducting gravimeter over 20 yr at Syowa Station ( $39.6\deg$ E, $69.0\deg$ S), East Antarctica. The third-generation instrument, OSG#058, was installed in January 2010 and was calibrated by an absolute gravimeter during January and February, 2010. The estimated scale factor was –73.823 ± 0.053 μGal V –1 (1 μGal = 10 –8 m s –2 ). The first 5 yr of OSG#058 data from 2010 January 7 to 2015 January 10 were decomposed into tidal waves (M3 to Ssa) and other non-tidal components by applying the Bayesian tidal analysis program BAYTAP. Long-term non-tidal gravity residuals, which were obtained by subtracting annual and 18.6 year tidal waves and the predicted gravity response to the Earth's variable rotation, showed significant correlation with the accumulated snow depth measured at Syowa Station. The greatest correlation occurred when the gravity variations lagged the accumulated snow depth by 21 d. To estimate the gravitational effect of the accumulated snow mass, we inferred a conversion factor of 3.13 ± 0.08 μGal m –1 from this relation. The accumulated snow depth at Syowa Station was found to represent an extensive terrestrial water storage (the snow accumulation) around Syowa Station, which was estimated from the Gravity Recovery and Climate Experiment satellite gravity data. The snow accumulation around Syowa Station was detectable by the superconducting gravimeter.

Description: The new release AIUB-RL02 of monthly gravity models from GRACE GPS and K-Band range-rate data is based on reprocessed satellite orbits referring to the reference frame IGb08. The release is consistent with the IERS2010 conventions. Improvements with respect to its predecessor AIUB-RL01 include the use of reprocessed (RL02) GRACE observations, new atmosphere and ocean dealiasing products (RL05), an upgraded ocean tide model (EOT11A), and the interpolation of shallow ocean tides (admittances). The stochastic parametrization of AIUB-RL02 was adapted to include daily accelerometer scale factors, which drastically reduces spurious signal at the 161 d period in C 20 and at other low degree and order gravity field coefficients. Moreover, the correlation between the noise in the monthly gravity models and solar activity is considerably reduced in the new release. The signal and the noise content of the new AIUB-RL02 monthly gravity fields are studied and calibrated errors are derived from their non-secular and non-seasonal variability. The short-period time-variable signal over the oceans, mostly representing noise, is reduced by 50 per cent with respect to AIUB-RL01. Compared to the official GFZ-RL05a and CSR-RL05 monthly models, the AIUB-RL02 stands out by its low noise at high degrees, a fact emerging from the estimation of seasonal variations for selected river basins and of mass trends in polar regions. Two versions of the monthly AIUB-RL02 gravity models, with spherical harmonics resolution of degree and order 60 and 90, respectively, are available for the time period from March 2003 to March 2014 at the International Center for Global Earth Models or from ftp://ftp.unibe.ch/aiub/GRAVITY/GRACE (last accessed 22 March 2016).

Description: We document two kinds of traveling ionospheric disturbances, namely, CTIDs (Co-tsunami-Traveling-Ionospheric-disturbances) and ATIDs (Ahead-of-Tsunami-Traveling-Ionospheric-disturbances) related to the Tohoku-Oki tsunami of 2011 March 11. They are referred to the disturbances that remain behind and ahead of the principal tsunami wave front, respectively. We first note their presence in a numerical experiment performed using a simulation code coupling the tsunami, atmosphere and ionosphere. This code uses the tsunami wavefield as an input and simulates acoustic-gravity waves (AGWs) in the atmosphere and TIDs, in the form of total electron content (TEC) disturbance, in the ionosphere. The simulated TEC reveals the excitation of CTIDs (at about 2 TECU) and ATIDs (at about 1 TECU), representing up to 5 per cent disturbance over the ambient electron density, and they arise from the dissipation of AGWs in the thermosphere. A novel outcome is that during the tsunami passage between ~6° and 12° of epicentral distance, strong ATIDs arrive ~20–60 min ahead of the tsunami wave front covering ~3°–10° of distance from the tsunami location. Simulation results are compared with the far-field observations using GNSS satellites and confirm that ATIDs are the first detected TEC maximum, occurring 20–60 min ahead of the tsunami arrival. Our simulation also confirms the presence of largest TEC maximum representing CTIDs, 10–20 min after the first tsunami wave. ATIDs reported in this study have characteristics that can be potentially used for the early warning of the tsunami.

Description: Apparent acceleration in Gravity Recovery and Climate Experiment (GRACE) Antarctic ice mass time-series may reflect both ice discharge and surface mass balance contributions. However, a recent study suggests there is also contamination from errors in atmospheric pressure de-aliasing fields [European Center for Medium-Range Weather Forecast (ECMWF) operational products] used during GRACE data processing. To further examine this question, we compare GRACE atmospheric pressure de-aliasing (GAA) fields with in situ surface pressure data from coastal and inland stations. Differences between the two are likely due to GAA errors, and provide a measure of error in GRACE solutions. Time-series of differences at individual weather stations are fit to four presumed error components: annual sinusoids, a linear trend, an acceleration term and jumps at times of known ECMWF model changes. Using data from inland stations, we estimate that atmospheric pressure error causes an acceleration error of about +7.0 Gt yr –2 , which is large relative to prior GRACE estimates of Antarctic ice mass acceleration in the range of –12 to –14 Gt yr –2 . We also estimate apparent acceleration rates from other barometric pressure (reanalysis) fields, including ERA-Interim, MERRA and NCEP/DOE. When integrated over East Antarctica, the four mass acceleration estimates (from GAA and the three reanalysis fields) vary considerably (by ~2–16 Gt yr –2 ). This shows the need for further effort to improve atmospheric mass estimates in this region of sparse in situ observations, in order to use GRACE observations to measure ice mass acceleration and related sea level change.

Description: The long-wavelength gravity field contains information about processes in the sublithospheric mantle. As satellite-derived gravity models now provide the long to medium-wavelength gravity field at unprecedented accuracy, techniques used to process gravity data need to be updated. We show that when determining these long-wavelengths, the treatment of topographic-isostatic effect (TIE) and isostatic effects (IE) is a likely source of error. We constructed a global isostatic model and calculated global TIE and IE. These calculations were done for ground stations as well as stations at satellite height. We considered both gravity and gravity gradients. Using these results, we determined how much of the gravity signal comes from distant sources. We find that a significant long-wavelength bias is introduced if far-field effects on the topographic effect are neglected. However, due to isostatic compensation far-field effects of the topographic effect are to a large degree compensated by the far-field IE. This means that far-field effects can be reduced effectively by always considering topographic masses together with their compensating isostatic masses. We show that to correctly represent the ultra-long wavelengths, a global background model should be used. This is demonstrated both globally and for a continental-scale case area in North America. In the case of regional modelling, where the ultra-long wavelengths are not of prime importance, gravity gradients can be used to help minimize correction errors caused by far-field effects.

Description: Constraining laterally varying structures in planetary interiors is important for understanding both the composition and the internal dynamics of a planet. Recognizing that seismic imaging technique is currently only viable for studying the Earth's interior structures, methods that can be supported by advanced space geodetic techniques may become alternatives to ‘image’ the interiors of other planets. The method of tidal tomography is one possibility, and it relies on high precision measurement of the response of a planet to its body tide. However, it is essential to develop an efficient analytical tool that computes the dependence of tidal response to 3-D interior structures. In this paper, we present a complete formulation of such an analytical tool, which calculates to high accuracy the tidal response of a terrestrial planet with lateral heterogeneities in its elastic and density structures. We treat the lateral heterogeneities as small perturbations and derive the governing equations based on the perturbation theory. In a spherical harmonic representation, equations at each order of perturbation are reduced into multiple matrix equations at harmonics that are allowed by mode couplings, and the total response equals the sum of all those single-harmonic responses, which can be solved semi-analytically. We test our perturbation method by applying it to the Moon with a harmonic degree-1 mantle structure for which the perturbation solutions of the tidal response are compared with those from a fully numerical method. The remarkable agreement between results from these two methods validates the perturbation method. As an example, we then use the perturbation method to evaluate the impact of lunar crustal thickness variations on tidal response of the Moon. We find that lunar crust produces much smaller degree-3 tidal responses than a relatively weak degree-1 structure in the deep lunar mantle. Our calculations show that degree-3 tidal response measurements may hold key constraints on possible degree-1 mantle structure of the Moon, as suggested from previous modelling results.

Description: Seismic data are primarily used in studies of the Earth's lithospheric structure including the Moho geometry. In regions, where seismic data are sparse or completely absent, gravimetric or combined gravimetric-seismic methods could be applied to determine the Moho depth. In this study, we derive and present generalized expressions for solving the Vening Meinesz–Moritz's (VMM) inverse problem of isostasy for a Moho depth determination from gravity and vertical gravity-gradient data. By solving the (non-linear) Fredholm's integral equation of the first kind, the linearized observation equations, which functionally relate the (given) gravity/gravity-gradient data to the (unknown) Moho depth, are derived in the spectral domain. The VMM gravimetric results are validated by using available seismic and gravimetric Moho models. Our results show that the VMM Moho solutions obtained by solving the VMM problem for gravity and gravity-gradient data are almost the same. This finding indicates that in global applications, using the global gravity/gravity-gradient data coverage, the spherical harmonic expressions for the gravimetric forward and inverse modelling yield (theoretically) the same results. Globally, these gravimetric solutions have also a relatively good agreement with the CRUST1.0 and GEMMA GOCE models in terms of their rms Moho differences (4.7 km and 4.1 km, respectively).

Description: Geophysical techniques are widely used to monitor volcanic unrest. A number of studies have also demonstrated that hydrological processes can produce or trigger geophysical signals. Hydrologically induced gravity signals have previously been recorded by specifically designed gravity surveys as well as, inadvertently, by volcano monitoring studies. Water table corrections of microgravity surveys are commonplace. However, the fluctuations of the water table beneath survey locations are often poorly known, and such a correction fails to account for changes in water-mass storage in the unsaturated zone. Here, we combine 2-D axis-symmetrical numerical fluid-flow models with an axis-symmetric, distributed-mass, gravity calculation to model gravity changes in response to fluctuating hydrological recharge. Flow simulations are based on tropical volcanic settings where high surface permeabilities promote thick unsaturated zones. Our study highlights that mass storage (saturation) changes within the unsaturated zone beneath a survey point can generate recordable gravity changes. We show that for a tropical climate, recharge variations can generate gravity variations of over 150 μGal; although, we demonstrate that for the scenarios investigated here, the probability of recording such large signals is low. Our modelling results indicate that microgravity survey corrections based on water table elevation may result in errors of up to 100 μGal. The effect of inter-annual recharge fluctuations dominate over seasonal cycles which makes prediction and correction of the hydrological contribution more difficult. Spatial hydrogeological heterogeneity can also impact on the accuracy of relative gravity surveys, and can even result in the introduction of additional survey errors. The loading fluctuations associated with saturation variations in the unsaturated zone may also have implications for other geophysical monitoring techniques, such as geodetic monitoring of ground deformation.

Description: Several attempts have been made to obtain a radiographic image inside volcanoes using cosmic-ray muons (muography). Muography is expected to resolve highly heterogeneous density profiles near the surface of volcanoes. However, several prior works have failed to make clear observations due to contamination by background noise. The background contamination leads to an overestimation of the muon flux and consequently a significant underestimation of the density in the target mountains. To investigate the origin of the background noise, we performed a Monte Carlo simulation. The main components of the background noise in muography are found to be low-energy protons, electrons and muons in case of detectors without particle identification and with energy thresholds below 1 GeV. This result was confirmed by comparisons with actual observations of nuclear emulsions. This result will be useful for detector design in future works, and in addition some previous works of muography should be reviewed from the view point of background contamination.

Description: Essential to understanding sea level change and its causes during the last interglacial (LIG) is the quantification of uncertainties. In order to estimate the uncertainties, we develop a statistical framework for the comparison of palaeoclimatic sea level index points and GIA model predictions. For the investigation of uncertainties, as well as to generate better model predictions, we implement a massive ensemble approach by applying a data assimilation scheme based on particle filter methods. The different runs are distinguished through varying ice sheet reconstructions based on oxygen-isotope curves and different parameter selections within the GIA model. This framework has several advantages over earlier work, such as the ability to examine either the contribution of individual observations to the results or the probability of specific input parameters. This exploration of input parameters and data leads to a larger range of estimates than previously published work. We illustrate how the assumptions that enter into the statistical analysis, such as the existence of outliers in the observational database or the initial ice volume history, can introduce large variations to the estimate of the maximum highstand. Thus, caution is required to avoid overinterpreting results. We conclude that there are reasonable doubts whether the data sets previously used in statistical analyses are able to tightly constrain the value of maximum highstand during the LIG.

Description: A 3-D density model of the crust and upper mantle beneath the Karoo basin is presented here. The model is constrained using potential field, borehole and seismic data. Uplift of the basin by the end of the Cretaceous has resulted in an unusually high plateau (〉1000 m) covering a large portion of South Africa. Isostatic studies show the topography is largely compensated by changes in Moho depths (~35 km on-craton and 〉45 km off-craton) and changes in lithospheric mantle densities between the Kaapvaal Craton and surrounding regions (~50 kg m –3 increase from on- to off-craton). This density contrast is determined by inverted satellite gravity and gravity gradient data. The highest topography along the edge of the plateau (〉1200 m) and a strong Bouguer gravity low over Lesotho, however, can only be explained by a buoyant asthenosphere with a density decrease of around 40 kg m –3 .

Description: A method to estimate the rotation change in the orientation of the centre-of-figure (CF) frame caused by earthquakes is proposed for the first time. This method involves using the point dislocation theory based on a spherical, non-rotating, perfectly elastic and isotropic (SNREI) Earth. The rotation change in the orientation is related solely to the toroidal displacements of degree one induced by the vertical dip slip dislocation, and the spheroidal displacements induced by an earthquake have no contribution. The effects of two recent large earthquakes, the 2004 Sumatra and the 2011 Tohoku-Oki, are studied. Results showed that the Sumatra and Tohoku-Oki earthquakes both caused the CF frame to rotate by at least tens of μas (micro-arc-second). Although the visible co-seismic displacements are identified and removed from the coordinate time-series, the rotation change due to the unidentified ones and errors in removal is non-negligible. Therefore, the rotation change in the orientation of the CF frame due to seismic deformation should be taken into account in the future in reference frame and geodesy applications.

Description: In this study, we propose an approach for determining the geopotential difference using high-frequency-stability microwave links between satellite and ground station based on Doppler cancellation system. Suppose a satellite and a ground station are equipped with precise optical-atomic clocks (OACs) and oscillators. The ground oscillator emits a signal with frequency f a towards the satellite and the satellite receiver (connected with the satellite oscillator) receives this signal with frequency f b which contains the gravitational frequency shift effect and other signals and noises. After receiving this signal, the satellite oscillator transmits and emits, respectively, two signals with frequencies f b and f c towards the ground station. Via Doppler cancellation technique, the geopotential difference between the satellite and the ground station can be determined based on gravitational frequency shift equation by a combination of these three frequencies. For arbitrary two stations on ground, based on similar procedures as described above, we may determine the geopotential difference between these two stations via a satellite. Our analysis shows that the accuracy can reach 1 m 2 s – 2 based on the clocks’ inaccuracy of about 10 –17 (s s –1 ) level. Since OACs with instability around 10 –18 in several hours and inaccuracy around 10 –18 level have been generated in laboratory, the proposed approach may have prospective applications in geoscience, and especially, based on this approach a unified world height system could be realized with one-centimetre level accuracy in the near future.

Description: A sequence of large earthquakes occurred along the North Anatolian fault in the 20th century. These earthquakes, including the 1999 Izmit/Düzce earthquakes, generally propagated westward towards the Marmara Sea, defining the Main Marmara fault as a potential seismic gap. It is important to conduct a detailed assessment of the seismic hazards along the main Marmara fault because the megacity Istanbul lies only approximately 10 km north of the eastern segment of the Main Marmara fault, which is referred to as the Princes’ Islands Fault segment (PIF). Here, we study the locking status of this fault segment to evaluate the seismic hazard potential. For the first time, combined ascending and descending Interferometric Synthetic Aperture Radar and Global Positioning System observations were used to investigate the crustal deformation associated with the PIF. After careful corrections of the estimated ground velocity, a deformation pattern relating to fault locking near the Princes’ Islands was identified. The modeling results revealed that the slip rate and locking depth of the fault segment show a clear trade-off, which were estimated as 18.9 ± 7.2 mm yr –1 and 12.1 ± 7.0 km, respectively. With a moment accumulation rate of 1.7 ± 0.4  x  10 17 Nm yr –1 (proportional to the product of slip rate and locking depth), our results imply a build-up of a geodetic moment on the PIF and therefore a potential for earthquake hazards in the vicinity of the Istanbul megacity.

Description: We present a distributed slip model for the 1999 M w 6.3 Chamoli earthquake of north India using interferometric synthetic aperture radar (InSAR) data from both ascending and descending orbits and Bayesian estimation of confidence levels and trade-offs of the model geometry parameters. The results of fault-slip inversion in an elastic half-space show that the earthquake ruptured a $9 _{ - 2.2}^{\circ + 3.4}$ northeast-dipping plane with a maximum slip of ~1 m. The fault plane is located at a depth of ~ $15.9_{ - 3.0}^{ + 1.1}$ km and is ~120 km north of the Main Frontal Thrust, implying that the rupture plane was on the northernmost detachment near the mid-crustal ramp of the Main Himalayan Thrust. The InSAR-determined moment is 3.35 x 10 18 Nm with a shear modulus of 30 GPa, equivalent to M w 6.3, which is smaller than the seismic moment estimates of M w 6.4–6.6. Possible reasons for this discrepancy include the trade-off between moment and depth, uncertainties in seismic moment tensor components for shallow dip-slip earthquakes and the role of earth structure models in the inversions. The released seismic energy from recent earthquakes in the Garhwal region is far less than the accumulated strain energy since the 1803 M s 7.5 earthquake, implying substantial hazard of future great earthquakes.

Description: We use a Bayesian formalism combined with a grid node discretization for the linear inversion of gravimetric data in terms of 3-D density distribution. The forward modelling and the inversion method are derived from seismological inversion techniques in order to facilitate joint inversion or interpretation of density and seismic velocity models. The Bayesian formulation introduces covariance matrices on model parameters to regularize the ill-posed problem and reduce the non-uniqueness of the solution. This formalism favours smooth solutions and allows us to specify a spatial correlation length and to perform inversions at multiple scales. We also extract resolution parameters from the resolution matrix to discuss how well our density models are resolved. This method is applied to the inversion of data from the volcanic island of Basse-Terre in Guadeloupe, Lesser Antilles. A series of synthetic tests are performed to investigate advantages and limitations of the methodology in this context. This study results in the first 3-D density models of the island of Basse-Terre for which we identify: (i) a southward decrease of densities parallel to the migration of volcanic activity within the island, (ii) three dense anomalies beneath Petite Plaine Valley, Beaugendre Valley and the Grande-Découverte-Carmichaël-Soufrière Complex that may reflect the trace of former major volcanic feeding systems, (iii) shallow low-density anomalies in the southern part of Basse-Terre, especially around La Soufrière active volcano, Piton de Bouillante edifice and along the western coast, reflecting the presence of hydrothermal systems and fractured and altered rocks.

Description: We compute the gravimetric factor at the Chandler wobble (CW) frequency using time-series from superconducting gravimeters (SG) longer than a decade. We first individually process the polar motion and data at each individual gravity station to estimate the gravimetric factor amplitude and phase, then we make a global analysis by applying a stacking method to different subsets of up to seven SG stations. The stacking is an efficient way of getting rid of local effects and improving the signal-to-noise ratio of the combined data sets. Using the stacking method, we find a gravimetric factor amplitude and phase of 1.118 ± 0.016 and –0.45 ± 0.66 deg, respectively, which is smaller in amplitude than expected. The sources of error are then carefully considered. For both local and global analyses, the uncertainties on our results are reliably constrained by computing the standard deviation of the estimates of the gravimetric factor amplitude and phase for increasing length of the time-series. Constraints on the CW anelastic dissipation can be set since any departure of the gravimetric factor from its elastic value may provide some insights into the dissipative processes that occur at the CW period. In particular, assuming given rheological models for the Earth's mantle enables us to make the link between the gravimetric factor phase and the CW quality factor.

Description: Analysing independent 1-yr data sets of 10 European superconducting gravimeters (SG) reveals statistically significant temporal variations of M2 tidal parameters. Both common short-term (〈2 yr) and long-term (〉2 yr) features are identified in all SG time-series but one. The averaged variations of the amplitude factor are about 0.2. The path of load vector variations equivalent to the temporal changes of tidal parameters suggests the presence of an 8.85 yr modulation (lunar perigee). The tidal waves having the potential to modulate M2 with this period belong to the 3rd degree constituents. Their amplitude factors turn out to be much closer to body tide model predictions than that of the main 2nd degree M2, which indicates ocean loading for 3rd degree waves to be less prominent than for 2nd degree waves within the M2 group. These two different responses to the loading suggest that the observed modulation is more due to insufficient frequency resolution of limited time-series rather than to time variable loading. Presently, SG gravity time-series are still too short to prove if time variable loading processes are involved too as in case of the annual M2 modulation known to appear for analysis intervals of less than 1 yr. Whatever the variations are caused by, they provide the upper accuracy limit for earth model validation and permit estimating the temporal stability of SG scale factors and assessing the quality of gravity time-series.

Description: A sequence of large earthquakes occurred along the North Anatolian fault in the 20th century. These earthquakes, including the 1999 Izmit/Düzce earthquakes, generally propagated westward towards the Marmara Sea, defining the Main Marmara fault as a potential seismic gap. It is important to conduct a detailed assessment of the seismic hazards along the main Marmara fault because the megacity Istanbul lies only approximately 10 km north of the eastern segment of the Main Marmara fault, which is referred to as the Princes’ Islands Fault segment (PIF). Here, we study the locking status of this fault segment to evaluate the seismic hazard potential. For the first time, combined ascending and descending Interferometric Synthetic Aperture Radar and Global Positioning System observations were used to investigate the crustal deformation associated with the PIF. After careful corrections of the estimated ground velocity, a deformation pattern relating to fault locking near the Princes’ Islands was identified. The modeling results revealed that the slip rate and locking depth of the fault segment show a clear trade-off, which were estimated as 18.9 ± 7.2 mm yr –1 and 12.1 ± 7.0 km, respectively. With a moment accumulation rate of 1.7 ± 0.4  x  10 17 Nm yr –1 (proportional to the product of slip rate and locking depth), our results imply a build-up of a geodetic moment on the PIF and therefore a potential for earthquake hazards in the vicinity of the Istanbul megacity.

Description: Deformation analysis in general and strain analysis in particular using permanent GPS networks require proper analysis of time-series in which all functional effects are taken into consideration and all stochastic effects are captured using an appropriate noise model. This contribution addresses both issues when considering the strain parameters of a GPS network. Estimates of spatial correlation, time correlated noise, and multivariate power spectrum for daily position time-series of the Southern California Integrated GPS Network (SCIGN) stations collected between 1996 and 2011 are obtained. Significant signals with periods of 13.63 d and those related to the GPS draconitic year are identified in these time-series. We aim to assess the effect of a realistic noise model of the series on the uncertainties of the strain parameters including displacements, normal and shear strains, and rotations. For the SCIGN network considered, the following results are highlighted. Contrary to the common belief, the uncertainties of the displacements parameters become smaller when taking a realistic noise model into account. This however was not the case when assessing the noise characteristics of the normal and shear strain, and rotation parameters. The uncertainties increase nearly by a factor of two, in agreement to what is expected. Some of the significant deformation parameters of the white noise model become less significant in case of the realistic noise model.

Description: We have extended backwards from 2001 to 1979 the current release 05 (RL05) of the Gravity Recovery and Climate Experiment (GRACE) Atmospheric and Oceanic De-aliasing Level-1B (AOD1B) product and studied the impact of this and a previous release 04 (RL04) of the AOD1B product on precise orbits of five altimetry satellites (ERS-1, ERS-2, TOPEX/Poseidon, Envisat and Jason-1) for the time span 1991–2012, as compared to the case when no AOD1B product is used. We have found that using AOD1B RL05 product reduces root mean square (RMS) fits of satellite laser ranging (SLR) observations by about 1.0–6.4 per cent, 2-d arc overlaps in radial, cross-track and along-track directions by about 1.3–12.0, 0.3–10.0 and 2.0–10.0 per cent, respectively, for various satellites tested, as compared to the case without AOD1B product. Using AOD1B RL05 product instead of RL04 one reduces SLR RMS fits by 0.1–0.7 per cent, 2-d arc overlaps in radial, cross-track and along-track directions by 0.1–0.6, 0.1–1.3 and 0.2–1.2 per cent, respectively, for the satellite orbits tested. The multi-mission crossover analysis shows that the application of an AOD1B product reduces the scatter of radial errors by 0.4–2.8 per cent for the satellite missions studied. At the regions with the most pronounced changes the use of the AOD1B products improves the consistency between the sea level as measured by the TOPEX and ERS-2 missions and by the Jason-1 and Envisat missions by 5 to 10 per cent (globally by about 2 per cent). The results of our study show that extended AOD1B RL05 product performs better than the AOD1B RL04 and improves orbits of altimetry satellites and consistency of sea level products.

Description: A new analytical method for the computation of a truncated series of solid spherical harmonic coefficients (HCs) from data on a spheroid (i.e. an oblate ellipsoid of revolution) is derived, using a transformation between surface and solid spherical HCs. A two-step procedure is derived to extend this transformation beyond degree and order (d/o) 520. The method is compared to the Hotine–Jekeli transformation in a numerical study based on the EGM2008 global gravity model. Both methods are shown to achieve submicrometre precision in terms of height anomalies for a model to d/o 2239. However, both methods result in spherical harmonic models that are different by up to 7.6 mm in height anomalies and 2.5 mGal in gravity disturbances due to the different coordinate system used. While the Hotine–Jekeli transformation requires the use of an ellipsoidal coordinate system, the new method uses only spherical polar coordinates. The Hotine–Jekeli transformation is numerically more efficient, but the new method can more easily be extended to cases where (a linear combination of) normal derivatives of the function under consideration are given on the surface of the spheroid. It therefore provides a solution to many types of ellipsoidal boundary-value problems in the spectral domain.

Description: We explore Earth's elastic deformation response to ocean tidal loading (OTL) using kinematic Global Positioning System (GPS) observations and forward-modelled predictions across South America. Harmonic coefficients are extracted from up to 14 yr of GPS-inferred receiver locations, which we estimate at 5 min intervals using precise point positioning. We compare the observed OTL-induced surface displacements against predictions derived from spherically symmetric, non-rotating, elastic and isotropic (SNREI) Earth models. We also compare sets of modelled predictions directly for various ocean-tide and Earth-model combinations. The vector differences between predicted displacements computed using separate ocean-tide models reveal uniform-displacement components common to all stations in the South America network. Removal of the network-mean OTL-induced displacements from each site substantially reduces the vector differences between observed and predicted displacements. We focus on the dominant astronomical tidal harmonics from three distinct frequency bands: semidiurnal (M 2 ), diurnal (O 1 ) and fortnightly (M f ). In each band, the observed OTL-induced surface displacements strongly resemble the modelled displacement-response patterns, and the residuals agree to about 0.3 mm or better. Even with the submillimetre correspondence between observations and predictions, we detect regional-scale spatial coherency in the final set of residuals, most notably for the M 2 harmonic. The spatial coherency appears relatively insensitive to the specific choice of ocean-tide or SNREI-Earth model. Varying the load model or 1-D elastic structure yields predicted OTL-induced displacement differences of order 0.1 mm or less for the network. Furthermore, estimates of the observational uncertainty place the noise level below the magnitude of the residual displacements for most stations, supporting our interpretation that random errors cannot account for the entire misfit. Therefore, the spatially coherent residuals may reveal deficiencies in the a priori SNREI Earth models. In particular, the residuals may indicate sensitivity to regional deviations from standard globally averaged Earth structure due to the presence of the South American craton.

Description: Planar faults are widely adopted during inversions to determine slip distributions and fault geometries using geodetic observations; however, little research has been conducted with respect to curved faults. We attribute this to the lack of an appropriate parameterized modelling method. In this paper, we present a curved-fault modelling method (CFMM) that describes a curved fault according to specific parameters, and we also develop a corresponding hybrid iterative inversion algorithm (HIIA) to perform inversions for parametric curved-fault geometries and slips. The results of the strike-component and dip-component synthetic tests show that a complex S-shaped fault surface and a circular slip distribution are successfully recovered, indicating the strong performance of the CFMM and HIIA methods. In addition, we describe and verify a scenario for determining the number of necessary geometrical parameters for the HIIA and examine the case study of the Wenchuan earthquake, which occurred on a complex listric fault surface. During the iteration process of the HIIA, both the fault geometry and slip distribution of the Beichuan and Pengguan faults converge to optimal values, indicating a Beichuan fault (BCF) model with a continuous listric shape and gradual steepening from the southwest to the northeast, which is highly consistent with geological survey results. Both the synthetic and real-world case studies show that the HIIA and the CMFF are superior to the conventional fault modelling method based on rectangular planes and that these models have the potential for use in more integrated research involving inversion studies, such as joint slip/curved-fault-geometry inversions that take into account data resolving power.

Description: Globally gridded estimates of monthly-mean anomalies of terrestrial water storage (TWS) are estimated from the most recent GRACE release 05a of GFZ Potsdam in order to provide non-geodetic users a convenient access to state-of-the-art GRACE monitoring data. We use an ensemble of five global land model simulations with different physics and different atmospheric forcing to obtain reliable gridded scaling factors required to correct for spatial leakage introduced during data processing. To allow for the application of this data-set for large-scale monitoring tasks, model validation efforts, and subsequently also data assimilation experiments, globally gridded estimates of TWS uncertainties that include (i) measurement, (ii) leakage and (iii) re-scaling errors are provided as well. The results are generally consistent with the gridded data provided by Tellus, but deviate in some basins which are largely affected by the uncertainties of the model information required for re-scaling, where the approach based on the median of a small ensemble of global land models introduced in this paper leads to more robust results.

Description: Regional recovery of the disturbing gravitational potential in the area of Central Europe from satellite gravitational gradients data is discussed in this contribution. The disturbing gravitational potential is obtained by inverting surface integral formulas which transform the disturbing gravitational potential onto disturbing gravitational gradients in the spherical local north-oriented frame. Two numerical approaches that solve the inverse problem are considered. In the first approach, the integral formulas are rigorously decomposed into two parts, that is, the effects of the gradient data within near and distant zones. While the effect of the near zone data is sought as an inverse problem, the effect of the distant zone data is synthesized from the global gravitational model GGM05S using spectral weights given by truncation error coefficients up to the degree 150. In the second approach, a reference gravitational field up to the degree 180 is applied to reduce and smooth measured gravitational gradients. In both cases we recovered the disturbing gravitational potential from each of the four well-measured gravitational gradients of the GOCE satellite separately as well as from their combination. Obtained results are compared with the EGM2008, DIR-r2, TIM-r2 and SPW-r2 global gravitational models. The best fit was achieved for EGM2008 and the second approach combining all four well-measured gravitational gradients with rms of 1.231 m 2  s –2 .

Description: We present efficient Fourier-domain algorithms for modelling gravity effects due to topographic masses. The well-known Parker's formula originally based on the standard fast Fourier transform (FFT) algorithm is modified by applying the Gauss–FFT method instead. Numerical precision of the forward and inverse Fourier transforms embedded in Parker's formula and its extended forms are significantly improved by the Gauss–FFT method. The topographic model is composed of two major aspects, the geometry and the density. Versatile geometric representations, including the mass line model, the mass prism model, the polyhedron model and smoother topographic models interpolated from discrete data sets using high-order splines or pre-defined by analytical functions, in combination with density distributions that vary both laterally and vertically in rather arbitrary ways following exponential or general polynomial functions, now can be treated in a consistent framework by applying the Gauss–FFT method. The method presented has been numerically checked by space-domain analytical and hybrid analytical/numerical solutions already established in the literature. Synthetic and real model tests show that both the Gauss–FFT method and the standard FFT method run much faster than space-domain solutions, with the Gauss–FFT method being superior in numerical accuracy. When truncation errors are negligible, the Gauss–FFT method can provide forward results almost identical to space-domain analytical or semi-numerical solutions in much less time.

Description: The gravity gradient tensor (GGT) has been increasingly used in practical applications, but the advantages and the disadvantages of the analysis of GGT components versus the analysis of the vertical component of the gravity field are still debated. We analyse the performance of joint inversion of GGT components versus separate inversion of the gravity field alone, or of one tensor component. We perform our analysis by inspection of the Picard Plot, a Singular Value Decomposition tool, and analyse both synthetic data and gradiometer measurements carried out at the Vredefort structure, South Africa. We show that the main factors controlling the reliability of the inversion are algebraic ambiguity (the difference between the number of unknowns and the number of available data points) and signal-to-noise ratio. Provided that algebraic ambiguity is kept low and the noise level is small enough so that a sufficient number of SVD components can be included in the regularized solution, we find that: (i) the choice of tensor components involved in the inversion is not crucial to the overall reliability of the reconstructions; (ii) GGT inversion can yield the same resolution as inversion with a denser distribution of gravity data points, but with the advantage of using fewer measurement stations.

Description: This paper compares GRACE (Gravity Recovery and Climate Experiment) and ICESat (Ice, Cloud and land Elevation Satellite) observations to confirm whether the observed gravity increase in the Tibetan Plateau (TP) was primarily caused by lake storage gain, and comprehensively analyses the changing pattern of lake level over 2003–2009. An improved automated method was used to obtain lake-level changes and the underestimation of lake water storage was considered due to lake area expansion and lake density. The result demonstrates that GRACE recorded a mass gain (16.43 ± 1.65/11.79 ± 1.25 gt a –1 ) in the total/inner TP, of which lake storage increase accounts for (8.78 ± 0.75/7.53 ± 0.56 gt a –1 ) based on ICESat. The northwestern residual may be stored in new lakes and soil moisture as a result of net precipitation gain. According to the character of the lake-level changes, we divide the TP into four subregions. Generally, the changing pattern of lake level concurs with the distribution of precipitation, which is increasing in the inner TP and decreasing in the upstream area of the Indus and Brahmaputra Rivers. An excess of rainfall in the northeastern TP in the summer of 2005 and 2009 caused a simultaneous large increase in water level in many lakes. The correlation of lake changes with precipitation demonstrates that precipitation rather than glacial melt is the main cause of lake-level change in most places. Nonetheless, the meltwater is a considerable supplement for lakes near glaciers such as Selin Co and Nam Co, which partly explains why GRACE indicates a much weaker signal in this region.

Description: Traditional processing of Global Navigation Satellite System (GNSS) data using dedicated scientific software has provided the highest levels of positional accuracy, and has been used extensively in geophysical deformation studies. To achieve these accuracies a significant level of understanding and training is required, limiting their availability to the general scientific community. Various online GNSS processing services, now freely available, address some of these difficulties and allow users to easily process their own GNSS data and potentially obtain high quality results. Previous research into these services has focused on Continually Operating Reference Station (CORS) GNSS data. Less research exists on the results achievable with these services using large campaign GNSS data sets, which are inherently noisier than CORS data. Even less research exists on the quality of velocity fields derived from campaign GNSS data processed through online precise point positioning services. Particularly, whether they are suitable for geodynamic and deformation studies where precise and reliable velocities are needed. In this research, we process a very large campaign GPS data set (spanning 10 yr) with the online Jet Propulsion Laboratory Automated Precise Positioning Service. This data set is taken from a GNSS network specifically designed and surveyed to measure deformation through the central North Island of New Zealand. This includes regional CORS stations. We then use these coordinates to derive a horizontal and vertical velocity field. This is the first time that a large campaign GPS data set has been processed solely using an online service and the solutions used to determine a horizontal and vertical velocity field. We compared this velocity field to that of another well utilized GNSS scientific software package. The results show a good agreement between the CORS positions and campaign station velocities obtained from the two approaches. We discuss the implications of these results for how future GNSS campaign field surveys might be conducted and how their data might be processed.

Description: We have developed a method to estimate the geometry, location and densities of anomalies coming from 2-D gravity data based on compact gravity inversion technique. Compact gravity inversion is simple, fast and user friendly but severely depends on the number of model parameters, that is, by increasing the model parameters, the anomalies tend to concentrate near the surface. To overcome this ambiguity new weighting functions based on density contrast, depth, and compactness models have been introduced. Variable compactness factors have been defined here to get either a sharp or a smooth model based on the depth of the source or existence of prior information. Depth weighting derived from one station of gravity data whereas the effect of gravity data is 2-D and 3-D. To compensate this limitation an innovating weighting function namely kernel function has been introduced which multiplies with weight and compactness matrixes to yield a general model weighting function. The method is tested using three different sets of synthetic examples: a body at various depths (20, 40, 80 and 140 m), two bodies at the same depth but various distances to estimate lateral resolution and three bodies with negative and positive density contrast in different depths. The method is also applied to three real gravity data of Woodlawn massive sulphide body, sulphides mineralization of British Colombia and iron ore body of Missouri. The method produces solutions consistent with the known geologic attributes of the gravity sources, illustrating its potential practicality.

Description: This paper resurrects a version of Poisson's Partial Differential Equation (PDE) associated with the gravitational field at the Earth's surface and illustrates how the PDE possesses a capability to extract the mass density of Earth's topography from land-based gravity data. Herein, first we propound a theorem which mathematically introduces this version of Poisson's PDE adapted for the Earth's surface and then we use this PDE to develop a method of approximating the terrain mass density. Also, we carry out a real case study showing how the proposed approach is able to be applied to a set of land-based gravity data. In the case study, the method is summarized by an algorithm and applied to a set of gravity stations located along a part of the north coast of the Persian Gulf in the south of Iran. The results were numerically validated via rock-samplings as well as a geological map. Also, the method was compared with two conventional methods of mass density reduction. The numerical experiments indicate that the Poisson PDE at the Earth's surface has the capability to extract the mass density from land-based gravity data and is able to provide an alternative and somewhat more precise method of estimating the terrain mass density.

Description: The Moho surface can be determined according to isostatic theories, and among them, the recent Vening Meinesz-Moritz (VMM) theory of isostasy has been successfully applied for this purpose. In this paper, this method is studied from a theoretical prospective and its connection to the Airy–Heiskanen (AH) and Vening Meinesz original theories are presented. Jeffrey's inverse solution to isostasy is developed according to the recent developments of the VMM method and both are compared in similar situations. It is shown that they are generalizations of the AH model in a global and continuous domain. In the VMM spherical harmonic solution for Moho depth, the mean Moho depth contributes only to the zero-degree term of the series, while in Jeffrey's solution it contributes to all frequencies. In addition, the VMM spherical harmonic series is improved further so that the mean Moho can contribute to all frequencies of the solution. This modification makes the VMM global solution superior to the Jeffrey one, but in a global scale, the difference between both solutions is less than 3 km. Both solutions are asymptotically convergent and we present two methods to obtain smooth solutions for Moho from them.

Description: In a pioneering study, Wahr & Bergen developed the widely adopted, pseudo-normal mode framework for predicting the impact of anelastic effects on the Earth's body tides. Lau et al. have recently derived an extended normal mode treatment of the problem (as well as a minor variant of the theory known as the direct solution method) that makes full use of theoretical developments in free oscillation seismology spanning the last quarter century and that avoids a series of assumptions and approximations adopted in the traditional theory for predicting anelastic effects. There are two noteworthy differences between these two theories: (1) the traditional theory only considers perturbations to the eigenmodes of an elastic Earth, whereas the new theory augments this set of modes to include the relaxation modes that arise in anelastic behaviour; and (2) the traditional theory approximates the complex perturbation to the tidal Love number as a scaled version of the complex perturbation to the elastic moduli, whereas the new theory computes the full complex perturbation to each eigenmode. In this study, we highlight the above differences using a series of synthetic calculations, and demonstrate that the traditional theory can introduce significant error in predictions of the complex perturbation to the Love numbers due to anelasticity and the related predictions of tidal lag angles. For the simplified Earth models we adopt, the computed lag angles differ by ~20 per cent. The assumptions in the traditional theory have important implications for previous studies that use model predictions to correct observables for body tide signals or that analyse observations of body tide deformation to infer mantle anelastic structure. Finally, we also highlight the fundamental difference between apparent attenuation (i.e. attenuation inferred from observations or predicted using the above theories) and intrinsic attenuation (i.e. the material property investigated through experiments), where both are often expressed in terms of lag angles or Q –1 . In particular, we demonstrate the potentially significant (factor of two or more) bias introduced in estimates of Q –1 and its frequency dependence in studies that have treated Q –1 determined from tidal phase lags or measured experimentally as being equal. The observed or theoretically predicted lag angle (or apparent Q –1 ) differs from the intrinsic, material property due to inertia, self-gravity and effects associated with the energy budget. By accounting for these differences we derive, for a special case, an expression that accurately maps apparent attenuation predicted using the extended normal mode formalism of Lau et al. into intrinsic attenuation. The theory allows for more generalized mappings which may be used to robustly connect observations and predictions of tidal lag angles to results from laboratory experiments of mantle materials.

Description: We consider a new approach to both the forward and inverse problems in post-seismic deformation. We present a method for forward modelling post-seismic deformation in a self-gravitating, heterogeneous and compressible earth with a variety of linear and nonlinear rheologies. We further demonstrate how the adjoint method can be applied to the inverse problem both to invert for rheological structure and to calculate the sensitivity of a given surface measurement to changes in rheology or time-dependence of the source. Both the forward and inverse aspects are illustrated with several numerical examples implemented in a spherically symmetric earth model.

Description: The Seiland Igneous Province (SIP) is the largest complex of mafic and ultramafic intrusions in northern Fennoscandia intruded at ca . 580–560 Ma. The depth extent and the deep structure of the SIP are mainly unknown apart from three profiles modelled by gravity and refraction seismic data. Utilizing 3-D gravity modelling, a complex model of the deep subsurface structure of the SIP has been developed. The structure is presented in a multiprofile model ranging from the surface to the Moho. The mafic/ultramafic rocks of the SIP are modelled with densities of 3100 and 3300 kg m –3 , the surrounding rocks by densities of 2700 and 2900 kg m –3 for upper and lower crust, respectively. This density model explains the pronounced positive Bouguer gravity anomaly of up to 100 mGal above background. Its minimum volume is estimated from the subsurface model to 17 000 km 3 and as such we revise downwards the earlier estimations of 25 000 km 3 . The new subsurface model suggests that most of the SIP has a thickness between 2 and 4 km. An area with roots in an annular pattern is found and two deep-reaching roots have been identified located below the islands of Seiland and Sørøy. The depth of these roots is estimated to approximatively 9 km. The SIP is presently interpreted to be in the Caledonian Kalak Nappe Complex and the roots depth constrains its minimum thickness which is larger than earlier estimated. Furthermore, the rather undisturbed shape of the annular root pattern indicates that the SIP has not been subjected to strong tectonic reworking during the Caledonian orogeny.

Description: Estimating the relief of the Moho from gravity data is a computationally intensive nonlinear inverse problem. What is more, the modelling must take the Earths curvature into account when the study area is of regional scale or greater. We present a regularized nonlinear gravity inversion method that has a low computational footprint and employs a spherical Earth approximation. To achieve this, we combine the highly efficient Bott's method with smoothness regularization and a discretization of the anomalous Moho into tesseroids (spherical prisms). The computational efficiency of our method is attained by harnessing the fact that all matrices involved are sparse. The inversion results are controlled by three hyperparameters: the regularization parameter, the anomalous Moho density-contrast, and the reference Moho depth. We estimate the regularization parameter using the method of hold-out cross-validation. Additionally, we estimate the density-contrast and the reference depth using knowledge of the Moho depth at certain points. We apply the proposed method to estimate the Moho depth for the South American continent using satellite gravity data and seismological data. The final Moho model is in accordance with previous gravity-derived models and seismological data. The misfit to the gravity and seismological data is worse in the Andes and best in oceanic areas, central Brazil and Patagonia, and along the Atlantic coast. Similarly to previous results, the model suggests a thinner crust of 30–35 km under the Andean foreland basins. Discrepancies with the seismological data are greatest in the Guyana Shield, the central Solimões and Amazonas Basins, the Paraná Basin, and the Borborema province. These differences suggest the existence of crustal or mantle density anomalies that were unaccounted for during gravity data processing.

Description: Geophysical parameters of the deep Earth's interior can be evaluated through the resonance effects associated with the core and inner-core wobbles on the forced nutations of the Earth's figure axis, as observed by very long baseline interferometry (VLBI), or on the diurnal tidal waves, retrieved from the time-varying surface gravity recorded by superconducting gravimeters (SGs). In this paper, we inverse for the rotational mode parameters from both techniques to retrieve geophysical parameters of the deep Earth. We analyse surface gravity data from 15 SG stations and VLBI delays accumulated over the last 35 yr. We show existing correlations between several basic Earth parameters and then decide to inverse for the rotational modes parameters. We employ a Bayesian inversion based on the Metropolis–Hastings algorithm with a Markov-chain Monte Carlo method. We obtain estimates of the free core nutation resonant period and quality factor that are consistent for both techniques. We also attempt an inversion for the free inner-core nutation (FICN) resonant period from gravity data. The most probable solution gives a period close to the annual prograde term (or S 1 tide). However the 95 per cent confidence interval extends the possible values between roughly 28 and 725 d for gravity, and from 362 to 414 d from nutation data, depending on the prior bounds. The precisions of the estimated long-period nutation and respective small diurnal tidal constituents are hence not accurate enough for a correct determination of the FICN complex frequency.

Description: We evaluate the benefit of different global geophysical loading products on the internal scatter of GPS position time-series from 240 globally distributed sites. We focus on the non-tidal atmospheric pressure loading predicted from NASA's Modern-Era Retrospective Analysis for Research and Applications (MERRA-NATML) and the European Center for Medium-Range Weather Forecasts operational model (ECMWF-NATML), non-tidal ocean loading predicted from the Ocean Model for Circulation and Tides model (OMCT-NTOL), and the continental water storage loading predicted from the MERRA model (MERRA-CWSL) and the GFZ's Land Surface Discharge Model (LSDM-CWSL), respectively. The result shows that the root mean square (RMS) discrepancy of different CWSL models is larger than that of NATML models in the vertical component due to the varying model parameters and input data choices. We discuss the performance of different loading models and their combination to reduce the weighted RMS of GPS up-coordinates. MERRA-NATML & OMCT-NTOL & MERRA-CWSL reduced the weighted RMS (WRMS) in 96 per cent (JPL solutions) and 86 per cent (SOPAC solutions) of the cases, and ECMWF-NATML & OMCT-NTOL & LSDM-CWSL reduced the WRMS in 95 per cent (JPL solutions) and 88 per cent (SOPAC solutions) of the cases. The result shows that local effects and technical uncertainties in GPS time-series hamper the meaningful comparison between GPS observations and mass loading models. Hence, simply using the RMS of the time-series as the assessment criteria may lead to biased comparison results. Nonetheless, we give a detailed comparison (differences in phase and amplitude at seasonal timescales) for eight representative stations located adjacent to great rivers, lakes and reservoirs. We find that LSDM can provide a complementary model to study the small-scale hydrological loading like loading extremes along river channels. However, such small-scale hydrological loading effects are still instable to be modelled in some regions with its current accuracy. Finally, we discuss the impacts of mass loading corrections on the velocity and noise estimates. The noise reductions have the consistent performance as WRMS reductions for most sites, whereas some sites have their noise increased even though GPS signal WRMS is decreased there, suggesting that our posterior correction is potentially feasible, but not sufficient.

Description: Although an amount of hydrocarbon has been discovered in the West Korea Bay Basin (WKBB), located in the North Korean offshore area, geophysical investigations associated with these hydrocarbon reservoirs are not permitted because of the current geopolitical situation. Interpretation of satellite-derived potential field data can be alternatively used to image the 3-D density distribution in the sedimentary basin associated with hydrocarbon deposits. We interpreted the TRIDENT satellite-derived gravity field data to provide detailed insights into the spatial distribution of sedimentary density structures in the WKBB. We used 3-D forward density modelling for the interpretation that incorporated constraints from existing geological and geophysical information. The gravity data interpretation and the 3-D forward modelling showed that there are two modelled areas in the central subbasin that are characterized by very low density structures, with a maximum density of about 2000 kg m –3 , indicating some type of hydrocarbon reservoir. One of the anticipated hydrocarbon reservoirs is located in the southern part of the central subbasin with a volume of about 250 km 3 at a depth of about 3000 m in the Cretaceous/Jurassic layer. The other hydrocarbon reservoir should exist in the northern part of the central subbasin, with an average volume of about 300 km 3 at a depth of about 2500 m.

Description: We develop a high-resolution regional gravity field model by a combination of spherical harmonics, band-limited spherical radial basis functions (SRBFs) and the residual terrain model (RTM) technique. As the main input data set, we employ a dense terrestrial gravity database (3–6 stations km –2 ), which enables gravity field modelling up to very short spatial scales. The approach is based on the remove–compute–restore methodology in which all the parts of the signal that can be modelled are removed prior to the least-squares adjustment in order to smooth the input gravity data. To this end, we utilize degree-2159 spherical harmonic models and the RTM technique using topographic models at 2 arcsec resolution. The residual short-scale gravity signal is modelled via the band-limited Shannon SRBF expanded up to degree 21 600, which corresponds to a spatial resolution of 30 arcsec. The combined model is validated against GNSS/levelling-based height anomalies, independent surface gravity data, deflections of the vertical and terrestrial vertical gravity gradients achieving an accuracy of 2.7 cm, 0.53 mGal, 0.39 arcsec and 279 E in terms of the RMS error, respectively. A key aspect of the combined approach, especially in mountainous areas, is the quality of the RTM. We therefore compare the performance of two RTM techniques within the innermost zone, the tesseroids and the polyhedron. It is shown that the polyhedron-based approach should be preferred in rugged terrain if a high-quality RTM is required. In addition, we deal with the RTM computations at points located below the reference surface of the residual terrain which is known to be a rather delicate issue.

Description: In this paper we present the potential of a new compact superconducting gravimeter (GWR iGrav) designed for groundwater monitoring. At first, 3 yr of continuous gravity data are evaluated and the performance of the instrument is investigated. With repeated absolute gravity measurements using a Micro-g Lacoste FG5, the calibration factor (–894.8 nm s –2 V –1 ) and the long-term drift of this instrument (45 nm s –2 yr –1 ) are estimated for the first time with a high precision and found to be respectively constant and linear for this particular iGrav. The low noise level performance is found similar to those of previous superconducting gravimeters and leads to gravity residuals coherent with local hydrology. The iGrav is located in a fully instrumented hydrogeophysical observatory on the Durzon karstic basin (Larzac plateau, south of France). Rain gauges and a flux tower (evapo-transpiration measurements) are used to evaluate the groundwater mass balance at the local scale. Water mass balance demonstrates that the karst is only capacitive: all the rainwater is temporarily stored in the matrix and fast transfers to the spring through fractures are insignificant in this area. Moreover, the upper part of the karst around the observatory appears to be representative of slow transfer of the whole catchment. Indeed, slow transfer estimated on the site fully supports the low-flow discharge at the only spring which represents all groundwater outflows from the catchment. In the last part of the paper, reservoir models are used to characterize the water transfer and storage processes. Particular highlights are done on the advantages of continuous gravity data (compared to repeated campaigns) and on the importance of local accurate meteorological data to limit misinterpretation of the gravity observations. The results are complementary with previous studies at the basin scale and show a clear potential for continuous gravity time-series assimilation in hydrological simulations, even on heterogeneous karstic systems.

Description: A thorough understanding of time-dependent noise in Global Navigation Satellite System (GNSS) position time-series is necessary for computing uncertainties in any signals found in the data. However, estimation of time-correlated noise is a challenging task and is complicated by the difficulty in separating noise from signal, the features of greatest interest in the time-series. In this paper, we investigate how linear trends affect the estimation of noise in daily GNSS position time-series. We use synthetic time-series to study the relationship between linear trends and estimates of time-correlated noise for the six most commonly cited noise models. We find that the effects of added linear trends, or conversely de-trending, vary depending on the noise model. The commonly adopted model of random walk (RW), flicker noise (FN) and white noise (WN) is the most severely affected by de-trending, with estimates of low-amplitude RW most severely biased. FN plus WN is least affected by adding or removing trends. Non-integer power-law noise estimates are also less affected by de-trending, but are very sensitive to the addition of trend when the spectral index is less than one. We derive an analytical relationship between linear trends and the estimated RW variance for the special case of pure RW noise. Overall, we find that to ascertain the correct noise model for GNSS position time-series and to estimate the correct noise parameters, it is important to have independent constraints on the actual trends in the data.