In 2011 a novel strategy for GPS raw observations processing was proposed in GPS seismology. This strategy, called variometric approach, was proven to be able to supply, in real time, three dimensional velocities, in the global reference frame, of a GPS receiver during an earthquake. The approach is based on the time single-differences of carrier phase observations collected at high-rate (1 Hz or more) by a stand-alone receiver and on the standard broadcast products (orbits, clocks). The approach was implemented for the first time in the VADASE (Variometric Approach for Displacements Analysis Stand-alone Engine) software demonstrating that the estimated velocities can be integrated over a short time period (when the earthquake occurs) in order to reconstruct three dimensional waveforms and coseismic displacements.
VADASE was applied for the first time to data not yet processed with other established approaches (i.e. Precise Point Positioning and Differential Positioning, both not able to provide a real-time stand-alone solution) supplying the first computation of the GPS displacement waveforms of March 11, 2011 Tohoku-oki (Japan) earthquake, using IGS high-rate data (1 Hz) recorded at MIZU (140 Km from earthquake) and USUD (430 km from earthquake) stations.
In the following years many earthquakes were analyzed using the variometric approach and significant refinements and different implementations of the model have been proposed by several different researchers also considering the comparison and integration with other sensors. In particular, the multi-constellation capability of the approach was developed enhancing the GALILEO signals leverage and the single-frequency application for low-cost receivers was consolidated.
On an industrial point of view, the variometric approach option, embedded in a GNSS receiver firmware, was presented for the first time in 2015.
In 2017 the variometric approach extended its application outside the GNSS seismology. A first feasibility demonstration of the kinematic application of VADASE was given in order to contribute to GNSS precise real time navigation. In the last two years, different authors started to consider the direct estimation of the 3D velocities provided by the variometric approach a significant contribution and aid to kinematic positioning models.
Finally, in the last year, after the new availability provided by the Android framework, the effectiveness of the approach was also proven on raw observations collected by mass-market Android devices paving the way for still not investigated mass-market application.

Mining exploitation significantly changes stress–strain conditions. Stress concentration may induce strong seismicity, therefore, reliable information on temporal elastic variations around the exploitation area is of crucial interest for hazard assessment and thus preventing damages in mines and protecting mine crews. Nowadays most of mines in Poland established classic seismometer networks in order to observe seismic activity during the exploitation. These data are used for studying distributions of seismic events and also for seismic hazard assessment. On the other hand, this method has several main drawbacks: the data are not temporally continuous and do not allow to precisely determine geometric metrics. To fill these gaps, we propose to employ multi-constellation GNSS-based seismography.
In this contribution we present the results of the project supported by ESA and devoted to development and application of advanced high-rate and multi-GNSS system to seismic analysis at the area of copper open pit mine in Poland. The system is based on the network of continuous operating receivers, central processing facility, which automatically process multi-station, high-rate GPS+Galileo data in a relative mode and web portal responsible for visualization of the station displacement results. The initial validation revealed that the system is capable to provide the continuous displacements of the monitored stations with the temporal resolution up to 50 Hz and sub-centimeter level precision. Detailed analyses showed that results comparable to multi-constellation may be obtained with the application of only Galileo observations. Hence, Galileo signals have proved their high usability in most demanding applications.

Since more than a century of geophysical and tectonic observations, studies give us a partial view of the dynamics of the Solid Earth, which most of the scientific community today interprets in the light of plate tectonics. However, many questions are still open. One of the most intriguing matter concerns the causes and the driving mechanisms of plate motions, which are a fundamental problem for the understanding of all the geological phenomena we observe on our planet. For years it has been preferred to leave out some inconsistencies between theoretical models and experimental data, among these the “westerly” polarization of the movement of the plates with respect to the mantle and the asymmetry of subduction zones remains completely unexplained. Lunisolar tides have often been invoked to trigger plate tectonics and earthquakes, but not straightforward correlation has ever been proven. In recent decades models of astronomically tuned tectonics have been developed, giving rise to a lively debate in the scientific community. In this work an analysis of the relative motions of the main plates will be performed, in search of one possible lunisolar tidal modulation. In particular we plan to perform both short and long period analysis of baselines bridging two tectonic plates measured by geodetic techniques: GNSS, Satellite Laser Ranging (SLR) and Very Long Baseline Interferometry (VLBI) data will be analyzed in order to check whether such lunisolar tidal modulation is really present. Long period harmonic analysis now is made feasible since we now have a lot of geodetic stations all over the world for which data time series span up to 20-30 years. Moreover, data of plate motions recorded by GPS stations are routinely filtered by the theoretical tide. The goal of our analysis will focus on cross-correlating different frequencies of the lunisolar tidal modulation, tidally unfiltered space geodesy data and seismicity.

Today's best practice earthquake early warning systems (EEW) are based on strong motion sensors recording accelerations over a broad frequency range. These instruments are capable of measuring translational motions of large earthquakes. Due to current technical limitations, they are capable of saturation at > 2 g and are insensitive to very low frequencies or permanent displacements. The fast development of GNSS high-precision positioning over the last years allows it nowadays to use a second type of instrumentation for ground motion monitoring. The progress in GNSS towards high sampling rates and increasing precision opens up new possibilities to support strong motion sensors in the lower frequency range and for very large dynamic and permanent displacements. The development of an optimal combination method requires an in-depth understanding of the quality of the recordings during motions with varying frequencies, amplitudes and static displacements. It also requires a verification of the combination results with realistic experiments. We use an industrial six-axis robot arm to perform experiments to validate and study the optimal frequency and amplitude range of each instrument. Both instruments were mounted on the arm simultaneously. We achieved a high pose repeatability of the robot of ±0.03 mm by careful calibrating the robot with our own methodology and instrumentation and improved the quality of motion, so that realistic earthquake motions could be executed. The feedback loop of the robot is responsible for its precise movement. It records the performed trajectory and is used for motion corrections. We used it as the robot ground truth to verify the instrument recordings and the combination thereof.
We were able to show that the strong motion sensor has considerable problems to resolve low frequencies due to a recurrent overestimation of the amplitudes. The GNSS sensor mounted simultaneously on the robot arm, however, shows great potential in complementing the strong motion sensor for slow and long-period motions. When motions with comparably high frequencies of up to 10 Hz were performed, the high-rate GNSS data showed significant biases regarding the accurate reconstruction of phase shifts and amplitudes. The optimum combination using a Kalman filter has, therefore, to carefully take into account these different sensor behaviors.