The zero-difference (ZD) and double difference (DD) approach for processing GNSS observations is mathematically equivalent – meaning that in DD case the huge number of clock parameters are just pre-eliminated. The results for all remaining parameters are identical, if no numerical shortcut is done (e.g., ignoring parts of the correlations introduced by the DD approach). The advantage of processing DD observations is the reduced number of parameters, easier detection and potential correction of cycle slips, and direct access to the integer ambiguities without any phase bias parameter. On the other hand, ZD processing offers greater flexibility in network configuration and parameter handling. This is particularly advantageous when modifying the list of stations in the processing, including Low Earth Orbiting satellites (LEOs), or implementing advanced clock models, e.g., for Galileo satellites.Based on an experimentally developed ZD-based ambiguity resolution method that introduces ambiguity clusters and satellite-wise consistency corrections, the original research prototype was translated into a robust and automated routine processing chain which is suitable for operational use.The new procedure follows the structure of the established CODE DD strategy but adapts the individual processing steps, e.g. pre-processing, estimation of global parameters, handling of receiver-dependent parameters, and ambiguity resolution. Special emphasis is placed on numerical stability, the reliable handling of real-valued ambiguities, and the introduction of quality-control mechanisms designed for long-term autonomous operation. The resulting procedure enables efficient parallelization and delivers consistent orbit, clock, and ambiguity products. We have investigated the requirements on the station density for all these steps in order to optimize also the processing time. Stations, that are not needed in this context can get pre-processed based on the PPP approach and get added to the final solution only.Initial results show that the operational ZD processing chain reaches the accuracy and stability of CODE DD-based products while offering greater flexibility for future extensions. The results demonstrate that the ZD-based GNSS processing is sufficiently mature to generate stable global products on a daily basis and therefore represents a promising foundation for next-generation GNSS solutions computed at CODE.

Abstract At the Astronomical Institute of the University of Bern (AIUB) the Celestial Mechanics Approach for the analysis of satellite orbital motion is currently employed for two space geodetic techniques, namely GNSS and Satellite Laser Ranging (SLR). Both are not only used for station network analyses but also for precise orbit determination (GNSS) and validation (SLR) of Low Earth Orbiting (LEO) satellites. In addition, GNSS-based LEO orbits and SLR observations to the spherical geodetic satellites are used in Earth gravity field recovery. Taking GRACE Follow-On as an example, we present temporal gravity field results, where we derive dynamic and reduced-dynamic orbits together with gravity field parameters from GPS carrier phase observations and combine them with SLR observations to the satellites. Hence, using the satellites as a space tie with respect to the gravity field. We validate our results by comparing them with solutions of superior quality derived from the operational GRACE Follow-On data processing at the AIUB, which include the K-band range-rate data, and we investigate the quality of the co-estimated orbits.

Since mid-2018, the satellite pair of the gravimetry satellite mission called Gravity Recovery And Climate Experiment Follow-on (GRACE-FO) has been providing observations to determine the time-variable Earth’s gravity field with high temporal and spatial resolution. However, some of the low-degree Earth’s gravity field coefficients can better be determined by using Satellite Laser Ranging (SLR) observations to spherical geodetic satellites. Consequently, individual Earth’s gravity field coefficients may be replaced in the GRACE-FO products, or alternatively, the parameter estimation of the Earth’s gravity field can be performed using multi-technique combinations.The Variance Component Estimation (VCE) is a well-established and widely adopted technique in satellite geodesy. As an example, VCE is used for combining observations from different geodetic space techniques or even to or from individual satellites to estimate precise geodetic parameters, e.g., to derive an international terrestrial reference frame.In this study, applications of the VCE method, implemented in a development version of the Bernese GNSS software, are assessed for improved and more automated gravity field parameter determination. The primary focus is on the use of the VCE method for data quality assessments and to strengthen the orbit parametrization. In GRACE-FO data processing, reduced-dynamic orbits, where the usual orbit parameter set is extended with regular Piece-Wise Accelerations (PCAs), are used. In case of SLR data, the strong correlation between the gravity field coefficients and some of the dynamic orbit parameters prevents certain orbit parameters from being estimated but to compensate for this, stochastic pulses are set up. Both PCAs and the stochastic pulses need to be constrained in a way, that they can account for mis-modelings but preserve the sensitivity to the gravity field signal. It is shown that these constraints can be computed directly from VCE rather than determined empirically.

As low-Earth orbit (LEO) satellites are increasingly equipped with highly precise Global Navigation Satellite Systems (GNSS) receivers, new opportunities for GNSS-based precise orbit determination (POD) arise. The availability of GNSS observations from satellites that are geometrically well distributed over the entire Earth may enable the formation of an independent GNSS network in LEO, which can be processed solely on the basis of double-difference (DD) observations. This study presents a first attempt to compute a network solution of GNSS DD observations in space that combines data from eight satellites, including GRACE-FO C/D, Sentinel-3A/B, Sentinel-6A and Swarm A/B/C. Using satellite laser ranging (SLR) validation, we show that the absolute position accuracy of an ambiguity-float network solution increases from about 4 cm 1D RMS in a constellation of three satellites to about 2 cm–2.5 cm in a constellation of eight satellites. Similar results are obtained by comparing the baseline lengths from our DD network solution with those derived from external ambiguity-fixed zero-difference (ZD) orbits, with differences at a level of 2 cm to 3 cm RMS for a network of eight satellites. For both the SLR residuals and the baseline consistency, an improvement of up to 0.5 cm can be achieved when resolving 60

The European Space Agency’s BepiColombo mission continues its pioneering voyage to Mercury, the innermost planet of the Solar System. Among the advanced instruments onboard the Mercury Planetary Orbiter (MPO), one of the two spacecraft that comprise the BepiColombo mission, is the Italian Spring Accelerometer. The instrument’s primary scientific goals are closely linked to the Mercury Orbiter Radio-Science Experiment. Together, these instruments aim to provide valuable data on the spacecraft’s orbit, as well as Mercury’s gravity field and internal structure. This simulation study examines how accelerometer measurements affect orbit and gravity field recovery, and it explores strategies to overcome the challenges associated with the co-estimation of all parameters. In it, we evaluate the accuracy of the integrated retrieval of spacecraft orbit, gravity field, and accelerometer parameters under different noise levels and varying observation geometries during the mission. To achieve these goals, we propagate the orbit of MPO and simulate Doppler observations and accelerometer measurements based on the available noise models. We consider two scenarios by including either a realistic or a more conservative accelerometer noise level. Our results indicate that postponing the estimation of accelerometer biases until a preliminary gravity field is established helps prevent gravity field mismodelings from being absorbed into the accelerometer parameters. The daily estimation of bias was found to be essential. Using one year of Doppler tracking data and applying Kaula regularization, we carry out orbit determination and gravity field recovery under varying assumptions regarding the accelerometer noise and the observation geometry. Improved recovery of low-degree coefficients was observed under realistic noise assumptions. Moreover, the strong influence of observation geometry in shaping the error behavior was observed. namely, errors in cross-track bias estimation increased sharply when the β-Earth angle dropped below 45°, corresponding to the degraded Doppler observability periods. Orbit determination achieved decimeters-level accuracy in the radial direction and metre-level accuracy in the along- and cross-track directions, with a substantial rise in cross-track orbit errors observed during periods of reduced Doppler observability. This study also demonstrates the applicability of the planetary extension of the Bernese GNSS Software as an independent tool for orbit determination and gravity recovery for the BepiColombo mission.

Accurate uncertainty information associated with essential climate variables (ECVs) is crucial for reliable climate modeling and understanding the spatiotemporal evolution of the Earth system. Recent developments in deep learning have remarkably advanced the estimation of ECVs with improved accuracy. However, the quantification of uncertainties associated with outputs of such deep learning models has yet to be widely adopted. This survey explores the types of uncertainties associated with ECVs derived from deep learning methods, including aleatoric (data) and epistemic (model) uncertainty, and the techniques to quantify them. The focus is on highlighting the importance of considering uncertainty associated with inputs in the deep learning models to account for the dynamic and multifaceted nature of satellite observations. The survey starts by clarifying the definitions of aleatoric and epistemic uncertainties and their roles in a typical satellite observation processing workflow, followed by bridging the gap between conventional statistical and deep learning views on uncertainties. Then, we comprehensively review the existing uncertainty quantification methods for deep learning algorithms and discuss their strengths and limitations. A comprehensive literature review about quantifying uncertainties in the deep learning estimates of ECVs follows the theoretical survey, covering a wide range of ECVs. The specific need for modification to fit the requirements from both the Earth observation side and the deep learning side in such interdisciplinary tasks is highlighted. We further demonstrate our findings with two selected ECV examples, snow cover and terrestrial water storage, to provide clear insights into different methods by promoting quantitative comparison. In the end, we summarize our findings and provide perspectives for future research.

Determining long-term changes in global ice and water storage from satellite gravimetry remains challenging due to the limited temporal coverage of high-resolution missions. Here, we combine Satellite Laser Ranging (SLR) and Gravity Recovery and Climate Experiment (GRACE) data to reconstruct large-scale, non-linear mass variations from 1995 to 2024, extending gravity-based observations into the pre-GRACE era while preserving spatial detail through backward extrapolation. The combined model reveals widespread and statistically significant accelerations in global water and ice mass changes and enables the identification of key turning points in their temporal evolution. Results indicate that in Svalbard, a non-linear transition in ice mass balance occurred in late 2004, followed by a pronounced acceleration of mass loss due to climate warming. Glaciers in the Gulf of Alaska exhibit persistent mass loss with a marked intensification after 2012, while in the Antarctic Peninsula, ice mass loss substantially slowed and a potential trend reversal emerged around 2021. The reconstructed mass anomalies show strong consistency with independent satellite altimetry and climate indicators, including a clear response to the 1997/1998 El Niño event prior to the GRACE mission. These findings demonstrate that integrating SLR with GRACE enables robust detection of non-linear, climate-driven mass redistribution on a global scale and provides a physically consistent extension of satellite gravimetry records beyond the GRACE era.

Abstract Satellite Laser Ranging (SLR) measures the roundtrip time-of-flight of short laser pulses between SLR ground station telescopes and satellites equipped with retro-reflector arrays. Over the years the technology has evolved towards higher repetition rates, which requires either faster moving mechanics or separate optical systems for the transmit and receive path. The approach presented in this article, however, does neither require fast moving mechanics nor separate optical systems. The Swiss Optical Ground Station (SwissOGS) Zimmerwald has carried out a major upgrade in 2025 and is now operating a 1 kHz SLR system. A special feature of the new system is that the same transmit and receive path is used. When transmitting and receiving in the telescope Coudé path, it is necessary to prevent the outgoing laser pulse from directly hitting the detector. Another well-known problem is the so-called afterglow, which results from intensely laser-illuminated optics. Many SLR stations contribute to the Global Geodetic Observing System (GGOS). Higher repetition rates are beneficial for improving the quality of the SLR normal point generation. But higher repetition rates require upgrades to the processing software and hardware, e.g. the laser. On the hardware side, transmit/receive switching techniques such as the mechanical rotating shutter are critical elements, as their rotational velocity cannot easily be increased. As in the early days of SLR technology, one option for implementation is to piggyback the laser on the telescope tube, with the disadvantage of a potentially varying alignment. Another option for implementation is to use a perforated mirror which has solved this problem at the Zimmerwald station for years. Our new technique also takes afterglow into account. In this article, we present the implementation of a kHz transmission and reception in the telescope Coudé path. Although the SLR systems share common components such as a telescope, laser, detector, timer, etc., the implementations at the SLR stations differ in the details. The presented solution could be implemented at other stations although they may differ in certain details. At Zimmerwald, 1 kHz is currently used for satellite tracking without any mechanical rotating parts tied to the high repetition rate. The laser and detector are still located in the telescope Coudé.

In recent years, the use of non-scientific low Earth orbiting (LEO) satellites for gravity field determination has been increasingly explored. In principle, the same methods as for non-dedicated scientific missions are applied by analysing the satellites’ orbital perturbations to gain information about the Earth’s gravity field. With currently over 100 CubeSats in orbit, the Spire commercial constellation provides a huge amount of GNSS tracking data in the same timeframe as scientific missions. This can be exploited to increase the spatio-temporal resolution of estimated gravity field solutions. Although the data quality is limited, previous analyses using data from a 2020 ESA Announcement of Opportunity project have shown that a combined processing of Spire CubeSats can achieve monthly gravity field solutions of similar quality as non-dedicated scientific missions. In our work, we make use of both the Spire data from 2020 and a new dataset from 2023 provided by EUMETSAT. At the Astronomical Institute of the University of Bern, non-gravitational forces acting on the satellite have so far not been explicitly modelled when determining the gravity field, but instead absorbed by pseudo-stochastic parameters such as piecewise constant accelerations. As these forces are prominent for CubeSats due to their low altitude and their large area-to-mass ratio, this study explores how the estimation of the orbits and gravity field solutions improves when explicitly modelling these forces. To achieve this, we determine kinematic orbit positions from Spire CubeSat GNSS phase and code data and introduce them as pseudo-observations in an orbit and gravity field recovery step, where dedicated non-gravitational force modelling is applied using macro-model information provided by Spire. The combination of the different Spire satellites is performed at the normal equation level using a least-squares approach.

MESSENGER’s Mercury Laser Altimeter (MLA) provided high-resolution topographic models and geodetic insights into Mercury’s internal structure, though its data was limited to the northern hemisphere. Due to its less elliptical orbit, the BepiColombo Laser Altimeter (BELA) will soon be delivering global altimetry coverage. We present expected improvements on our knowledge of Mercury’s geodetic parameters, such as its obliquity, short- and long-period librations, rotation rate, and the amplitude of tidal deformations, by comparing crossover analyses of the MLA and BELA altimetry datasets within consistent closed-loop simulations. We discuss the potential of cross-dataset crossovers (consisting of a BELA track crossing with an MLA track) to refine the determination of key geodetic parameters by providing extended coverage at lower latitudes. Our simulations incorporate realistic altimetric ranges, orbital dynamics, and perturbations due to the imperfect knowledge of, e.g., MESSENGER orbit and attitude, and of Mercury’s orientation. We show the improvement of crossover-derived uncertainties of the geodetic parameters, calibrated using variance component estimation. In particular, rotation parameters, including long-period forced librations, significantly benefit from combining inter-instrument crossovers. These uncertainties are finally mapped through Markov Chain Monte Carlo to study BELA’s impact on constraining plausible models for Mercury’s internal structure.

Coronal mass ejections (CMEs) from the Sun can cause geomagnetic storms which cause the thermosphere to expand. This leads to enhanced air drag for satellites in low Earth orbit (LEO). This work focuses on the evaluation of orbital decay with a focus on selected geomagnetic storm events. Using the Bernese GNSS Software (BSW), reduced-dynamic orbits of different LEO satellites are computed from GNSS data of on-board receivers, where non-gravitational accelerations are modelled by means of estimated empirical piecewise-constant accelerations (PCAs). The orbital decay is then calculated by using the PCAs, or, in case of the GRACE Follow-On satellite, calibrated accelerometer data, to solve Gauss’s perturbation equation for the satellite’s semi-major axis. This method is compared to an approach where a fit model is applied to the osculating semi-major axis derived from the reduced-dynamic orbits computed by BSW. The fit model consists of a piece-wise linear model of the time-varying mean orbital decay and the time-varying amplitudes of the most dominant periodic oscillations. The results of both methods are compared and found to be similar for large orbital decays induced by CMEs. But the fit model struggles with low orbital decay. The Gaussian perturbation equation approach is far more precise than the fit model and can react, e.g., instantaneously to satellite maneuvers which change the semi-major axis. For satellites without an on-board accelerometer, PCAs can be estimated and used for the numerical integration.

An accurate knowledge of the orientation, the tidal deformability, and the gravity field of a celestial body is fundamental to provide constraints on its internal structure. These quantities may be retrieved by processing radiometric tracking and altimetry data from a probe in orbit around such body. This paper presents a method to combine altimetry crossovers with two-way Doppler tracking observations at normal equation level, using the Bernese GNSS Software and the pyXover software library. This method was applied to a proposed 200km altitude orbiter around Callisto, a privileged destination for the upcoming phase of Solar System exploration. Enhancing “standard” Doppler tracking with altimetry generally benefited both orbit determination and a joint estimation of the orientation of the north pole and of planetary librations. The retrieval of low-degree gravity field parameters was also improved by the addition of altimetry data. However, the improvements on the estimated parameters were highly dependent on the characteristics of the simulation, e.g., the underlying topography roughness. Overall, combining radioscience with altimetry data accounted for a visible reduction of correlations among estimated parameters, while also allowing for a consistent estimation of the “vertical” Love number h2 along with gravity.

Laser ranging to spherical satellites is a major source for the determination of geophysical and geometric parameters like the scale of the reference system and the lowest degree (1-2) gravity field coefficients, i.e. geocenter motion, dynamic oblateness and the orientation of the rotation axis of the Earth. The International Laser Ranging Service (ILRS) is collecting Satellite Laser Ranging (SLR) observations of a global station network, providing the range data as normal points to its analysis centers, which perform precise orbit determination (POD) and network solutions based on 7 day orbital arcs. Currently, efforts are undertaken to extend the classical ILRS processing of the LAGEOS and ETALON satellites by LARES 2, as well as lower-flying satellites, i.e. LARES, Stella and Starlette, necessitating an adaption of the SLR-POD model and parametrization due to the increased sensitivity to orbit perturbations at low orbit altitudes. We present the status of the SLR-POD at AIUB, where in support of the ILRS analysis center at BKG the incorporation of the low-flying SLR satellites into the 7 day POD and network solution of LAGEOS/ETALON/LARES-2 is beeing tested, making use of long-arc stacking techniques of daily arcs to continuous 7 day arcs. All orbit and geophysical parametes are estimated in one common estimation process to avoid implicit regularization by apriori information. Special attention is paid to the co-estimation of the low-degree gravity field coefficients and the correlations with the empirical dynamic parameters deployed in the classical LAGEOS/ETALON POD model.

The ESA GENESIS mission, which obtained green light at ESA's Council Meeting at Ministerial Level in November 2022 and which is expected to be launched in 2027, aims to significantly enhance the accuracy and stability of the Terrestrial Reference Frame (TRF). This shall be achieved by equipping one satellite at approximately 6000 km altitude with well-calibrated instruments for all four space-geodetic techniques contributing to TRF realizations, i.e., Global Navigation Satellite Systems (GNSS), Satellite Laser Ranging (SLR), Very Long Baseline Interferometry (VLBI) and Doppler Orbitography and Radiopositioning Integrated by Satellite (DORIS), and by exploiting the such realized very precise space collocations. The GENESIS satellite is foreseen to carry a zenith- and nadir-pointing GNSS antenna to track (at least) GPS and Galileo signals. Because of its very high altitude, GENESIS will cover much larger nadir angles as seen from the GNSS transmitting antennas, compared to receivers at ground or in low Earth orbit. This will partly result in GNSS observations with lower signal-to-noise ratios. Furthermore, to date, only little reliable information is available on the GNSS transmit antenna gain and carrier phase patterns at very large nadir angles. These problems lead to questions with respect to the best possible exploitation of GNSS data by GENESIS, e.g., whether phase pattern calibrations will need to be performed by means of tracked GNSS data (which would weaken the GNSS contribution to GENESIS). The aim of this study is to assess the impact of GNSS transmit antenna phase pattern errors on the GNSS-based POD of GENESIS, as well as global GNSS network solutions for GNSS orbits and clock corrections, Earth rotation and geocenter parameters and station coordinates based on GNSS observations from GENESIS and terrestrial stations. To accomplish this, we employ simulated GNSS pseudo-range and carrier phase data for GENESIS and ground stations, which have been generated based on detailed link-budget computations and a comprehensive set of transmit antenna gain patterns. The data are used for closed-loop simulation investigations, which allow to compare the reconstructed orbit and geodetic parameter solutions to the simulation truth and offer a quantification of the impact of transmit antenna phase pattern uncertainties on the estimated parameters.

Satellite Laser Ranging (SLR) observations are provided by a global station network of the International Laser Ranging Service (ILRS). Since almost all SLR stations are unique in terms of utilized equipment, e.g., laser system, photo-detector or timing devices, their measurement performances may slightly differ. Furthermore, the tracked satellites have various properties, e.g., material composition (number and type of retro-reflectors or mantel material), diameter, area-to-mass ratio, altitude or inclination, which have an impact on the measurement precision and the requirements on the background force modeling. A reliable estimation of geodetic parameters solely based on SLR can only be performed by combining SLR observations to several spherical satellites provided by the entire ILRS network. To take the quality of each individual SLR measurement into account, several stochastic models, e.g., static or time-variable station- and/or satellitespecific weights, are introduced and their impact on the parameter estimation is studied.

The European Space Agency's BepiColombo mission continues its pioneering voyage to Mercury, the innermost planet of the Solar System. Among the advanced instruments onboard the Mercury Planetary Orbiter (MPO) is the Italian Spring Accelerometer (ISA), whose scientific objectives are closely linked to the Mercury Orbiter Radio-Science Experiment (MORE). Together, these instruments aim to provide high-precision data on the spacecraft's orbit, as well as Mercury's gravity field and internal structure. This simulation study investigates how the modeling and parametrization of accelerometer measurements influence orbit and gravity field recovery and explores strategies to overcome the challenges associated with the co-estimation of all parameters. We assess the integrated retrieval accuracy of the spacecraft orbit, gravity field, and accelerometer parameters under different noise levels and varying observation geometries during the mission. The MPO orbit is propagated, and Doppler and accelerometer measurements are simulated based on available noise models. Our results indicate that postponing the estimation of accelerometer biases until a preliminary gravity field is established helps prevent gravity field mismodelings from being absorbed into the accelerometer parameters. The daily estimation of bias was found to be essential. Using one year of Doppler tracking data and applying Kaula regularization, we demonstrate the potential to recover Mercury's gravity field up to degree and order 40. Low-degree coefficients improve under more optimistic noise conditions. Errors in cross-track bias estimation rise sharply when the beta-Earth angle falls below 45°, corresponding to the degraded Doppler observability. Orbit determination achieved centimetre-level radial accuracy and metre-level along- and cross-track accuracy, with increased cross-track errors during low-observability periods.

A growing number of Low Earth Orbiting (LEO) satellites are collecting GNSS tracking data that allows to recover the long-wavelength part of the Earth’s time-variable gravity field. Besides scientific LEO missions, commercial satellite constellations consisting of a huge number of nano-satellites are moving into focus. Due to an improved ground track coverage, such constellations offer the opportunity to increase the spatio-temporal resolution of derived gravity field models and can contribute to reduce temporal aliasing errors of dedicated gravity field missions. The Spire constellation is of particular interest as it consists of more than 100 nano-satellites (standardized CubeSats), all equipped with high-quality GNSS receivers. Furthermore, the Spire constellation offers a variety of orbital characteristics with different inclinations at altitudes of about 400–650 km. In this study, we use GNSS data from nine Spire CubeSats to derive monthly gravity field solutions covering a six-month period. The orbit and gravity field recovery is performed with the Bernese GNSS Software, which applies the Celestial Mechanics Approach. We demonstrate that the 2–3 times larger noise level of the Spire GNSS observations affects the quality of the retrieved gravity field solutions in the same order of magnitude. Therefore, a single Spire CubeSat solution cannot compete with those obtained from scientific LEO missions. However, with an increasing number of CubeSats, the performance improves so that a combination based on data from all nine Spire CubeSats can achieve a quality level comparable to a solution derived from ESA’s Swarm-B satellite.

We study temporal gravity field determination from GRACE Follow-On satellite-to-satellite tracking data using the inter-satellite link of the K-Band Ranging System (KBR) and kinematic positions of the satellites as observations and pseudo-observations, respectively. In addition, we introduce the range-rate observations collected by the Laser Ranging System (LRI) as further observation group. We compute our solutions with the Celestial Mechanics Approach using next to the kinematic positions either only LRI range-rate observations or combining them with the KBR derived range-rate data by the means of Variance Component Estimation (VCE).The stochastic noise of the KBR and LRI data is modelled with an empirical description of the noise based on the post-fit residuals between the final GRACE Follow-On orbits, that are co-estimated together with the gravity field, and the observations, expressed in position residuals to the kinematic positions and in KBR and/or LRI range-rate residuals. We validate the LRI-only and KBR+LRI monthly solutions by comparing them with the KBR-only derived models from the operational GRACE Follow-On processing at the AIUB, by examining the stochastic behaviour of respective post-fit residuals and by inspecting areas where a low noise is expected. Last but not least, we investigate the influence of LRI-only and KBR+LRI monthly gravity field solutions in a combination of monthly gravity fields based on other approaches as it is done by the Combination Service for Time-variable Gravity fields (COST-G) and make use of noise and signal assessment applying the quality control tools routinely used in the frame of COST-G.

The Combination Service for Time-variable Gravity fields (COST-G) of the IAG looks back at an eventful and very successful year. The operational combination of the monthly GRACE-FO gravity fields now comprises eight analysis centers, providing high-quality solutions with short latency on a regular basis. When the new release 06.3 of the GRACE-FO Science Data System (SDS) time-series became available in September 2024, COST-G generated test combinations and could confirm the quality gain compared to the former release 06.2. Meanwhile, release 06.3 is routinely incorporated in the operational combination.The number of analysis centers providing complete time-series of monthly gravity fields of the GRACE mission to COST-G has more than doubled compared to the original COST-G GRACE RL01, published in 2019. The current COST-G GRACE RL02 is aweighted combination of 11 time-series, where the weighting scheme was adapted to be consistent with the operational GRACE-FO combination. The quality gain of the new combination is most pronounced during the early and late GRACE mission period, when data quality issues and environmental conditions were challenging.

Within the IGS, it was agreed that Precise Point Positioning (PPP) based on satellite orbit and clock corrections of the IGS analysis centers allow a direct access to the IGS realization of the International Terrestrial Reference Frame (ITRF). This convention is considered convenient for all PPP users and should not be changed in future. On the other hand, the groups determining the GNSS satellite orbits do prefer an origin of the frame that is related to the Earth instantaneous center of mass since this is the reference of the gravitational orbit force model. With this background, some discussions take place to change the convention related to the origin of the terrestrial frame because it is convenient for the orbit determination. This convention is, however, in contradiction to the expected needs of the PPP users that do prefer a stable coordinate origin in time. We will introduce a strategy to serve both needs by applying the center of mass corrections for the orbit determination only.