While orbit determination for space debris and cooperative satellites is usually based on radar and lidar systems, whose data are integrated by the North American Aerospace Defense Command (NORAD), optical observations could be exploited to offer a more precise ground-based orbit determination. The common observational methodologies require the target to be in sunlight, while the observation site is in darkness, and they present an obvious requirement for acceptable weather conditions. In order to eliminate these dependencies and to acquire a higher amount of data, multiple observatory sites can be used to significantly increase the coverage over the celestial vault and to decrease the time between couples of passages above an optical ground station. The Sapienza Space Systems and Space Laboratory (S5Lab) Research team at Sapienza University of Rome is managing a network of several observatories in Italy, as well as two optical observatories at low latitudes (located in Malindi, Kenya), for the orbit determination of non-cooperative targets.

Moreover, a fine time synchronization among all the utilized observational sites allow to achieve a higher precision in the orbit determination, as well as permitting an easier definition of the target height and distance from the observatories. Indeed, with a limited amount of observational data, it is possible to intersect the acquired celestial coordinates of the observable with its known orbital plane to consequently evaluate the object height and distance from the observatory. If the observatories network presents a precise time synchronization between multiple observatories, it is possible to evaluate the object position while integrating data from multiple observatories at a time, with a determination precision improvement that depends on the ground distance between the currently used observational sites, also defined as baseline. Data from the aforementioned observatory network has been utilized for the determination of the height of in-orbit objects by exploiting the first exposed methodology. Meanwhile, a higher time synchronization is being achieved by the S5Lab observatory network in order to allow a more precise height and distance determination by using multiple observatories data.

On the other hand, the implementation of Light Emitting Diodes (LEDs) on-board Low Earth Orbit satellites is able to enhance the orbit determination precision and to allow for an increase of the acquirable data. Indeed, an autonomously self-illuminating system, such as LED-based boards to be installed on the satellite external surface, would allow to completely eliminate the traditional constraint on lighting conditions related to optical observations, i.e. making the satellite observable during the whole eclipse time. The optimization of the flashing patterns related to orbit determination would ease the identification of the spacecraft over the background stellar field and improve the efficiency of the determination process. Indeed, the orbit determination would be performed by exploiting the already mature astrometry software, if the spacecraft is able to perform short flashes that would result as dot-like trails in the images acquired by the telescopes. On this topic, the S5Lab team has conceived, in collaboration with the University of Michigan, the LEDSAT (LED-based small SATellite) CubeSat mission. LEDSAT is a 1-Unit CubeSat, currently part of the European Space Agency (ESA) Fly Your Satellite! Programme, and included in the Italian Space Agency (ASI) IKUNS Programme. The satellite is currently under development at S5Lab and it will be deployed from the International Space Station (ISS) in 2020. The satellite will test the effectiveness of a LED-based payload for the verification and improvement of the current methodologies of orbit determination based on optical data.

All the mentioned methodologies are applicable to Precise Orbit Determination (POD) tasks for Earth Observation satellites (EO). While the non-cooperative observations remain constrained to the lighting conditions requirements for optical observation, they would be applicable to all the already existing spacecraft addressed at EO, without the need of any cooperation between the satellite and the observational site. On the other hand, the confirmation of the applicability of the LED-based technology for LEO spacecraft would allow a faster and easier orbit determination at a relatively low cost, in terms of mass, consumed power, surface and volume. In particular, it would be possible to implement LED boards on-board all spacecraft, including EO satellites, with a significant improvement of the optical orbit determination acquirable data amount and precision.

This work will deal with the applicability of optical observations to POD for EO satellites. After the description of the up-to-date techniques for non-cooperative optical orbit determination, the S5Lab network and the major completed tasks will be presented. Then, the current improvements in the field of fast range and altitude determination based on the integration of optical data will be described. The expected improvements of optical orbit determination when applying the LED-based technology on spacecraft will also be presented, with an overall description of the LEDSAT mission. Finally, the future perspectives will be presented, with particular regards to the orbit determination of EO satellites.

Area B - Board 163

Seven years from 2008 to 2014 of Doppler Orbitography and Radiopositioning Integrated by Satellite (DORIS) and Satellite Laser Ranging (SLR) observations to the satellites JASON-1, JASON-2, and ENVISAT are evaluated in a Precise Orbit Determination (POD) process. The DORIS residuals are sampled in 3x3 degree bins of azimuth and elevation seen from the ground stations. The resulting stochastic and systematic patterns show specific characteristics for each mission. The stochastic patterns are then considered in the POD analysis in form of elevation dependent weighting, the systematic deviations in form of phase center correction masks. We present the results of the effects on orbit quality in terms of orbital fits and differences in orbit positions. We finally solve also for station positions and velocities, and for Earth Rotation Parameters (ERPs). The effects of the phase center variations are assessed in terms of improvement of the formal errors of the estimated coordinates, velocities and ERPs. We also look into the defining parameters of the terrestrial reference frame where origin and scale are of particular interest.

Area B - Board 164

Regularly involved in Precise Orbit Determination for altimetry missions (Jason, Sentinel-3, not forgetting the pioneers TOPEX-Poseidon, CryoSat and Envisat), the DORIS system will also participate to future missions Sentinel-3C/D, Jason-CS/Sentinel-6, SWOT and HY-2C/D.
With these missions, the system and its beacon network are bound to be operational until at least year 2033.
In the 15 years mean time, improvements are still going on to make Orbit Determination always more precise :

New generation beacons are currently being designed and validated. The third generation beacons, developed in 2000-2010, have been operational and efficient for almost 20 years now. Sixty of them were spread all around the world hosted by 35 countries . The fourth generation will progressively replace them in the next years. Long coordinates series of DORIS stations are useful for space geodesy throuh decades.

A recent version of the flight software was recently uploaded on-board Sentinel-3B (and A) producing real-time estimations at the 2-3cm level on the Radial component. Near-Real Time altimetry can count on this accurate and reliable on-board product (not a 10 second unavailabilty on-board Jason-2 since 2008). Hosted by current DORIS receivers provided by Thales, these software are also able to provide on-board altimeters with accurate altitude information to preprogram altimeter tracking loops.
French Earth Observation Pleiades mission has highly contributed to DORIS permanent improvement by bringing a strong robustness to the on-board software series, and to the DORIS receivers.

In parallel, on-ground, an efficient integrity team accurately checks each day the telemetry of the DORIS system, ensuring DORIS users that measurements are accurate. This good health and quality control is very important to bring scientists with accurate and reliable data.

Last but not least, the new POE-F standards have significantly improved the accuracy of DORIS-only orbits (1.2cm RMS SLR residuals all elevations and 8mm RMS at high elevations) for Saral/AltiKa and Sentinel-3A/B. These orbits are necessary to precisely process the altimeter data, and thus monitor the sea level with subcentimetric accuracy, revealing future trends in it.

As an alternate future possibility, a new kind of twin GNSS-DORIS receiver is currently being studied to provide real-time orbits with an interesting accuracy (below 1m, down to a few centimeters) within much reduced weight and cost.

Renewed with all these features, the DORIS system is ready for intensive use of a wide panel of new user satellites in the next two decades at least. Permanently compared with GNSS, SLR and VLBI data, its products will be valuable for GGOS goals : monitor the Earth (both liquid and solid) with a millimeter accuracy, in line with a United Nations recommandation.

Area B - Board 165

A Precise Orbit Determination (POD) status for DORIS satellites, in particular for the two Sentinel-3 and Jason (2&3) missions, is presented. We now use body and solar array quaternions for Jason-2 and 3 satellites. We will also use the last standards and IDS recommendations given for the next ITRF realization. We will give statistical results such as one per revolution empirical acceleration amplitudes and orbit residuals. We will also give some comparisons to the CNES precise orbit used for altimetry and to GPS-only orbits contributing to the Copernicus POD Quality Working Group of Sentinel. Some external validations of our orbits will be done, as independent SLR measurements processing as well the use of altimeter crossovers.

Area B - Board 166

The key factor for the reliable Earth observation using space techniques is the precise positioning of the satellites. If one wanted to observe the phenomena changes characterized by the cm-level amplitude, the precision of the determined orbit should be at the order of magnitude better. Orbits of the low Earth orbit (LEO) missions devoted to the Earth observation usually are determined using GPS and DORIS data with the support of Satellite Laser Ranging (SLR) and other navigation devices. The European navigation satellite system Galileo developed by the European Space Agency (ESA) has reached operational capability for initial services in late 2016. The current Galileo constellation comprises 26 satellites which allow for the precise positioning of the ground stations. Precise Galileo orbits may be useful for precise orbit determination of LEO satellites if the future Earth-observing missions were equipped with the Galileo signal receiver.

In this work, we present the result for the precise orbit determination of the Galileo satellites. We mainly focus on the modeling of the direct solar radiation pressure (SRP) – the greater non-gravitational force which perturbs the Global Navigation Satellite System orbits. For the coping with the direct SRP we test several approaches (1) using purely empirical orbit modeling, (2) using physical models based on the geometrical and optical properties of the Galileo satellites, and (3) the hybrid solution which is based on the physical parameters with the estimation of different sets of the empirical parameters. We present the pros and cons of particular solutions in terms of the reliability of the orbit determination including SLR validation, day boundary discontinuities, and orbit predictions.
The results of the precise Galileo orbit determination are validated using SLR residual analysis which indicated that the best solution is based on the hybrid models achieving the RMS of the SLR residuals at the level of 2.5 cm as compared to 2.7 and 2.9 cm for the solutions based on the empirical and physical models, respectively. We checked also the stability of the orbit predictions with the result at the level of 3 meters after 48 hours.

The satisfying quality of the Galileo orbit determination and the stability of the orbit prediction shed in a good light the performance of the Galileo navigation system which has a great potential when achieves the fully operational capability. With the precisely determined Galileo orbit, one could provide more robust and independent reference for the LEO satellites. Therefore, that may improve our knowledge about different global phenomena. This has to be further investigated when the future Earth-observing satellites were to be equipped with the Galileo signal receiver.

Area B - Board 167

The Copernicus POD (Precise Orbit Determination) Service delivers as part of the PDGS of the Sentinel-1, -2, and -3 missions orbital products and auxiliary data files for their use in the corresponding PDGS processing chains. The most precise results are the NTC (Non-Time Critical) orbit products delivered for the Sentinel-1 and -3 missions. These orbits are part of the final SAR (Synthetic Aperture Radar) products in the case of Sentinel-1 and are serving as backup (for the CNES orbits) for the generation of the final altimeter products in the case of Sentinel-3. Consistent time series of the orbit products are a requirement for highest-level final products of the different satellite missions.
The Copernicus POD Service is compelled to deliver consistent orbit time series to the extent possible. However, due to external developments such as the updates of the ITRF (International Terrestrial Reference Frame) the time series get inconsistent. In addition, necessary model improvements and updates or correction of wrongly used instrument reference points lead to inconsistencies. A rigorous reprocessing is the usual way to provide consistent orbit time series after major switches necessary in the operational processing.
The Copernicus POD Service plans to perform an offline reprocessing to have consistent time series available for all Sentinel satellites. This includes all model updates and improvements made in the processing system since the launch of Sentinel-1A (April 2014). Updates (S-3) and correction of wrongly used antenna reference points (S-1) are also included as well as a full reprocessing of the GPS orbits and clocks in the ITRF14/IGS14 reference frame. The reprocessed GPS orbits and clocks have an arc length of 36 h instead of the usual 24 h, because all Sentinel orbit products cover more than 24 h. Thus the common discontinuity of the GPS orbit and clock products at the day boundaries is avoided.
The results of this offline reprocessing are presented including a summary of all updates and improvements made in the processing. The improvements of the resulting Sentinel orbit time series are discussed and analysed. The final goal is to have the reprocessed orbits included in the official Sentinel processing procedures at the individual PDGS of the missions.

Area B - Board 168

The Copernicus POD (Precise Orbit Determination) Service is among others responsible to deliver orbital and auxiliary data files to the PDGS of the Sentinel-1 mission. The orbits are needed to provide highest level (In)SAR products of the mission.
To guarantee this highest level a consistent time series of the orbits is essential. The operational orbit processing at the Copernicus POD Service is done as consistent as possible over time but nonetheless several model updates and improvements have been made during the years. This is in particular true for the orbit time series of Sentinel-1A being the oldest satellite in the Sentinel satellite series.
The Copernicus POD Service plans to perform an offline reprocessing to have consistent time series available for all Sentinel satellites. This includes all model updates and improvements made in the processing system since the launch of Sentinel-1A (April 2014). A major change for Sentinel-1 is the correction of wrongly used coordinates of the antenna reference points for the GPS antennas of both satellites (not yet corrected for the operational products). This implies a change of the absolute orbit positions of several cm. Additionally, analyses of differences between the results from both satellites revealed a possible update of the Center of Gravity coordinates of the satellites.
Sentinel-1A and -1B orbit results from this offline reprocessing are presented in detail considering the large impact of the antenna reference point updates. The improvements gained by the consistent reprocessing are presented among others in terms of orbit comparisons, orbit overlap analysis and analysis of the estimated orbit parameters.

Area B - Board 169

Single-receiver GPS ambiguity resolution methods are started to be deployed for precise orbit determination (POD) of the Sentinel satellites. Several analyses from DLR, CNES, TUD and ESOC showed the improvement of the orbit results of Sentinel-3 when applying ambiguity-fixing. Prerequisite for these methods are dedicated information and products providing the necessary information to perform the single-receiver ambiguity resolution, e.g., the GPS wide-lane bias product together with the corresponding orbit and clock products from the CNES/CLS analysis center of the IGS.
The Sentinel-1, -2, and-3 orbit products delivered from the Copernicus POD Service to the PDGSs (payload data ground segment) of the individual missions are not yet based on fixed carrier phase ambiguities. Effort is taken to develop and implement corresponding interfaces and ambiguity fixing strategies. Aim is to be able to deliver ambiguity-fixed Sentinel orbits in the operational environment of the Copernicus POD Service.
The impact of the ambiguity-fixed carrier phases on the resulting Sentinel orbit products is analysed. Due to the different latencies of the orbital products it has to be carefully investigated for which product lines dedicated bias products are available in time. Backup procedures for the case that the necessary information is missing have to be developed as well. Additionally, Sentinel-1 and -2 do not provide observations at multiples of 10 seconds. This might complicate the applicability on those missions.
This study will give an overview of the work done at the Copernicus POD Service to follow up recent developments in the frame of single-receiver ambiguity resolution for POD of the Sentinel satellites. This will also include the impact on the future C&D units that will also track Galileo signals.

Area B - Board 170

The Sentinel satellites are developed for the specific needs of the Copernicus programme. Precise orbit determination (POD) is supported through an onboard Global Positioning System (GPS) receiver. In this study, we assess ambiguity fixed POD solutions for a GPS-only receiver on each SENTINEL spacecraft. Rather than fixing double-difference ambiguities based on a network of reference stations, ambiguity resolution of a single receiver requires dedicated satellite orbit, clock, and wide-lane bias products. The IGS (International GNSS Service) CNES/CLS (Centre National d'Etudes Spatiales/ Collecte Localisation Satellites) analysis center provides such GPS products with respect to P1, L1, P2, L2 observations. Satellite Laser Ranging observations on SENTINEL-3A and SENTINEL-3B satellites are employed to validate the radial accuracy of reduced-dynamic satellite orbits. Compared to float solution, the ambiguity-fixed reduced-dynamic solution reduces the standard deviation of satellite laser ranging residuals by about 10%. In addition, we introduce the fixed ambiguities into kinematic orbit determination. Ambiguity fixed kinematic solution marks a notable improvement of more than a factor of two over the float kinematic solution by comparing to the reduced-dynamic solution for all the SENTINEL satellites except for SENTINEL-1B. The reason might be due to the unclear phase center corrections and variations (PCO/PCV) of GPS antenna. We analyze these corrections for Sentinel-1B.

Area B - Board 171

In this presentation we will present our Sentinel and GPS orbits generated at the European Space Operation Centre (ESOC) as part of our Copernicus activities. We will highlight the improvements that have been made in our software (NAPEOS) as well as in our satellite modelling in recent times.

The main focus of this presentation will be on the integer ambiguity resolved orbits for the Sentinel missions as well as the benefits of a full combined processing (i.e., combining Sentinel together with the GPS satellites). In NAPEOS we are able to either resolve the integer ambiguities in a network approach in which the satellites are included in an IGS like setup or we can resolve the integer ambiguities using the GFZ/CNES method. The results and benefits of both these methods will be presented.

We evaluate the orbit accuracy by analysing post-fit residuals, orbit overlap errors, and orbit differences between our orbits and external orbits generated by other analysis centres. The consistency between our solutions and external orbits is below the 1 cm level in all directions.

Area B - Board 172

The Gravity Field and Steady-State Ocean Circulation Explorer (GOCE) orbited the Earth between March 2009 and November 2013 and conducted gradiometry measurements of Earth's gravity field. The resulting global GOCE-based gravity field models are of high spatial resolution and constitute a major source of information on the static part of Earth's gravity field. In the framework of the GOCE High-level Processing Facility (HPF), the Astronomical Institute of the University of Bern (AIUB) was responsible for the generation of the GPS-based Precise Science Orbits (PSOs) of the GOCE mission.

Since the end of the GOCE mission the understanding of remaining artifacts, especially in gradiometer data, but also in GPS data, has improved a lot and the corresponding processors and processing strategies have been significantly advanced. To benefit from these improvements, ESA has decided to reprocess the entire GOCE data set with the prospect of significantly improved GOCE-derived gravity field models. In the frame of this reprocessing campaign, AIUB is reprocessing the GOCE PSOs, using the latest version of the Bernese GNSS Software and the homogeneous time series of GPS products from a reprocessing effort conducted at AIUB in the frame of the Horizon 2020 project European Gravity Service for Improved Emergency Management (EGSIEM). Furthermore, refined data handling strategies to mitigate ionosphere-induced artifacts along the geomagnetic equator in GPS-only gravity field models derived from GOCE kinematic orbits are applied.

We give an overview on the applied processing strategy and compare the quality of the reprocessed PSOs to the previous ones in terms of consistency between reduced-dynamic and kinematic orbits and orbit overlaps, as well as by means of Satellite Laser Ranging (SLR). We demonstrate that the newly applied GPS data handling strategy is capable to strongly reduce the ionosphere-induced artifacts in derived GPS-only gravity field models, while not significantly degrading the orbit quality.

Area B - Board 174

Sentinel-1 A and B are twin spaceborne C-band synthetic aperture radar (SAR) sensors developed and operated by the European Space Agency under the Copernicus Earth Observation programme. An accurate geometric calibration has been performed since the Commissioning Phases on operationally generated Level-1 SAR products. Improvements introduced in the latest year on the geometric calibration of Sentinel-1A and -1B products allowed to achieve accuracies not only well within the mission requirements, but also well below the sample level.

On June 2016 an electronic problem on Sentinel-1A antenna tile #11 occurred stopping S-1A operation for eleven days (from 16th to 27th June). To resume the operations a new antenna configuration was identified with only the lower half of the tile #11 transmitting at full power in order to reduce power consumption. The new configuration allowed resuming S-1A operations and was further optimized on October 2017 to improve the electronic status of the antenna.

The analyses performed in the framework of the Mission Performance Centre (MPC) activities showed that this issue had a not negligible impact on target localization performances, introducing a swath-dependent bias in the along-track direction.

In this paper we investigate the origins of this bias and its impacts on data quality.

In particular, we show how this antenna issue has caused a change in the Sentinel-1A antenna patterns, introducing a phase trend along azimuth direction. The consequence of not compensating it during processing, combined to the antenna electronic steering implemented for the TOPSAR acquisition mode, resulted in the azimuth Absolute Location Error (ALE) bias observed on the data.

In addition to the theoretical demonstration, we validate the measured bias making use of the ARESYS SAR Data End-to-end Simulation Kit (SAR DESK), opportunely customized to reflect the Sentinel-1A configuration, and the ESA Sentinel-1 Antenna Model, to generate the antenna patterns in different conditions, i.e.:
a. before tile #11 issue
b. after tile #11 issue
c. after tile #11 reconfiguration
We show how well the simulations describe the measured biases and how, through the implemented antenna reconfiguration, it has been possible to reduce such biases and remove their swath-dependency.

To complete the picture, we describe also the impacts on Sentinel-1A interferometry, showing the effects on interferometric phase when combining images acquired with different configurations of the antenna and when using different strategies for data coregistration.

Area B - Board 175

An infrastructure of three corner reflectors for Sentinel-1 was deployed at the National Laboratory for Civil Engineering campus, in Lisbon, Portugal, with the collaboration of the IDL, FCUL. The infrastructure is formed by two triangular trihedral reflectors (1 x 1 m) and by a third one following a model proposed by the team. The third reflector is formed by two rectangular trihedrals with the vertical axis in common, each of the individual reflectors is oriented towards the ascending and the descending passes of Sentinel-1. Additionally, a Global Navigation Satellite System (GNSS) antenna is permanently installed on the reflector’s axis.
This study presents the first precision assessment for displacements measured at the infrastructure using Persistent Scatterer Interferometry (PSI). The PSI processing was performed with SARPROZ software using 19 Sentinel-1A images acquired between October 2017 and May 2018. The reflectors’ displacements are being monitored through levelling and GNSS. Levelling is performed for all reflectors every two weeks. Nearby persistent scatterers are included in the levelling line. A second permanent GNSS antenna is installed on a building roof in the campus to be used as a reference. PSI double-differences were compared to levelling and GNSS ones, projected onto the sensor line-of-sight. Precisions of 1.1 mm and 1.6 mm were achieved for the triangular trihedral reflectors, while 1.2 mm and 1.5 mm were obtained for the rectangular trihedral ones. However, during March 2018, strong winds have damaged the structure on which the rectangular trihedral reflectors are installed. If the evaluation is performed considering only the dates before the failure or only the dates after the damage was fixed, precisions of 0.7 mm and 0.8 mm are obtained. These results suggest submillimetre precision might be possible to achieve. Processing a larger time series of images acquired after the damage fix will be needed to confirm.

Area B - Board 176