Initially developed for monitoring the performance of TOPEX/Poseidon and follow-on Jason legacy satellite altimeters at Senetosa Cape, Corsica, this calibration/validation site, has been extended to include a new location at Ajaccio. This addition enables the site to monitor Envisat and ERS missions, CryoSat 2 and, more recently, the SARAL/AltiKa mission and Sentinel 3A&B satellites. Sentinel-3 carry altimeters that use a SAR (Synthetic Aperture Radar) mode that is different to the conventional pulse-bandwidth limited altimeters often termed “Low Resolution Mode” (LRM). The aim of this study is to characterize the Sea Surface Height (SSH) bias of the new SAR altimeter instruments and to demonstrate the improvement of data quality close to the coast. Moreover, one pass of Sentinel 3A&B overfly both Senetosa and Ajaccio with only a few seconds time difference allowing us to evaluate the reliability and homogeneity of both ground sites in term of geodetic datum. The Sentinel 3A and Sentinel 3B mean SSH biases are from this study not statistically different from 0 and the respective SSH biases look very similar at a cycle-by-cycle basis.

Area D - Board 385

Sentinel-3 is the Earth observation satellite mission designed to ensure the long-term collection and operational delivery of high-quality measurements of, among others, the sea surface topography. Post-launch calibration and validation of the satellite measurements is a prerequisite to achieve the desired level of accuracy and ensure the return on the investment. These Calibration/Validation (Cal/Val) services are provided by independent, external Cal/Val facilities that determine the error in satellite measurements, using known and controlled signal inputs on the ground. Sentinel-3 altimeter calibration site was established in West Crete, Greece. This site has been named CDN1 Cal/Val site and is located at an elevation of Transponders are commonly used to calibrate absolute range from conventional altimeter waveforms because of their characteristic point target radar reflection (not in the case of regenerative transponders). The waveforms corresponding to the transponder distinguish themselves from the other waveforms resulting from natural targets, in power and shape.

The transponder is used to calibrate SRAL’s range and datation to meet the mission requirements. For this calibration, the S3 L1A data is processed with a specialised transponder processor. Atmospheric delays are acquired directly from the calibration site providing better accuracy to the final range measurement. Ideally, the comparison between the theoretical values provided by the well-known target, and the measurement by the instrument to be calibrated provides us with the error that the instrument is introducing when performing its measurement. When this error can be assumed to be constant regardless of the conditions, it will provide the bias of the instrument. If the measurements can be repeated after a certain period of time, it can also provide an indication of the instrument drift.

The range and datation results of Sentinel-3A and Sentinel-3B using the Crete transponder are presented in this paper. This work is been carried out within the Sentinel-3 Mission Performance Center activity S3MPC.

Area D - Board 386

The Sentinel-3A&B are the most recent altimetric missions of a long historical series, and the first altimeters of the Copernicus constellation.

It is well known the crucial importance of the instrumental calibration when processing the altimeter data, in order to produce reliable and accurate L2 geophysical retrievals, such as sea surface height, significant wave height, or wind speed over the oceans.

isardSAT, as ESA Expert Support Laboratory within the Sentinel-3A Mission Performance Centre team, is responsible of monitoring the calibration parameters during the Sentinel-3A SRAL mission. This presentation addresses the monitoring results of up to date SRAL calibration data, acquired during the Commissioning and Routine Phases of the Sentinel-3A&B altimeter missions.

The SRAL altimeter permits to operate in two modes, LRM (low resolution) and SAR (high resolution) in two frequency bands, Ku and C. For the 4 combinations, CAL-1, CAL-2 and Autocal calibration modes are monitored and analysed from L1b data. Also the on-board thermal variations are monitored from L0 data. The on-board tracker performance is studied from L2 data.

Area D - Board 387

The tandem mission at the start of Sentinel-3B's life was designed to enable a close comparison of simultaneous measurements from the altimeters and other sensors. This was both to facilitate an accurate intercalibration between two SAR altimeters, and to permit an investigation of differences between instruments in SAR and LRM operation; this latter comparison being a key step towards integrating both SAR and LRM datasets into a seamless climate data record.
In this work, we consider the high-resolution (20 Hz) records from both SRAL instruments and characterise the bias and variability of one instrument relative to the other in terms of various operating regimes (low/high sea state, small/large radial velocity etc.). A strong covariance is noted between the local anomalies induced by multiplicative noise for i) backscatter strength (sigma0) and mispointing (psi-squared) in LRM mode, and ii) range (epoch) and wave height in both LRM and SAR modes. Although the high spatial variability of MLE4 estimates of sigma0 has been known for a long while, it still remains important to characterise this for each altimeter. Indeed, the observed covariance between estimates of different parameters has been shown to be a property of both the instrument and the associated retracking scheme.
For the standard MLE4 retracker (applied to LRM or PLRM waveforms), an empirical correction can be applied to reduce the short-scale variability in sigma0 by about 90%. The gain in performance for the range estimates is about 40% for these modes, but much less for SAR processing (which would thus appear to be better optimised). Application of these corrections improves the agreement on all parameters concerned. This produces not only an increased confidence in the relative biases between two instruments operating in SAR mode, but also improves the alignment with respect to LRM implementation. Although the methodology is based on a premise of independent 20 Hz realizations of fading noise, spectral analysis shows the improvement to reduce power levels at all scales below 50 km.

Area D - Board 388

Waves data from Sentinel-3 plays an important role for the improvement of wave products provided by the Copernicus Marine Environment and Monitoring Service (CMEMS-MFC). The use of significant wave heights (SWH) from altimeters in wave model is crucial for the correction of the wave model misfiting induced by the uncertainties from the atmospheric wind forcing. Since May 2018 SWH from S3A have been assimilated in operational wave model of Météo-france in the frame of the global CMEMS-MFC. During the calibration/validation of S3A we have demonstrated that significant wave heights from SAR mode of S3A induces a slight increase of bias in ocean areas dominated by the swell wave regime such as the agulhas and the south of Australia ocean zones.
The goal of this work is on the one hand to evaluate the quality of the wave data from S3B, and on the other hand to compare the performance of the assimilation when using SAR and LRM modes retrieval both processed in the tandem phase of S3B.
Several assimilation runs of S3A and S3B wave data have been performed during the tandem phase in June 2018. The validation of the results is performed with independent wave data from altimeters and buoys. The results will be discussed for the analysis and forecast periods. Particular attention will be considered for swell dominant ocean areas. Comparison will be also performed when using combined assimilation of SWH from S3 and SAR directional wave spectra from Sentinel-1.
Further discussions and conclusions on the results will be presented in the final paper.

Area D - Board 389

Sentinel-3 is part of a series of Sentinel satellites responsible for taking care of a continuous ‘health check’ of the Earth planet under the umbrella of the Copernicus program. The Copernicus program will launch four Sentinel-3 satellites (from A to D) to achieve this goal from 2016 to 2030. EUMETSAT’s Marine Centre is responsible for the processing of the Sentinel-3 science data in the marine environment.

Sentinel-3B was launched in April 2018 to join its companion mission Sentinel-3A, launched in February 2016. In order to compare the two missions, Sentinel-3B was flying in tandem with Sentinel-3A between May and October 2018. That means that the two satellites were trailing each other by only 30 seconds on the same ground-track. This unique opportunity allowed very accurate comparison of the two missions to assess:
• Relative biases between the retrievals of altimeter range, backscatter coefficient, and significant wave height on both altimeter frequencies (Ku- and C-band);
• Relative biases in the auxiliary measurements of wet tropospheric pass delay and ionospheric correction;
• Regional variations in their differences;
• A unique direct comparison between the conventional LRM and more modern SAR altimeter operating modes and how this is impacted by e.g. swell.

This work aims at presenting those commissioning results of Sentinel-3B in comparison with Sentinel-3A. This includes cross-comparisons of sea level, significant wave height, backscatter coefficients and wind speed; both their absolute value, relative drift, regional variations, and performance. We will also indicate how these commissioning results where fed into the processing baseline and reprocessing.

The latest Sentinel-3A and -3B multi-mission cross-calibration results are to be provided: quantification of cross-overs (mono- and multi-mission); multi-mission time series and collinear track analyses (including Jason-2, Jason-3, and SARAL/AltiKa).

Area D - Board 390

Sentinel-3 is part of a series of Sentinel satellites responsible for taking care of a continuous ‘health check’ of the Earth planet under the umbrella of the Copernicus program. The Copernicus program will launch four Sentinel-3 satellites (from A to D) to achieve this goal from 2016 to 2030. EUMETSAT’s Marine Centre is responsible for the processing of the Sentinel-3 science data in the marine environment.

Since Spring 2016 Sentinel-3A’s SRAL Synthetic Aperture Radar Altimeter has been successfully contributing to the continuity of the sea level climate data record. To further the quality of this records, its processing baseline has considerably evolved during the first three years of the mission.

This presentation will provide an overview of the latest evolutions of the Sentinel-3 SRAL processing, the relations between Sentinel-3A and -3B processing, and the strategy that EUMETSAT has adopted to provide the consistent long-term data set while continuing to evolve and improve the processing algorithms and standards. The latest reprocessing “Spring 2019” is the baseline for a quality analysis of the Sentinel-3A Marine Centre data in a multi-mission setting. The latter shall allow for revisiting the status of the Jason-3 and Jason-2, in comparison with Sentinel-3A. To this goal this presentation aims at: providing multi-mission time series of the main climate records (sea level, significant wave height and wind speed); quantifying cross-overs (mono- and multi- mission); as well as provide a 3-year global assessment of SAR mode versus Pseudo-LRM.

Area D - Board 391

Satellite altimetry provides valuable information for water level monitoring of inland water bodies, such as lakes, reservoirs and rivers, for many years. However, the performance of classical pulse-limited radar systems (e.g. as flying on board of Jason-3) is limited due to the relatively sparse along-track resolution and the large footprints that are prone to land contamination. More modern techniques based on the Delay-Doppler principle provides SAR data, which are characterized by an improved along-track resolution and smaller, elongated footprints (stripes). Cryosat-2 was the first mission operating in SAR mode in selected geographical regions. The satellites of the Sentinel-3 mission provides SAR mode data globally while they are also able to work in the classical low resolution mode (LRM). However, a parallel dissemination of SAR and LRM data is not possible.

Existing studies indicate the improved performance of SAR altimetry for monitoring inland water bodies, especially for small lakes and rivers. However, until now, a direct performance comparison between LRM and SAR mode was impossible without accepting either a significant time difference or different locations between the measurements. The usage of SAR data processed as Pseudo LRM (PLRM) was a workaround but not the optimal solution. The tandem phase of Sentinel-3A and Sentinel-3B offers now the first possibility to directly compare and explore both measurement approaches using similar instruments and sampling the same Earth environment just a few seconds apart from each other.

This contribution will use data from the first part of the tandem phase in summer 2018 when the Sentinel-3B satellite was operating for one cycle in LRM close mode loop while Sentinel-3A provides SAR open loop data. A set of inland water bodies beneath the Sentinel-3A track will be identified as study cases characterized by different sizes, shape and surrounding topography. These will be used to investigate and compare the different performances of the two altimeter tracking modes for inland water applications. The investigation is limited to single overflights since only one Sentinel-3 cycle of 27 days is available for comparison.

Area D - Board 392

Among several instruments, Sentinel-3 mission payload includes an advanced radar altimeter and a 2-band microwave radiometer (MWR), from which accurate measurements of the sea surface topography can be estimated. The MWR instrument is similar to that of Envisat and aims at the precise retrieving of the wet tropospheric correction (WTC) through collocated measurements with the Synthetic Aperture Radar Altimeter (SRAL) instrument.
Aiming at maximizing the space-temporal coverage of the two satellites, Sentinel-3A (S3A) and Sentinel-3B (S3B), launched on 16 February 2016 and 25 April 2018, respectively, have been planned to fly on identical high-inclination, sun-synchronous polar orbits with 27-day repeat cycles, S3B shifted 140º with respect to S3A. In view to perform accurate inter-calibration of the instruments carried by both satellites, between June and October 2018 they have been in the so-called tandem mission, with S3B flying approximately 30 seconds ahead of S3A.
The absolute and relative calibrations of the S3A and S3B MWR are crucial tasks to ensure the retrieval of sea level climate data records from the respective SRAL instruments. To this end, the assessment of the performance of S3A and S3B MWR and that of the corresponding MWR-based WTC are of up-most importance. Equally relevant is the derivation of enhanced WTC for both satellites using the GPD+ algorithm, giving continuity to the GPD+ WTC dataset, continuous and valid over all surface types, over the period ERS/Envisat.
Following the similar study on S3A, now extended to the whole period since launch, an in-depth study of the S3B MWR performance has been done by means of the comparison of S3B MWR-derived WTC i) with independent WTC datasets computed using Global Precipitation Measurement (GPM) Microwave Imager (GMI), Jason-3 and GNSS observations, and with a GPD+ WTC using only third-party data; ii) with respect to S3A. In what concerns the analysed parameters, both the brightness temperature (TB) of the two MWR channels and the WTC were used, paying particular attention to the performance of the instrument, as well as that of the 5-input retrieval algorithm, over coastal regions.
Initial results reveal a performance of the S3B radiometer very close to that of S3A. Using only valid open-ocean points, absent of land and ice contamination, S3B 23.8 GHz channel has a bias of 2 K and a standard deviation of 0.6 K with respect to that of S3A, the corresponding values for the 36.5 GHz channel being 1.1 K and 0.7 K respectively. Using the same open-ocean data-set, the S3B MWR-derived WTC shows a -0.6 cm bias with respect to S3A (S3A measuring wetter than S3B) and a 0.3 cm standard deviation. When coastal points are included, these numbers increase to -0.8 cm and 3.0 cm respectively, an evidence of the land contamination present in the observations from both instruments.
In agreement with previous assessments for S3A, the MWR-derived WTC from both instruments agree well with that from the GMI and Jason-3 MWR, with RMS differences around 1-1.3 cm. Strong ice and land contamination can be observed in both instruments, the latter one being mainly up to 25 km, as shown by comparisons with GNSS. At the time of the Living planet Symposium, it will be possible to present the long-term evolution of the S3A radiometer for the first 3 years of this mission, by comparison with GMI, Jason-3, GNSS and the GPD+ WTC solely derived with third-party data.
In spite of the good overall performance of S3A/B radiometers over open-ocean, in view to get accurate WTC, valid everywhere, in particular in coastal and polar regions, it is clearly preferable to use the continuous, homogenous and consistent WTC from GPD+ rather than either the original MWR-derived WTC or the Composite WTC.

Area D - Board 393

This work describes the absolute and relative bias estimation of Sentinel-3A and Sentinel-3B altimeters during their tandem mission using the Permanent Facility for Altimeter Calibration (PFAC) in west Crete, Greece. The performance of both altimeters is also compared against other reference missions.

This PFAC has been providing for more than 15 years calibration services for all past (i.e., Envisat, Jason-1, Jason-2, SARAL/AltiKa, HY-2A) and current (i.e., Sentinel-3A, Sentinel-3B, Jason-3) satellite altimeters. It consists of: a) the Sentinel-3 transponder Cal/Val site established on the mountains of Crete at an elevation of 1100m, b) three permanent sea-surface Cal/Val sites, c) a permanent network of 17 continuously operating Global Navigation Satellite System stations, and d) an Operations Control Center for data archival, processing, analysis and interpretation of results.

The Sentinel-3B mission was launched on 25 April 2018 and achieved its tandem phase, flying 30 seconds ahead of its twin satellite Sentinel-3A, on 6 June 2018. In their ascending Pass No.14, Sentinel-3A and Sentinel-3B have been passed over an open sea area west of the Gavdos Cal/Val facility and some seconds later fly over the transponder Cal/Val site on the mountains of Crete. Thus, these Sentinels have been calibrated with two diverse techniques. Sea-surface Cal/Val results are also presented for their descending Pass No.335. An uncertainty analysis for the altimeter biases produced by the transponder and also by sea-surface calibrations is also presented following the new concept of Fiducial Reference Measurements.

The performance of the SRAL altimeter of the Sentinel-3A and Sentinel-3B is also compared against the Jason-3 reference altimeter. This is accomplished at a crossover location south of Gavdos, where the Sentinel-3 Pass No. 14 and Jason-3 Pass No.109 intersect. On 21 Aug. 2018, these three altimeters flew also over this location within some hours. A similar crossover occurred over the transponder on 17 Sept. 2018. On that day, transponder calibrations were carried out at 02:20:08 UTC (Jason-3), 19:58:42 UTC (Sentinel-3B) and 20:00:12 UTC (Sentinel-3A).

The preliminary results for the absolute calibration of Sentinel-3A and Sentinel-3B altimeters during their tandem mission indicate a difference of the order of +1cm (S3B minus S3A). The absolute bias of Sentinel-3A is estimated to be +0.6mm. The relative bias between Sentinel-3A and Jason-3 is estimated to be +3.75cm (S3A minus Jason-3) at their crossover location.

Area D - Board 394

The Sentinel-3 mission is a two-satellite operational system designed to provide wide coverage and high-quality data for the Copernicus environmental monitoring programme. Sentinel-3B, the second satellite of this system, has been successfully launched in April 2018, and will arrive at its baseline position (at ±140° to its twin, Sentinel-3A) at the very end of 2018. Before being inserted into its final orbit, S3B has been positioned in a four-month-long close formation with S3A, during which the two satellites were observing the same scene a mere 30 seconds apart. Such flight configuration allowed us to perform very accurate cross-calibration of the Sentinel-3 constellation instruments, in order to produce homogeneous and unbiased time-series observations for climate record as requested by the user community.

The Sentinel-3 tandem phase also offers a unique and unprecedented opportunity to provide a more complete assessment of the main altimeter measurement approaches (SAR and LRM/PLRM modes) and identify any potential processing improvements that would enhance the altimeter data quality. Additional studies were undertaken focusing on the exploitation and combination of altimeter data from the two-satellite constellation in an attempt to provide further information on the different measurement approaches that may not be accessible from mono-satellite mission.

This paper reports the most significant findings emerging from those studies based on joint use of S3A and S3B in tandem phase.

Area D - Board 395

The Sentinel-3 Marine Altimetry L2P-L3 Service supplies value-added Sea Level Anomaly L2P and global L3 products for Sentinel-3A and—very soon—to Sentinel-3B Mission. In addition to the Sea Level Anomaly products, the Service is being evolved to produce and distribute NRT Wave L2P and L3 products for both, Sentinel-3A and Sentinel-3B missions.
The service was developed and is operated by CNES/CLS in the frame of the Copernicus Programme, funded by the European Union, under a EUMETSAT Contract. It is based on the existing operational system in place for Jason-3, Jason-2, CryoSat-2, SARAL.
The Service was pre-operational since mid-December 2016 and is fully operational since 27 June 2017 for the distribution of Sea Level Anomaly products:
- Marine L2P products in near-real-time (NRT), short-time-critical (STC) and non-time-critical (NTC) timelinesses on AVISO+ ftp; NRT and STC products are also distributed via EUMETCast, EUMETSAT’s multi-service dissemination system;
- Global L3 products in near-real-time (NRT, based on short-time-critical L2P data completed with near-real-time L2P data) and delayed time (DT) timeliness via CMEMS.
A Service extension is currently on-going to produce and distribute Wave products at global scale. The Service will be pre-operational early 2019 and is planned to become fully operational in Summer 2019 with the distribution of Wave products to the users:
- Marine L2P products in near-real-time (NRT) timeliness on AVISO+ ftp and EUMETCast, EUMETSAT’s multi-service dissemination system;
- Global L3 products in near-real-time (NRT) timeliness via CMEMS.
Sea Level Anomaly Products: The L2P products are along-track mono-mission pole-to-pole products, providing as much as possible the same updated corrections and models for several altimeter missions, in order to facilitate inter-mission comparisons. In addition, they are intercalibrated versus the reference mission in order to have consistent timeseries since TOPEX/Poseidon. The sea level anomaly is provided with a validity flag, enabling users to discard data with spurious measurements. The L3 products are along-track products, 1 day in duration, that contain the valid and low-pass filtered sea level anomaly and absolute dynamic topography after inter-calibration with a reference mission.
Wave Products: the L2P products are along-track pole-to-pole products that contain time, significant wave height and information of validity of the data. They have a homogenized format and content for all altimeter missions. They are calibrated versus buoys. The L3 products are along-track products, 3 hours in duration, that contain time and significant wave height for valid marine surfaces. They have under-gone a filtering process.
This service and its products will be presented in detail.

Area D - Board 396

Sentinel-3A has been in orbit since February 2016 and has been providing data on sea state (significant wave height; SWH and sigma0/windspeed). With the launch of Sentinel-3B in April 2018 there is a requirement to compare the two missions with each other and also with independent data. The operation of Sentinel-3B in tandem with Sentinel-3A provides a unique opportunity to obtain data close in space and time (Sentinel-3B is 30 seconds ahead of Sentinel-3A) and for many purposes can essentially be considered to be instantaneous. In addition, the operation of the two Sentinel missions in different operating modes (Low Resolution Mode; LRM and Synthetic Aperture Radar Mode; SARM) provides a unique opportunity to investigate the impact of these two approaches on the measurements retrieved. The independent data includes in situ buoy data, model output and other satellite data (e.g. Jason and CryoSat-2) all of which, like the Sentinel-3 data, include uncertainties.

In this study, we present ongoing work concerned with the calibration and validation of data from the two Sentinel-3 missions. Using the independent data, we will examine the geographical distribution and uncertainty characteristics of SWH and windspeed from the Sentinel-3 missions as well as investigating any offsets. Longer term, we will apply the triple collocation method (Stoffelen, 1998) to the Sentinel-3A, Sentinel-3B and in situ data to better characterize the errors of the measurements, as well as examining dependence of SWH and wind speed errors on sea state parameters for LRM and SARM.


Reference
A. Stoffelen, Toward the true near-surface wind speed: Error modeling and calibration using triple collocation. Journal of Geophysical Research-Oceans 103, 7755-7766 (1998).

Area D - Board 397

During the Sentinel-3 Tandem Mission, the altimeters on Sentinel-3 A and B each acquired observations along the same ground tracks, separated in time by only 30 seconds. Sentinel-3 B operated sometimes in SAR and sometimes in LRM mode, while Sentinel-3 A remained in SAR mode. The primary goal of a tandem mission is to inter-calibrate biases between the two satellite measuring systems, so that there is no change in accuracy in moving between the two data sets. However, the availability of both LRM and SAR mode measurements from the Sentinel-3 Tandem Mission also allows investigation of the relative precision of different radar processing schemes, and several such studies have been or are being carried out.
We have developed a processor for Sentinel-3 SAR mode Level-1A data that allows us to process the same radar pulse echoes by three different algorithms, generating Pseudo-LRM, Delay/Doppler SAR, and Fully-Focused SAR waveforms and geophysical retrievals. Our processor can create all three outputs at any desired sampling and posting rate along the ground track, such that these different measurement varieties are all geo-located at the same points on Earth's surface. We can produce these three different kinds of estimates from either Sentinel-3 A or B or both, so long as Level-1A SAR mode data are available.
Tandem Mission Sentinel-3 B SAR mode data allow us to inter-compare and inter-calibrate between six varieties of data (pseudo-LRM, D/D SAR, F-F SAR from each of Sentinel-3 A and B). Tandem Mission Sentinel-3 B LRM mode data allow us to compare pseudo-LRM, D/D SAR and F-F SAR data from Sentinel-3 A with LRM data from Sentinel-3 B. This comparison has not been possible previously, as simultaneous LRM (not Pseudo-LRM) and SAR data collection is furnished for the first time by the Sentinel-3 Tandem Mission.
We present an inter-comparison of these altimeter operating and processing modes. Our analysis begins by exploiting flights over transponders to verify that the F-F SAR algorithm developed for Sentinel-3A also works for Sentinel-3B. We then move to data from other Earth surface types to explore the possibilities and relative precisions of these different algorithms.

Area D - Board 398

The Sea and Land Surface Temperature Radiometers (SLSTR) on the Copernicus Sentinel-3 A and B missions are multi-channel scanning radiometers primarily designed to measure global sea-surface temperatures (SST) to an uncertainty < 0.3 K and land-surface temperatures (LST) to an uncertainty <1 K for climate monitoring. The first instrument was launched in February 2016 and was joined by its sister in April 2018. In principle, and to ensure seamless SST and LST data products from the two instruments, the instruments are identical and both are self-calibrating with an uncertainty in brightness temperature of 0.1 K over the main operating range. Nevertheless, small differences between the instrument builds and uncertainties in the calibration will inevitably lead to offsets in the radiometry of the two.

The tandem phase of the Sentinel-3A and B satellites provides a unique opportunity to directly compare the radiometric calibrations of the SLSTR A and B instruments. The tandem comparison data will be analysed for many long-term applications. During the commissioning phase of the SLSTR-B instrument, comparisons of the L1 tandem data have been used to verify the radiometric performance and to adjust the calibration factors of the VIS and SWIR channels to align to SLSTR-A.

Direct comparisons of the L1 products produced by the SLSTR-A and SLSTR-B instruments showed that there is a strong sensitivity to geometric differences, particularly where there are scenes with strong spatial and temporal variability, in particular over clouds. Therefore filtering of individual scenes is needed to reduce the scatter on the pixel-to-pixel comparisons.

An approach used for this paper is to remap the L1 radiometric data to a L3 grid with a resolution of 0.05degrees. This was found to be more efficient that comparing individual L1 images and allows comparisons to be performed over a wider range of scene temperatures/radiances. We present the results of the preliminary comparisons of the analysis.

Area D - Board 399

The European Copernicus programme’s Earth-observation satellite-mission Sentinel-3 carries a suite of instruments aimed to produce high-quality, high-frequency, long-term measurements of the Earth’s oceans, land, ice and atmosphere. In order to achieve the required coverage the mission will be composed of a pair of identical satellites, Sentinel-3A (S3A) and Sentinel-3B (S3B). After their 7.5 year lifespan has been completed, S3A and S3B will be replaced by two further satellites, Sentinel-3C (S3C) and Sentinel-3D (S3D), to provide collectively a 20 year period of continuous observational coverage.

Despite the fact that the set of instruments carried by all four Sentinel-3 satellites are designed and built to be nominally identical, it is always a challenge to produce consistent, long-term data records from the combination of observations by a series of multiple sensors. This is due to small differences in behaviour and calibration between the instruments in a given series, which are particularly significant for the study of subtle climatic trends. These inter-sensor inconsistencies, therefore, must either be corrected for or understood within the context of the measurement uncertainties. This is achieved through the process of comparison of the measurements from satellites in a series and external references.

When, after 2 years in orbit alone, S3A was joined by S3B in April 2018 the two satellites entered a so-called tandem phase. This six month phase, designed to understand inter-sensor differences, involved S3B flying close to S3A, separated by just 30 seconds, before moving away to its nominal orbit. Unlike typical satellite sensor inter-comparison exercises, which introduce error due to imperfect comparison conditions (for example comparatively large time separations) and rely on a small number of sensor-to-sensor match ups, the tandem phase provides a unique opportunity to directly compare the two satellites’ measurements, investigate any differences found and ultimately inter-calibrate, or harmonise, them to reconcile observed differences.

Rigorous comparison of measurements is integral to metrology, the science of measurement. Formal, global comparisons are undertaken regularly by National Metrology Institutes (NMIs) to ensure worldwide consistency of the international system of units (SI). The work presented here applies insights from metrology to the analysis of the tandem phase data for one of Sentinel-3’s instruments, the Sea and Land Surface Temperature Radiometer (SLSTR) Level 1 (L1) product. SLSTR is a dual-view, scanning radiometer which has 11 spectral channels spanning from the visible to the thermal infrared, with the primary objective of accurately measuring the surface temperature over sea and land.

Presented here is a metrological methodology and preliminary results for SLSTR L1 tandem analysis, including the inter-comparison of the measurement data and validation of the measurement uncertainties. Of course there are challenges associated with the tandem phase analysis not typically encountered in standard metrological comparisons, in particular the high data volume with a complex error-covariance structure, and so methodology for handling this is also presented.

Area D - Board 400

Quantifying lava flow hazards by combining field observations, satellite data and numerical modeling has immediate applications to the real time monitoring of effusive eruptions. By monitoring, we mean here both following the manifestations of the eruption once it has started, as well as forecasting the areas potentially threatened by lava in an effusive scenario. The strategy implemented in the Lav@hazard web-GIS framework for the assessment of lava flow inundation hazards integrates the HOTSAT satellite monitoring system to analyze multispectral images and determine hotspot location, lava thermal flux, and effusion rate estimation, and the physics-based MAGFLOW model to simulate the spatial and temporal evolution of lava flow fields. The satellite-derived TADR (Time Averaged Discharge Rate, i.e. an estimation of the effusion rate) provided by HOTSAT is used to drive MAGFLOW simulations in order to obtain updated lava flow scenarios in near real time. Recently the HOTSAT system, initially designed for low to moderate spatial resolution data (e.g. MODIS, AVHRR, SEVIRI) has being extended to ingest recent data from SLSTR instrument (Sentinel-3 mission) and higher spatial resolution satellite data (e.g. ASTER, Landsat-8, Sentinel-2) particularly precious to retrieve detailed thermal maps of the lava flow field.
These thermal maps are then compared with the eruptive scenarios simulated by MAGFLOW for iterative validation. This methodology, designed for Etna volcano and validated with recent eruptions, has been successfully exported to poorly monitored volcanic areas. This new satellite-driven modelling strategy represents an operational monitoring system that should give during an eruption: (i) the current state of the effusive activity; (ii) the probable evolution of the lava flow field; (iii) the potential impact of lava flows. Here, we will describe and demonstrate the operation of this system by using a retrospective analysis of recent lava flow-forming eruptions at Etna (Italy), Fogo (Cape Verde) and Nabro (Eritrea) volcanoes.

Area D - Board 401

The main objective of the Sentinel-3 mission is to provide accurate and reliable measurements of sea surface topography, sea and land temperature, and ocean and land surface color. The mission foresees the simultaneous operation of a constellation of up to three satellites and it is vital to ensure the delivery of consistent, high quality and well calibrated data to the scientific community.
The deployment of the Sentinel-3 constellation started with the launch of the A platform in 2016, followed after two years by the launch of the B platform, on 25 April 2018. During the commissioning phase of Sentinel-3B, it has been foreseen to operate the satellite in “tandem” configuration with the A module: the “tandem” phase consists in the operations of both satellites flying on the same orbit, with the B module preceding the A module of ~30 seconds with exactly the same viewing geometry. This unique configuration provides a huge number of very refined matching dataset of the Sentinel-3A and -3B observations, in the sense of strongly reducing matching errors (because of the simultaneity and the similar geometry), which allows powerful statistical analysis, for instance the relative non-linearity behavior, the relative temporal evolution, the intercalibration of all bands, and for the whole field of view of the instrument.
In this paper, we propose a technique to systematically process and match Sentinel-3 data (OLCI and SLSTR) in tandem configuration by means of an image registration approach using the maximization of the Enhanced Correlation Coefficient (ECC) [1]. This technique allowed to create tandem registered datasets for OLCI A/B, SLSTR A/B Nadir/Oblique and to perform analyses at pixel level.
The orbital stability achieved during the operation of the two platform in tandem configuration allows to easily reach a very good level of registration (mis-registration within 1 pixel) between products almost everywhere, with the only exception of cloudy targets, which usually showed a mis-registration of 1 to 3 pixels due to clouds’ movement. The considered tandem dataset consists of a time series of one full day every 10 days (23.06, 04.07, 14.07, 28.07, 08.08, 18.08, 30.08, 08.09, 18.09).
The statistical analysis of the tandem couples has been carried out selecting homogeneous areas with very high level of registration and low standard deviation. The selection of the candidates for the intercomparison can be also optionally triggered by the cloud masks or land/sea mask or over particular geographic areas.
The OLCI tandem dataset has been analyzed at camera and band level. The relative evolution of the B instrument with respect to the A one has been also addressed and compared to the relative evolution obtained by the operational monitoring of the on-board diffusers. The relative inter-band calibration and residual non-linearity have been addressed as well.
The same analysis has been repeated systematically for SLSTR couples, for Nadir and Oblique views, evaluating the relative differences and residual non-linear behavior between the two modules for the VIS/NIR and TIR bands.
The results are critically discussed and presented, which reveal specific signatures for both OLCI and SLSTR sensors with a very high accuracy.RTM simulations have been also carried out to better understand the large differences observed between –A and –B for bands for which the instrument spectral response may slightly differs, which could be crucial for some bands operating close to strong atmospheric absorption lines.
[1] Evangelidis and Psarakis, Parametric Image alignment Using Enhanced Correlation Coefficient Maximization, IEEE Transaction on pattern Analysis and Machine Intelligence, vol.30, N10, October 2008

Area D - Board 402

The future FLEX/Sentinel-3 tandem mission will provide unique information on vegetation dynamics by exploiting solar-induced fluorescence and reflectance indices at the unprecedented spatial resolution of 300m x 300m. This magnified view into the photosynthetic machinery will enhance our ability to face the actual and future challenges related to Earth biogeochemical and climate system. FLEX will fly in tandem with Sentinel-3 in order to make use of its payload for atmospheric characterization and thermal measurements.

In summer 2018, during the tandem phase with Sentinel-3A, the Sentinel-3B OLCI sensor was reconfigured to acquire data in 44 spectral bands in the visible and near-infrared region, mimicking as much as possible the future FLEX data (“FLEX mode”). This provides the opportunity to exploit for the first time real satellite data for the retrieval of solar-induced fluorescence in the red and far-red spectral regions at 300m x 300m spatial resolution. Moreover, in correspondence to the reconfigured S-3B overpass, the ESA-funded FLEXSense campaign was organised in order to collect multi-scale and multi-temporal measurements of vegetation fluorescence and reflectance in different sites across Europe and US, both from near-ground and airborne sensors, at a relevant spatial scale for comparison with the satellite estimates.

In this contribution, we present the concept and the initial stages of the processing of Sentinel-3B “FLEX mode” data from Level 0 to Level 1B through the adaptation of the OLCI ground processor, together with the first attempts to obtain higher level products from these data, and perspectives on the retrieval of solar-induced fluorescence from space.
Moreover, the quality of these products was evaluated using top of canopy reflectance and fluorescence data coming from the FLEXSense campaign, resampled at the 300m x 300m spatial resolution to match the Sentinel-3B grid, together with the comparison of Sentinel-3A and Sentinel-3B OLCI products. The outcomes of this activity will support the further development of FLEX, highlighting challenges and opportunities for the future calibration and validation stages of the mission.

Area D - Board 403

The main objective of the SENTINEL-3 mission is to measure sea surface topography, sea and land surface temperature, and ocean and land surface colour with high accuracy and reliability to support ocean forecasting systems, environmental monitoring and climate monitoring, through acquisitions by multiple sensing instruments. Two of them, OLCI (Ocean and Land Colour Imager) and SLSTR (Sea and Land Surface Temperature Radiometer), are optical sensors designed to provide a continuity with Envisat's MERIS and AATSR instruments. OLCI is based on the opto-mechanical and imaging design of MERIS instrument, i.e. a push-broom imaging spectrometer with five cameras for a 1270 km total FOV. The instrument flies 21 spectral bands from 0.4 to 1.02 microns with an IFOV of 300 m over land and ocean. SLSTR is based on the design of AATSR instrument, i.e. a conical scanning imaging radiometer employing the along track scanning dual view technique for a total swath of 1420km nadir (and 750km backwards). The acquisition is made for a spatial sampling of 500m (VIS-SWIR) and 1000m (MWIR, TIR) and in 9 spectral bands from 0.55 to 12 microns.

At the start of the Sentinel-3B mission, between June and October 2018, the two S3A&B satellites were flown in close Tandem formation. The Tandem formation entails Sentinel-3B flying 30 seconds ahead of Sentinel-3A. This formation has been an unique opportunity to derive an accurate cross-calibration between the optical sensors of these two satellites.

Usually, optical sensors are recalibrated and recharacterized on-orbit particularly using on-board devices which include diffusers (OLCI, SLSTR) and/or black bodies (SLSTR). Afterwards, vicarious calibration methods and instrumented sites are used in order to validate the radiometry for the reflective bands and carry out cross-calibrations with many other instruments.

Cross-calibration between two optical sensors is usually done using stable target sites like Desert or Antarctic sites, which can be classified as bright sites. The Tandem phase allows us to use other various types of natural targets for the cross-calibration between S3A&B.

This paper presents the cross-calibration results between S3A and S3B (for OLCI and SLSTR) using the standard approach. Then, a TOA direct comparison between S3A&B over dedicated natural targets such as Rayleigh scattering, sunglint, desert sites, Antarctica, and deep convective clouds will be presented. Based on measurements on these sites, it is possible to derive a cross-calibration between the sensors of the two satellites over various site types using Tandem phase measurements. Then, a comparison between the two approaches will be presented and discussed in order to interpret the different results.

Area D - Board 404

The Atlantic Meridional Transect (AMT) programme has been one of the most valuable sources of high-quality in situ measurements for remote sensing over the past 20 years. In particular, the Copernicus Atlantic Meridional Transect for Sentinel Fiducial Reference Measurements (AMT4SentinelFRM - 2016-2017) and Atlantic Meridional Transect Ocean Flux from Satellite (AMT4OceanSatFlux - 2018) campaigns have established the AMT platform as one of the key providers of Fiducial Reference Measurements (FRM) for the evaluation of the Sentinel data streams.

In this work we use data collected on the first AMT4OceanSatFlux field campaign (AMT-28 , September-October 2018) for the validation of Sentinel-3A and Sentinel-3B OLCI (Ocean and Land Colour Instrument) tandem-phase products. We employ pre- and post-cruise calibration to generate post-cruise uncertainty budgets for our SeaBird HyperOCR radiometer, to quantify the overall uncertainty of the FRMs and to identify and correct for biases. Quality-control procedures are then defined to ensure FRMs are of accuracy compatible with satellite validation. To minimise the effects of ocean variability in the inter-comparison of the two OLCI instruments, we employ a robust match-up procedure that takes into account quality and temporal/spatial homogeneity issues.

Using these tools, we perform a comprehensive accuracy assessment of Sentinel-3A and Sentinel-3B OLCI level 2 remote sensing reflectances and chlorophyll-a over a range of conditions. The performance of the two sensors are assessed in both independent and coincident match-up analysis, for the standard level 2 products and other relevant processors (POLYMER, L2GEN). A comparison with concurrent ocean colour sensors – Aqua MODIS and Suomi-NPP VIIRS – is also performed to evaluate the consistency in the common bands and the potential for a merged product. The methodology presented here offers a clear advantage in the number and quality of match-ups over traditional techniques, allowing us to fully exploit the inter-calibration opportunity that the tandem phase presented.

Area D - Board 405

As from November 24th 2018, the Sentinel-3B (S3B) mission will fly 140 degrees apart from Sentinel-3A (S3A), as requested by the users, to provide the best possible spatial and temporal coverage and sampling. This Sentinel-3 constellation will be the heart of the Ocean monitoring system in the Copernicus program.
Before this final operational constellation configuration, S3B has been flying in tandem with its twin S3A for a period of 6 months between May and October 2018. These few months of data, where S3B flew 30 seconds ahead of S3A are an invaluable source of information, to assess the performance of the instruments and processing.
In this presentation, we propose to give an overview of the activities performed at Eumetsat during this tandem phase and to present our main results on the comparison of the OLCI, SLSTR, and SRAL instruments.
We proposed and implemented an innovative technique to systematically process and match Sentinel-3 optical data (OLCI and SLSTR) in tandem configuration by means of an image registration approach using the maximization of the Enhanced Correlation Coefficient (ECC). This technique allowed to create tandem registered datasets for OLCI A/B, SLSTR A/B Nadir/Oblique and to perform analyses at pixel level. The orbital stability achieved during the operation of the two platforms in tandem configuration allows to easily reach a very good level of registration (mis-registration within 1 pixel) between products almost everywhere, with the only exception of cloudy targets, which usually showed a mis-registration of 1 to 3 pixels due to clouds’ movement.
The OLCI tandem dataset has been analyzed at camera and band level. The relative evolution of the B instrument with respect to the A one has been also addressed and compared to the relative evolution obtained by the operational monitoring of the on-board diffusers. The relative inter-band calibration and residual non-linearity have been addressed as well. The same analysis has been repeated systematically for SLSTR couples, for Nadir and Oblique views, evaluating the relative differences and residual non-linear behavior between the two modules for the VIS/NIR and TIR bands.
For OLCI instruments, the tandem phase enabled large-scale direct inter-comparisons of OLCI-A and –B radiometry at Level-1B products, which would be impossible otherwise. The results indicate on average a 2% difference between OLCI-A and –B, -B being less bright and closer to other missions. The analysis of L2 marine products highlighted a high consistency.
For SLSTR instruments, we also performed an in-depth analysis of L1 data of both missions through Level3 binning. For SLSTR IR, we highlighted some non linear effect outside of nominal dynamic range for S7, some small non linearities for S8 and S9 channels, and significant differences in oblique view for S8 and S9 for cold BTs that may indicate some S3A straylight below 240 K. For SLSTR VIS/SWIR, we highlighted some remaining offsets based on vicarious calibration results in both S3A and S3B.
For the Surface topography part of the mission, the different analyses showed a very good agreement between the missions. The Sea Surface Height Anomaly differences between both missions have been analysed in the different instrumental configurations: LRM mode, SAR mode open loop, SAR mode closed loop. Observed differences are small with mean bias less than 2 cm and standard deviation around 4 cm. The agreement on the significant wave height is excellent except for unexpected jumps at Open Loop / Closed Loop transitions. A difference of 0.5 dB on Ku-band backscattering coefficient, implying a bias in wind speed, has been observed, explained and will be corrected in the future processing baseline. The initial difference of 7 mm on the wet tropospheric correction based on MWR data will be corrected thanks to the in-flight calibration performed during the commissioning phase.
For the analysed instruments on board, this tandem phase has been very useful to assess effectively and rapidly the performance of both missions (S3A and S3B). It has allowed to check the consistency between A and B products, to identify/characterize the differences between A and B missions, to analyse the differences while operating A and B in different operating modes, to identify errors/anomalies in the processing of either A or B, and finally to improve our knowledge on measurement uncertainties.
The feedback of this tandem phase is very positive and should be taken into account for the future S3C/D missions.

Area D - Board 406

We compare remote sensing reflectance (Rrs) products of OLCI on Sentinels 3A and 3B with MODIS-Aqua, VIIRS-NPP and VIIRS-JSPP1 and evaluate how the differences affect the estimation of chlorophyll-a concentration (Chla). We then evaluate the statistical differences between standard Chla products of these five satellite sensors and compare to in situ measurements collected by the California Cooperative Oceanographic and Fisheries Investigations (CalCOFI) and the California Current Ecosystem Long Term Research (CCE-LTER) programs. We evaluate those differences as functions of time lag, atmospheric conditions, distance to clouds and to coast, sun elevation, pixel location across the swath and in situ Chla level.

Area D - Board 407