The CYGNSS constellation of eight satellites was successfully launched in December 2016 into a low inclination (tropical) Earth orbit. Each satellite carries a four-channel bistatic radar receiver that measures GPS signals scattered by the Earth surface. Over ocean, surface roughness, near-surface wind speed, sea surface height, and microplastic concentration can be estimated from the measurements. Over land, estimates of near-surface soil moisture and imaging of inland water body extent are possible. The measurements are able to penetrate through all levels of precipitation due to the low microwave frequency at which GPS operates. The density and revisit time of sampling afforded by the number of satellites in the constellation makes possible the detection of tropical cyclone intensification, the diurnal cycle of tropical winds, and day-to-day changes in flood inundation, among other possibilities. Engineering commissioning of the constellation was completed in March 2017 and the mission is currently in its science operations phase.

Level 2 science data products have been developed for near surface (10 m referenced) ocean wind speed and ocean surface roughness (mean square slope), latent and sensible heat fluxes at the air/sea interface, and ocean microplastic number density. Level 3 gridded versions of the L2 products have also been developed, as well as an L3 soil moisture product and an L3 storm centric gridded product for tropical cyclone overpasses. Assimilation of L2 wind speed data into hurricane weather prediction models and of L3 soil moisture into regional and global hydrological models have also been developed.

A number of new measurement capabilities are also in development, including the generation of inland water masks, flood inundation imaging, and river width and slope determination for the estimation of streamflow rate.

An overview and the current status of the mission will be presented, together with an update on the status of science data products produced by the mission and highlights of recent scientific results.

Spire Global, Inc., a leading global provider of space-based data, analytics, and space services, designs, builds, and operates a growing constellation of Earth observation nanosatellites in low-Earth orbit (LEO). Spire satellites use a state-of-the-art GNSS receiver to observe the Earth in a variety of ways: GNSS radio occultation (GNSS-RO) for atmospheric sounding, ionospheric slant total electron content and scintillation for space weather monitoring, GNSS-Reflectometry (GNSS-R) for soil moisture, sea ice, and ocean wind remote sensing.

Spire’s GNSS-Reflectometry missions consist of 2 types of payload, conventional GNSS-R and grazing angle GNSS-R. The defining characteristics of the conventional GNSS-R are observations around nadir (elevation 20 to 90 degrees) for which the payload uses LHCP antennas and processing of Delay-Doppler Maps to map out the reflection power from incoherent surfaces. For the grazing angle GNSS-R (elevation angles 5 to 30 degrees), the payload uses RHCP antennas and processing tuned to coherent signals with outputs of phase and amplitude.

Since starting in mid-2018, and with support from the ESA Pioneer program, Spire has launched four conventional GNSS-R satellites of two separate designs: “Batch-1” and “Batch-2”. In 2019, Spire launched its first pair of “Batch-1” GNSS-R satellites into a 37 deg inclination orbit. These Batch-1 satellites adapted a heritage Spire GNSS-RO bus and the STRATOS GNSS receiver to perform GNSS-R with deployable LHCP GNSS-R antennas. The STRATOS GNSS receiver firmware and software was updated to perform onboard, real-time delay-Doppler map (DDM) processing. Following the success of the Spire GNSS-R Batch-1 satellites, a new “Batch-2” GNSS-R nanosatellite was designed that leveraged a new wide-band front-end version of the STRATOS GNSS receiver and an additional LHCP reflection antenna. A pair of Spire GNSS-R Batch-2 satellites were launched into polar Sun-synchronous orbit in January 2021.

Spire has demonstrated a number of firsts for a GNSS-R mission. Digital beamforming between the two nadir reflection antennas has been demonstrated to improve the system gain over single antenna observations. The first satellite-based GNSS-R DDMs generated from all L1 band GNSS signals except GLONASS have been generated to improve the coverage per receiver. Additionally, the concept of relative amplitude calibration among the direct and reflected signal channels has successfully compensated for transmission power differences, such as the GPS Flex-Power mode. The triple GNSS-R antenna on “Batch-2” achieves an approximate 2X increase in the quantity of viable land reflections over “Batch-1.

In addition to these conventional geometry GNSS-R satellites, Spire operates a constellation with RHCP GNSS antennas designed for GNSS-RO atmospheric observations. In 2019, the receivers were configured to additionally measure GNSS reflections at grazing angles, i.e., between 5 and 30 degrees elevation. As of November 2021, over 20 satellites continuously measure GNSS-R grazing angle reflection events in areas of high-coherence, i.e., calm waters and glaciated surfaces. With these, the feasibility has been demonstrated of phase delay altimetry using coherent reflections.


This presentation will describe the separate Spire GNSS-R missions and accomplishments and derived data products of ocean roughness, soil moisture, sea-ice classification and phase altimetry. Studies of our data are encouraged and they are freely available in agency portals NASA CSDAP and ESA Earthnet.

The GNOS II payload onboard the FengYun-3E (FY-3E) meteorological satellite launched on July, 5, 2021 is the upgraded GNSS remote sensor of GNOS onboard FY-3C and FY-3D. It has both the GNSS radio occultation (GNSS RO) and GNSS reflectometry (GNSS-R) functions that can monitor atmosphere, ionosphere and the Earth surface simultaneously. In particular, the observations of the FY-3E GNOS II sensor mainly involves the Earth’s atmospheric refractivity, temperature, humidity, pressure, ionospheric electron density profiles, and the ocean surface wind speed, which is an additional data product.

Firstly, this presentation will introduce the FY-3E GNOS II GNSS RO atmospheric and ionospheric results, comparing with the corresponding data from the models (e.g., ECMWF, NCEP, IRI, etc.), the other missions such as the MetOp, COSMIC, especially the FY-3C and FY-3D GNOS missions. The GNOS RO data quality consistency of different FengYun-3 meteorological satellites i.e., FY-3C/-3D/-3E, as well as different GNSS systems i.e., GPS and BDS will be analyzed.

Secondly, the retrieval algorithm and validation results of the ocean surface wind product will be presented. The GNSS-R L1 product of GNOS II is the 122x20 non-uniform delay-Doppler map. The delay resolution is 1/8 chip near the specular point and 1/4 chip away from the specular point. The Doppler resolution is 500 Hz. The L2 ocean surface winds are retrieved by geophysical model functions related to the delay-Doppler map average (DDMA) and Leading Edge Slope (LES) observables computed from the L1 DDM. The retrieved L2 winds are validated by comparing to the ECMWF model and HY-2B, HY-2C scatterometer winds. The results are also analyzed by different GNSS systems (GPS and BDS). Furthermore, the preliminary results for the retrieval of sea ice coverage and soil moisture will be presented. The sea ice coverage retrieval results are validated by the OSI SAF product and the soil moisture retrieval results are validated by the SMAP data.

Finally, the additional value of the combination of the GNSS RO and GNSS-R techniques in one payload will be investigated and analyzed.

Starting with TDS-1, and confirmed by the CYGNSS constellation, it was realized that GNSS reflections are very sensitive to scenes on land containing water, because they exhibit a strong coherent component in the specular direction. Soon a research community interested in detecting extent and monitoring dynamics of tropical inundations and wetlands and other flooded areas emerged. This communication reviews our team’s effort of the past five years. We first focused our efforts on modeling expected reflected signals from heterogeneous water scenes including water extent/depth, different vegetation and terrain type, and atmospheric conditions leading to winds, and leveraged the CYGNSS End-to-End Simulator (EES) to understand the role of coherent reflections, confounding variables, and error sources. Increased understanding of scattering mechanisms occurring in heterogeneous water scenes led us to develop alternative coherence detectors, also working with complex waveforms. Validation with SAR data is ongoing, specifically in partnership with the NISAR mission through use of the Yucatan Lake cal/val site and pre-launch data obtained from airplane SAR campaigns.
In parallel, we developed a wet-dry classifier based on DDM peak SNR to generate dynamic maps over the Everglades with CYGNSS data, starting with available water masks and compared consistently with in-situ data, showing seasonal variability of inundation extent, detectable also in the presence of obstructing vegetation, and sensitivity to water depth. Our analysis shows that dynamic changes can be monitored over small spatial scales at reasonable accuracy, and this motivated us to apply our analysis and retrieval algorithms in support of the needs of the community of users of flood data products, as opposed to the research community per se, by exploring the benefits of the CYGNSS data for use by practitioners in decision making. In collaboration with the Global Flood Working Group, we are developing and testing CYGNSS derived flood maps for a number of representative scenarios in terms of spatial and temporal scales, to compare with existing operational products for preliminary validation.
The presentation will share our preliminary recommendations for designing future GNSS instruments and observation strategies suitable for generating products for operational use in the management of inland water bodies and floods.

Recent ESA studies on precise altimetry using carrier phase information of GNSS reflected signals: suitability to New Space missions:

The signals transmitted by the Global Navigation Satellite Systems (GNSS) can be used for other applications beyond navigation and positioning. Earth remote sensing is one of the opportunistic applications of the GNSS, based on signals that bounce off the Earth surface as a bi-static radar (GNSS reflectometry, GNSS-R), or on signal refraction within the atmosphere in the so called radio occultation technique (GNSS RO).

The surface height, for altimetry, is one of the geophysical parameters that can be inferred from GNSS-R signals. It is obtained from the inversion of delay measurements: the time required by the signals to travel from the GNSS transmitter down to the surface and up again to the receiver antenna. This delay is measured as the delay of the overall ‘echo’ or ‘waveform’ (group-delay measurement) or through the evolution of the phase of the electromagnetic field that carries the GNSS modulations and information (carrier phase measurements). Given the narrow bandwidth of the GNSS modulations, the group-delay measurements tend to present poor precision compared to those achieved with monostatic and dedicated radar altimeters. Nevertheless, the carrier phase measurements are very precise, at a level of few cm after a few millisecond integration. GNSS-R Carrier Phase Altimetry (CaPA) has been proved in several experiments from ground-based, airborne and even spaceborne receiving systems [e.g. 1-4]. The drawback is that the reflecting surface must be smooth enough to enable coherent scattering and to preserve the carrier phase information. Earth surfaces such as the ocean – where altimetric measurements are required – tend to scatter GNSS signals in a diffuse regime, with a loss of carrier phase information.

The surface must be smooth with respect to the electromagnetic wavelength for the scattering to preserve the carrier phase information: the whole reflecting volume ought to be within the first Fresnel zone. Geometry plays a role: the smaller the incidence angle the thinner the vertical component of the Fresnel zone and the higher the chances that peaks and troughs of the ocean waves do not fit within it, thus resulting in diffuse scattering. As the incidence angle increases, the Fresnel zone thickens in its vertical component and eventually captures the whole sea surface wave structure. At these Grazing Angle (GA) geometries – low elevation angles of observation – the chances of coherent scattering increase. GNSS-R at these geometries for carrier phase altimetry will be called hereafter GA-CaPA, and its applicability to sea surface altimetry from spaceborne GNSS-R payloads was demonstrated in [5], using a limited set of CyGNSS raw data samples. Moreover, as the GA geometries are compatible with GNSS RO payloads, sea ice altimetric GNSS GA-CaPA tracks are regularly acquired by the Spire Global constellation of GNSS RO satellites, after an upgrade in their firmware [6].

In view of this promising new altimetric approach, a set of ESA studies have further investigated different aspects of GA-CaPA: during Q2-Q3 2021, a dedicated coastal experiment was installed and operated at the highest mountain in Majorca (Balearic Islands, Spain), to collect GNSS-R in GA geometries at two GNSS frequency bands and polarization states, as sea winds and sea waves conditions changed. Together with the deployment of an oceanographic buoy for sea state monitoring and the Sentinel-3 passes across the observational area for altimetric comparisons, the study aimed to understand the underlying conditions of this type of reflectometry and identify the major limitations. In parallel, another study collected and analysed selected raw intermediate frequency samples of the Spire GNSS RO spaceborne constellation. The acquisitions were programmed when tracks of reflected GA-CaPA were predicted within the GNSS RO signals, and these were co-incident with passes of Sentinel-3 radar altimetry tracks. A third study targets to build the theoretical body of the GNSS GA-CaPA including its validation with the former and new Spire raw data acquisitions.
The technique can be embedded in small low-consuming payloads. The ESA CubSat mission PRETTY, to be launched in 2022, is designed for GNSS-R altimetry at GA geometries, and as mentioned before, Spire Global already implemented this technique into its constellation of GNSS RO nanosatellites. The combined outcome of these three ESA projects, which will be presented, permits to get a more comprehensive understanding of the performance, systematic effects, underlying conditions and limitations of the GA-CaPA, its potential complementarity to the current meta-constellation of dedicated radar altimeters and its potential use in the frame of the New Space paradigm.

[1] Treuhaft et al., 2001, doi:10.1029/2001GL013815
[2] Cardellach et al., 2004, doi:10.1029/2004GL019775.
[3] Semmling et al., 2014, doi:10.1002/2013GL058725.
[4] Li et al., 2017, doi:10.1002/2017GL074513.
[5] Cardellach et al., 2020, doi:10.1109/JSTARS.2019.2952694.
[6] Nguyen et al, 2020, doi:10.1029/2020GL088308.

We have conducted two sets of observing system experiments (OSEs) to assess the impact of Spire data. First is an OSE assimilating additionally three months in 2020 of about 6,000 Spire occultations and about 5,000 COSMIC-2 occultations per day in the ECMWF and Met Office systems. We found, that adding Spire and COSMIC-2 data to the 4D-Var systems clearly improves forecast scores. For example, tropical forecast scores in humidity are very much improved by about 1.3% at 500hPa for COSMIC-2 compared to 0.5% for Spire at day 2 at ECMWF. Fits to independent observations sensitive to temperature, wind and humidity improve largely when adding Spire. For example, fits to radiosonde temperature observations at 150hPa improve by 1.2% and 4%, to microwave imager observations sensitive to total water vapour by 0.5% and 0.5%, and to wind observations at 850hPa from atmospheric motion vectors by 0.3% and 0.1% for ECMWF and Met Office, respectively.
The second set of OSEs assimilates Spire bending angles processed by EUMETSAT. One OSE assimilates up to 10,000 additional Spire occultations per day, which brings the total of GNSS-RO profiles to around 18,000, which is close the daily number targeted by the International Radio Occultation Working Group (IROWG). These experiments with a large volume of GNSS-RO observations show a large positive impact on the forecast score, demonstrating the potential power of this observation type. Furthermore, more subtle impacts from the assimilation of GNSS-RO can be detected more easily. One interesting feature through the addition of a large number of GNSS-RO data is the improvement in the fit to wind observations in the troposphere and stratosphere, especially in the Tropics. In addition, the assimilation of a high number of observations enables us to investigate the sensitivity of temperature and humidity forecast scores to the presence of GNSS-RO observations in the lower troposphere.