This paper introduces the Chinese Tianmu-1 GNSS-R constellation of 22 small satellites launched in 2023–2024 and comprehensively evaluates the latest version of the in-orbit data. First, the mission design and instrument technology are described, which largely builds on the Fengyun-3/GNOS-II missions. Notable innovations include full-GNSS compatibility and dual-polarization antenna. Then, the spatiotemporal characteristics of the constellation are analyzed—specifically, coverage percentage and mean revisit time at different latitudes. Next, the accuracy of its science products including ocean surface winds and land soil moisture has been assessed, with two application cases demonstrating the mission’s utility for monitoring tropical cyclones and flooding. Finally, this paper for the first time evaluates Tianmu’s polarized observations including Horizontal (H), Vertical (V), Left-Hand Circularly-Polarized (LHCP) and Right-Hand Circularly-Polarized (RHCP). Analysis of the signal-to-noise ratio and reflectivity shows that the dual-polarimetric observations follow the trend of theoretical models and hold promise for advancing land remote sensing.

The GNOS-II global navigation satellite system reflectometry (GNSS-R) Level 1 products from FENGYUN-3E, which include delay-Doppler map (DDM) and normalized bistatic radar cross section (NBRCS), significantly affect the inversion accuracy of geophysical parameters such as sea wind and soil moisture. This study uses data collected from FY-3E GNOS-II GNSS-R and data from ECMWF ERA-5 sea surface wind speed, covering the period from July 2022 to June 2023. In addition, CYGNSSv3.1 Level 1 data are used as a reference to evaluate NBRCS, observation geometry, GNSS satellite types, equivalent isotropic radiated power (EIRP) of GNSS and peak signal-to-noise ratio (SNR). The results show a strong correlation between FENGYUN-3E and approximately 1.5 million colocated cyclone global navigation satellite system (CYGNSS) specular points, with an overall NBRCS correlation coefficient of 0.882. Compared to CYGNSS, the FY-3E/GNOS-II specular reflector point offers broader coverage and the capability to process reflected signals from GPS, BDS, and GALILEO systems. However, FY-3E operates at a higher orbital altitude, resulting in reduced range-corrected gain (RCG), lower peak SNR, and a more scattered distribution of NBRCS. FY-3E successfully achieves stable NBRCS measurements across multiple GNSS systems. However, slight bias in NBRCS values is observed between different GNSS systems due to variations in EIRP calibration methods. These findings provide critical information for the practical application and future calibration of FY-3E Level 1 data.

Fengyun-3 (FY-3) series are operational satellite missions that can provide global GNSS-R observations using multiple GNSS systems. This abstract highlights the advancements in GNSS-R product development from FY-3E, FY-3F, and FY-3G. Currently available to the public are operational Level 1 and Level 2 wind products with a 25-km resolution. Additionally, a raw intermediate frequency product is accessible for scientific research. Upcoming product developments include the release of Level 2 12.5-km wind, Level 2 land soil moisture, Level 2 sea ice thickness, and Level 3 wind. Their algorithms and scientific impact will be discussed.

Reflected global navigation satellite system (GNSS) signals from Earth surface can be received by receivers at low Earth orbit for the remote sensing of geophysical parameters. While the technique has been studied for around 30 years, most early spaceborne GNSS reflectometry missions only adapted to receive GPS signals and the studies of reflected Galileo (GAL) signals in space are limited. The Navigation Satellite System Occultation Sounder II (GNOS-II) payload onboard the FY-3E satellite is the first mission that can operationally receive reflected GPS, BeiDou (BDS), and GAL signals at the same time. This letter presents the GAL reflectometry measurements from GNOS-II together with their calibration and wind speed (WS) retrieval methods. Results show that while GAL has a different signal modulation, the observables can be used to retrieve WSs using the same geophysical model functions (GMFs) of GPS after a dedicated calibration. The retrieved WSs from GAL also have a comparable accuracy as those from GPS and BDS.

In this study, we propose a method for wind speed retrieval using a random forest (RF) algorithm for Cyclone Global Navigation Satellite System (CYGNSS) data. We first compared CYGNSS data with Soil Moisture Active Passive (SMAP) data and found a certain deviation in the CYGNSS ”young sea, limited fetch” (YSLF) data product for high winds. Then, we used SMAP as the ”ground truth” to train an RF model and applied it to the wind speed retrieval of CYGNSS data. The experimental results show that using the RF algorithm for wind speed retrieval can eliminate noise in the CYGNSS YSLF wind speed data and improve retrieval accuracy. In addition, we explored the impact of different input parameter combinations on model performance and found that using an 11-parameter model in CYGNSS wind speed retrieval can achieve optimal performance. This can provide valuable reference for rapid near-real-time retrieval of tropical cyclones using CYGNSS.

With low signal power, Global Navigation Satellite System (GNSS) reflections in space are vulnerable to radio frequency interference (RFI) across the globe which can contaminate observations of GNSS reflectometry (GNSS-R) and introduce bias in the retrieved geophysical parameters. This paper presents a comprehensive analysis of the RFI in the Fengyun-3E Navigation Satellite System Occultation Sounder II (FY-3E/GNOS-II) GNSS-R data over the ocean. It was found that major sources of interference are the reflections of regional GNSS and Satellite-Based Augmentation Systems (SBAS), and external L-band interference signals from the ground. Furthermore, interference can have different impacts on the observations through increasing the noise power, decreasing the signal-to-noise ratio (SNR) or biasing observables depending on the type of interference and retrieval method. Some geography-dependent biases in the retrieved wind speeds were found to be related to RFI. An RFI mitigation method is then presented to calibrate observables and reduce geographical wind speed bias for the operational retrieval system. The bias and root-mean-square error of retrieved wind speeds from a region that is affected by RFI were reduced from 1.26, 2.78 to 0.04, 1.59 m/s, respectively.

Global Navigation Satellite System Reflectometry (GNSS-R) is a burgeoning remote sensing observation technology that can retrieve global sea surface wind speeds using satellite signals reflected from the sea surface. The improvement of data quality and the accumulation of data volume of this technology provides data support for constructing interdisciplinary-based retrieval models. This article constructs a hybrid deep neural network model based on deep learning for wind speed retrieval, which can receive and perform feature mining on the entire delay waveform while simultaneously supporting multiple auxiliary features input and achieving joint fitting. Then a bias correction method based on cumulative distribution function (CDF) matching is introduced to mitigate bias, especially at high wind speeds. We verify the contribution of different attribute features in wind speed retrieval by designing a feature ablation analysis. The fluctuation variation of the retrieval accuracy in the time dimension and the retrieval results distribution in space are compared and analyzed. The root mean square error (RMSE) of retrieval results is 1.486m/s and can reach 1.399m/s under the 94.25% wind speed condition. After bias correction based on CDF matching, the retrieval accuracy at high wind speed is improved by 7.19%. Besides, this model also has good temporal stability and can reproduce large-scale wind fields while effectively mitigating retrieval bias on a global scale, showing great potential for operational applications.

The Global Navigation Satellite System Reflectometry (GNSS-R) technique can measure ocean surface winds and other geophysical parameters from forward scattered GNSS (GPS, BeiDou, Galileo, etc.) signals. However, most early spaceborne missions only captured GPS signals while the study of spaceborne reflectometry using BeiDou signals (BDS-R) is limited. The GNOS-II payload onboard China’s FY-3E satellite has been operationally collecting a large number of BDS-R data since July 10, 2021. BDS is different from GPS in the orbit, signal frequency, chipping rate and Effective Isotropic Radiated Power (EIRP). Furthermore, BDS satellites have different generations and orbits. This paper, for the first time, comprehensively characterizes the spaceborne BDS-R observations over the ocean using the FY-3E GNOS-II data from different BDS satellite types including BDS-2 IGSO, BDS-2 MEO, BDS-3 IGSO and BDS-3 MEO. Their spatial coverage, spatial resolution, effective scattering area, incidence angle, range corrected gain, EIRP and signal-to-noise ratio have been analyzed and compared to those of GPS-R. Spaceborne BDS-R shows a lot of uniquenesses compared to GPS-R. The BDS-R observables are then calibrated separately for each type. An intercalibration is also applied to correct extra calibration errors. After the calibration, BDS-R observables and retrieved winds are consistent between each type and compared to those of GPS-R. Calibrated observables from FY-3E GNOS-II are also evaluated by comparing them to those measured by the Cyclone Global Navigation Satellite System (CYGNSS) mission and a theoretical model. The results of this paper can provide a reference for future BDS-R studies and spaceborne GNSS-R mission design.

Delay-Doppler maps (DDMs) are generally the lowest level of calibrated observables produced from global navigation satellite system reflectometry (GNSS-R). A forward model is presented to relate the DDM, in units of absolute power at the receiver, to the ocean surface wind field. This model and the related Jacobian are designed for use in assimilating DDM observables into weather forecast models. Given that the forward model represents a full set of DDM measurements, direct assimilation of this lower level data product is expected to be more effective than using individual specular-point wind speed retrievals. The forward model is assessed by comparing DDMs computed from hurricane weather research and forecasting (HWRF) model winds against measured DDMs from the Cyclone Global Navigation Satellite System (CYGNSS) Level 1a data. Quality controls are proposed as a result of observed discrepancies due to the effect of swell, power calibration bias, inaccurate specular point position, and model representativeness error. DDM assimilation is demonstrated using a variational analysis method (VAM) applied to three cases from June 2017, specifically selected due to the large deviation between scatterometer winds and European Centre for Medium-Range Weather Forecasts (ECMWF) predictions. DDM assimilation reduced the root-mean-square error (RMSE) by 15%, 28%, and 48%, respectively, in each of the three examples.

Global Navigation Satellite System Reflectometry (GNSS-R) observations from the Cyclone Global Navigation Satellite System (CYGNSS) mission are expected to improve numerical weather prediction (NWP) models. Level 1 GNSS-R observables, delay-Doppler maps (DDMs), contain information that is lost in producing the Level 2 wind speed retrievals at the specular point. DDMs could, therefore, prove to be a better observable for data assimilation. In this study, assimilation of GNSS-R DDMs into both global and regional NWP models is demonstrated. In the global case, DDM assimilation shows improvements on the European Centre for Medium-Range Weather Forecasts (ECMWF) background for one month of data. In the regional case, DDM assimilation shows its impact on the hurricane structure and intensity. Results using a two-dimensional variation analysis method (VAM) are presented. A plan for an Observation System Experiment (OSE) experiment is proposed.

Spaceborne Global Navigation Satellite System Reflectometry (GNSS-R) measurements of ocean winds present a challenge because the surface wind speed cannot be assumed homogeneous over the large glistening zone observable from orbit. The 25-km resolution requirement of the Cyclone GNSS (CYGNSS) mission limits wind speed retrievals to using only a small fraction of the delay-Doppler map (DDM) near the specular point. This paper presents a new method to invert a larger portion of the DDM and estimate the wind field within a swath defined by the maximum delay. An extended Kalman filter (EKF) is applied to combine a series of DDMs and estimates the wind speed on a uniformly gridded ocean surface exploiting the large overlap between sequential looks. Wind retrievals at the specular point using simulated data generated using wind fields from two hurricanes (Danielle and Earl, 2010) were found to meet the CYGNSS measurement requirements (2 m/s for U20 m/s) and perform better than a typical single-point observable. Wind retrievals were also found to meet these requirements below 20 m/s within a 90-km swath and meet them for hurricane-force winds ($17 \times 11$) DDM, demonstrating benefits of sequential processing within a limited data budget. Delay-Doppler ambiguities introduced noticeable artifacts in circumstances where the wind varies asymmetrically within the observed ($90\times 90$km) surface area but did not appear to affect specular point wind retrievals.

The Cyclone Global Navigation Satellite System (CYGNSS) constellation was launched for the purpose of improving tropical cyclone forecasts using GNSS Refiectometry (GNSS-R). CYGNSS wind speed estimates have been based on only a small window of the Delay-Doppler Maps (DDM) due to the resolution requirement. Direct assimilation of DDM data into a forecast model is an alternative approach, that could take advantage of contribution to the DDM from regions on the ocean away from the specular point. This paper will present a generalized forward model for assimilation of DDMs into a weather model. The forward operator and Jacobian matrix are derived and structured for use in data assimilation systems. The model has also been assessed using CYGNSS Level 1 data from the 2017 Hurricane season.