FengYun meteorological satellites, as the major components of the world’s earth observation constellations, play a crucial role in global meteorological monitoring, yet they are confronted with increasing demands driven by technological progress and higher user expectations. Meanwhile, small satellite constellations have demonstrated remarkable potential in enhancing meteorological observation capabilities through rapid deployment, cost efficiency, and high-density observations. This study introduces FengYun virtual constellation, namely ‘FengYun+’, an integrated system that consists existing FengYun backbone satellites with complementary small satellites. The small satellite framework is designed to fulfill three main goals: new technology verification, thematic earth system observation, and networked collaborative observation. These small satellites encounter key technological challenges related to miniaturization, systematization, rapid processing, and standardization across satellite platforms, payloads, inter-satellite interconnection, and data processing. The collaborative architecture of the virtual constellation, enabled by inter-satellite connectivity, dynamic task allocation, and AI-enhanced processing, supports multi-element, high-frequency, and near real-time global meteorological services. Collaboration with international programs ensures alignment with global meteorological constellation. FengYun+ will advance the Sustainable Development Goals (SDGs) and Early Warnings Initiatives of United Nations (UN) by delivering critical data for weather prediction, climate research, and disaster mitigation.

The bottomside thickness parameter (B-0) is a critical component for accurately representing electron density profiles in the International Reference Ionosphere (IRI) model. This study presents a comprehensive validation and global analysis of B-0 derived from the Fengyun-3 (FY-3) GNSS radio occultation (RO) mission during the high solar activity period of 2022-2024. By integrating observations from the FY-3 constellation, this work effectively complements the data coverage of FORMOSAT-7/COSMIC-2 (F7/C2) by extending analysis to mid- and high-latitude regions. Validation against global digisonde measurements demonstrates that FY-3 derived B-0 achieves high reliability, yielding correlation coefficients exceeding 0.86 and an RMSE of approximately 20 km in low and midlatitudes. Comparisons with three IRI-2020 model options reveal that the ABT-2009 option offers the best overall agreement with observations, particularly in reproducing hemispheric asymmetries, whereas the Gul-1987 and Bil-2000 options exhibit notable deficiencies in capturing geomagnetic modulation and spatial variability. Global morphological analysis identifies a synchronized but inversely correlated relationship between B-0 and peak electron density (NmF2) in the equatorial ionization anomaly region. Furthermore, distinct longitudinal structures are observed, characterized by wavenumber-4 patterns during equinoxes, and wavenumber-2 and -3 patterns during summer and winter solstices, respectively. In high latitudes, observations suggest a B-0 enhancement near the South Magnetic Pole during the Southern Hemisphere winter, which likely reflects the influence of geomagnetic control. These findings confirm the utility of FY-3 RO data for characterizing the global ionosphere and offer valuable constraints for future refinements of the IRI model.

This article introduces the Chinese Tianmu-1 Global Navigation Satellite System Reflectometry (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 (CP) 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 (TCs) and flooding. Finally, this article, 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 (SNR) and reflectivity shows that the dual-polarimetric observations follow the trend of theoretical models and hold promise for advancing land remote sensing.

The Fengyun-3 G (FY-3 G) satellite equipped with a Global Navigation Occultation Sounder II (GNOS-II) is capable of tracking BeiDou Navigation Satellite System (BDS) and Global Positioning System (GPS) dual-frequency signals by its zenith precise orbit determination antenna. Currently, most global navigation satellite system (GNSS)-related studies derived the topside total electron content (TEC) only by GPS signals, while there are relatively few studies using BDS signals. In addition, as FY-3 G satellite operates at a lower orbit with an altitude of 410 km, it is closer to the pivotal regions of the ionosphere, particularly the F2 layer. This grants it the capability of capturing dynamic changes and intricate details of the ionosphere with enhanced resolution and sensitivity. In this article, we derive the topside TEC from the six months of FY-3 G GNOS-II data and analyze the performance of differential code bias (DCB). The final retrieved topside BDS and GPS TEC has also been evaluated. The internal assessment of the FY-3 G receiver DCB is conducted by evaluating the long-term stability, which shows that the standard deviation (STD) of the FY-3 G receiver DCB reaches 0.86 TECU for the GPS C1C-C2W observation type and 0.83 TECU for BDS B2I-B6I type. The FY-3 G topside slant TEC (STEC) shows a high agreement with the space weather and atmospheric response mission TEC product, with a mean bias of 0.836 TECU and STD of 2.044 TECU. A comparative analysis in FY-3 G STEC between BDS and GPS satellites reveals no significant systematic bias, and the TEC retrieval performance of BDS-3 is superior to that of BDS-2. Moreover, the ionospheric characteristics of the monthly average FY-3 G topside vertical TEC map is consistent with the International GNSS service Global Ionospheric Map product in latitude, season, and local time. Above all, this research shows that FY-3 G GNOS-II can provide high-quality topside TEC observations, which can serve as a valuable dataset for research in ionospheric and plasma sciences.

The spatial–temporal sampling errors arising from the differences in geographical locations and measurement times between co-located Global Navigation Satellite System (GNSS) radio occultation (RO) and radiosonde (RS) data represent systematic errors in the three-cornered hat (3CH) method. In this study, we propose a novel spatial–temporal sampling correction method to mitigate the sampling errors associated with both RO–RS and RS–model pairs. We analyze the 3CH processing chain with this new correction method in comparison to traditional approaches, utilizing Fengyun-3E (FY-3E) GNSS Occultation Sounder II (GNOS II) RO data, atmospheric models, and RS datasets from the Hailar and Xisha stations. Overall, the results demonstrate that the improved 3CH method performs better in terms of spatial–temporal sampling errors and the variances of atmospheric parameters, including refractivity, temperature, and specific humidity. Subsequently, we assess the error variances of the FY-3E GNOS II RO, RS and model atmospheric parameters in China, in particular the northern China and southern China regions, based on large ensemble datasets using the improved 3CH data processing chain. The results indicate that the FY-3E GNOS II BeiDou navigation satellite system (BDS) RO and Global Positioning System (GPS) RO show good consistency, with the average error variances of refractivity, temperature, and specific humidity being less than 1.12%2, 0.13%2, and 700%2, respectively. A comparison of the datasets from northern and southern China reveals that the error variances for refractivity are smaller in northern China, while temperature and specific humidity exhibit smaller error variances in southern China, which is attributable to the differing climatic conditions.

With long-term mission planning extending well into the next decade, the Chinese FengYun-3 (FY-3) meteorological satellite series acts as an important backbone for the provision of radio occultation (RO) data for climate science and numerical weather prediction (NWP). Independent evaluation is an essential part for considerate application of the data. Here we present results of the first cross-evaluation of RO low-level data, obtained by the GNSS occultation sounder (GNOS) onboard of FY-3 satellites, from two independent processing chains (Wegener Center, Graz; National Space Science Center–National Satellite Meteorological Center, Beijing). We analyzed and compared the orbit solutions, atmospheric excess phase, and bending angle profiles, for 3-month evaluation periods of FY-3C and FY-3D data in 2014 (FY-3C only) and 2019. We found the orbit data to be within requirements for high-quality RO data processing. Cross-comparison at excess phase level as well showed high-quality consistency overall, while some differences were revealed in the closed-loop to open-loop transition of FY-3D data. The cross-evaluation of bending angle profiles against European Centre for Medium-Range Weather Forecasts (ECMWF) forecast data also showed good agreement and adequate quality in general, while some distinct deviations occurred below 26 km altitude in the early FY-3C rising event data in 2014. The results underscore the high basic quality of the GNOS data for applications in climate research and NWP. At the same time they valuably inform subsequent work to resolve remaining biases, improve current processing algorithms, and include data from FY-3E, FY-3F, and FY-3G satellites.

The global navigation satellite system radio occultation (GNSS-RO) is an important means of space-based meteorological observation. It is necessary to test the Global Navigation Satellite System Occultation signal receiver on the ground before the deployment of space-based occultation detection systems. The current approach of testing the GNSS signal receiver on the ground is mainly the mountaintop-based testing approach, which has problems such as high cost and large simulation error. In order to overcome the limitations of the mountaintop-based test approach, this paper proposes an accurate, repeatable, and controllable GNSS atmospheric occultation simulation system and builds a load performance evaluation approach based on the ground-based GNSS atmospheric occultation simulation system on the basis of it. The GNSS atmospheric occultation simulation system consists of the visualization and interaction module, the GNSS-RO simulation signal generation module, the GNSS-RO simulator module, the GNSS-RO signal receiver module, and the GNSS-RO inversion and evaluation module, combined with the preset atmospheric model to generate GNSS-RO simulation signals with a high degree of simulation, and comparing the atmospheric parameters of the inversion performance of the GNSS-RO signal receiver with the parameters of the preset atmospheric model to obtain the error data. The overall performance of the GNSS-RO signal receiver can be evaluated based on the error information. The novel approach to evaluate the GNSS-RO signal receiver performance proposed in this paper is validated by using the FY-3E (FengYun-3E) receiver qualification parts that have been verified in orbit, and the results confirm that the approach can meet the requirements of the GNSS-RO receiver performance test. This study shows that the novel approach to evaluate the GNSS-RO signal receiver performance in terms of the ground-based atmospheric occultation simulation system can efficiently and accurately be used to carry out the receiver test and provides an effective solution for the ground-based test of GNSS-RO signal receivers.

Higher-order residual ionospheric errors in Global Navigation Satellite System (GNSS) radio occultation (RO) data can induce a systematic residual ionospheric bias (RIB) in RO bending angles, which can impact stratospheric climate monitoring. The main RIB causes are the ray path splitting of dual-frequency GNSS signals, the electron density distribution along the ray path and at the RO receiver location, and the geomagnetic field. In this study, we investigate the ionospheric and geomagnetic effects on RO-retrieved stratospheric bending angle and temperature profiles by inspecting the kappa and bi-local RIB correction methods, the current state-of-the-art methods, using multiple RO satellite mission data in different ionospheric and geomagnetic conditions. We find that, globally, the layer mean/median RIBs of the kappa and bi-local correction methods exhibit similar behaviors for different RO missions: the estimated bending angle RIBs reach about -0.025/-0.01 and -0.024/-0.008 mu rad in the upper (40-45 km) and lower (30-35 km) stratosphere, respectively, while the temperature RIBs reach about -1.0/-0.3 and -0.2/-0.1 K in these layers. However, in the equatorial day time region, the RIB statistics of the mission results diverge. Both the kappa and bi-local RIBs increase in magnitude with increasing ionization and geomagnetic field strength but show no increase with an increasing degree of ionospheric asymmetry. Overall, the more refined bi-local method has higher capacity than the comparatively simple kappa method to represent the variability of the ionospheric and geomagnetic conditions that affect RO events. This is important for regional-scale studies, where the geomagnetic term can be of key relevance. Ionospheric, geomagnetic and orbit-height effects play a role for ionospheric errors in radio occultation bending angle and temperature Residual ionospheric biases (RIBs) in bending angle and temperature were modeled for these effects using the kappa and bi-local methods The kappa and bi-local methods show larger RIBs for higher ionization and geomagnetic activity and clear deviations in equatorial daytime

Objectives: Using global navigation satellite system（GNSS） reflectometry（GNSS-R） signal to do remote sensing research has become a hotspot in recent years. We has simulated Fengyun 3 E the Ⅱgeneration of GNSS occultation sounder（GNOSⅡ） GNSS-R parameters, which includes the average number of specular points,the maximum number of specular points,the average distance of specular point（SP）-GNSS,the average distance of SP-low Earth orbit（LEO）,the average path lost,the average incident angle on LEO and the average reflected angle on SP.Methods:The relationships between these simulated parameters and antenna parameters（antenna angle, beam width, and inclination direction） are analyzed, and the corresponding results are demonstrated using snow flake method.Results: Through the statistical analysis results, It can be easily learned the conclusion that the beam width has the maximum influence on all of these antenna observations, the antenna angle is the second, and the inclination direction has the minimum influence.Conclusions:The snow flake method mentioned is able to provide a new method for data analysis of GNSS-R especially space-borned ones, and may help to clear some obstacles which obstruct the successful development of corresponding payloads.

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 1.486 m/s and can reach 1.399 m/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 Occultation Sounder II (GNOS-II) payload onboard the Chinese Fengyun-3E (FY-3E) satellite is the world’s first operational spaceborne mission that can utilize reflected signals from multiple navigation systems for Earth remote sensing. The satellite was launched into an 836-km early-morning polar orbit on 5 July 2021. Different GNSS signals show different characteristics in the observations and thus require different calibration methods. With an average data latency of less than 3 h, many near real-time applications are possible. This article first introduces the FY-3E/GNOS-II mission and instrument design, then describes the extensive calibration methods for the multi-GNSS measurements, and finally presents application results in the remote sensing of ocean surface winds, land soil moisture and sea ice extent. Especially, the ocean surface wind product has been used in operational applications such as assimilation in the numerical weather prediction model and monitoring of tropical cyclones. Currently, GNOS-II has been carried by FY-3E, FY-3F (launched in August 2023) and FY-3G (launched in April 2023). It will be also carried by future follow-on FY series and a more complete multi-GNSS reflectometry constellation will be established.

The earth's climate has undergone significant changes due to the combined effects of natural changes and human activities. To understand the impact of climate change, the most fundamental work is to establish high-quality data required for climate purposes. Currently, the long series observations mainly come from satellites and sites. However, most satellite sensors are designed for short-term and imminent weather monitoring and numerical prediction, rather than long-term climate monitoring. To meet future research needs, more efforts are needed in data reprocessing such as satellite calibration and multi-source data fusion.Global Navigation Satellite System Radio Occultation (GNSS-RO) is a system that carries a receiver on low orbit satellite to receive radio signals transmitted by the global navigation satellite system. GNSS-RO detects the earth's atmosphere in a borderline manner during relative motion. When propagating in non-vacuum atmosphere, radio signals may appear bent or delay due to different atmospheric physical characteristics. After complex processing, physical parameters such as atmospheric temperature, humidity, and density can be inverted. Each receiver observes approximately 500 occultation events per day, which are almost randomly distributed on the earth and not affected by clouds and underlying surfaces. These data provide a source of observational information with high vertical resolution and long-term stability, extending from near surface to upper stratosphere. The original occultation observation is based on time and position measurements, needing no calibration, which has advantages in climate change study.The occultation receiver on FY-3C/3D/3E meteorological satellite can receive GPS and Beidou Navigation Satellite System (BDS) signals, and the records are almost nine years long. To analyze the accuracy and stability of temperature records from multiple radio occultation, the mean and standard deviation of the dry temperature of FY-3C/3D/3E GPS and BDS radio occultation are studied using ERA5 data. It demonstrats that the accuracy of the dry temperature profile is the highest between 200 hPa and 20 hPa and the error characteristics of GPS and BDS radio occultation are similar. The stability of the average temperature deviation of FY-3C GPS for 5-year time series is very good, which is -0.0055 K·a-1. After several algorithm improvements, the standard deviation of FY-3C GPS dry temperature decreased to about 1 K at the beginning of 2018. BDS radio occultation products are operationally provided since April 2021, and there is a good consistency between FY-3C/3E and between GPS and BDS radio occultation. Due to the algorithm adjustment at the beginning of 2021, the average deviation between FY-3D radio occultation and ERA5 data shows a significant jump. In general, the stability of multiple radio occultation dry temperature records is good and promising for climate change monitoring and research. It is necessary to carry out homogeneity reprocessing.

The FengYun-3E Global Navigation Satellite System (GNSS) occultation sounder II (FY-3E GNOS II) was launched on 5 July 2021. For the first time, based on the new GNOS II sensor, this mission realizes radio occultation (RO) and reflectometry observations using the navigation signals from the third-generation BeiDou System (BDS-3), and it is hence important to assess and analyze the BDS-3 remote sensing performances relative to other systems. Here, we assessed FY-3E GNOS II RO atmospheric retrievals by inter-comparing with corresponding data from the NCEP FNL global atmospheric analysis and FY-3D GNOS mission. The GNOS RO data quality and consistency of the different FY-3 meteorological satellites, i.e., FY-3D and FY-3E, as well as different GNSS systems (GPS, BDS-2, BDS-3) were analyzed. We find that the FY-3E GNOS II RO data exhibit better quality than FY-3D GNOS, particularly in the number, penetration height toward surface, and global coverage by BDS RO profiles, due to the integration of BDS-2 and BDS-3. Additionally, comparing with co-located NCEP FNL analysis profiles, the mean difference (and standard deviation) of the FY-3E GNOS II RO atmospheric refractivity profile retrievals is found to be smaller than 0.2% (and 1%), in the upper troposphere and lower stratosphere, from 5 to 30 km, and remains consistent at this accuracy and precision level with the FY-3D GNOS RO data. These features provide clear evidence for a high utility of the new GNOS II RO data for weather and climate research and applications.

FengYun-3E (FY3E), launched on 5 July 2021, is one of China’s polar-orbiting meteorological satellite series. The GNOS II onboard FY3E is an operational GNSS remote sensor that for the first time combines GNSS radio occultation (GNSS RO) and GNSS reflectometry (GNSS-R). It has eight reflection channels that can track eight specular points at the same time, receiving reflected signals from multiple GNSS systems, including GPS, BeiDou and Galileo. The basic GNSS-R output generated by GNOS II is a 122 × 20 non-uniform delay-Doppler map whose high resolution portion captures more information near the specular point. This paper introduces the GNSS-R aspect of the FengYun-3E GNOS II, including the instrument, power calibration and wind speed retrieval algorithm. Preliminary validation results for its first four months of data are also presented. After preliminary quality control, the overall wind speed error is less than 2 m/s at wind speeds below 20 m/s for data from both GPS satellites and BeiDou satellites when compared to the ECMWF reanalysis winds.

Global navigation satellite system (GNSS) radio occultation (RO) is a novel detection technique that can provide global ionospheric products with high vertical resolution, high precision, and low cost. In recent years, China has launched the Fengyun-3(FY3) series of meteorological satellites carrying the first RO payload to simultaneously receive global positioning system (GPS) and BeiDou navigation system (BDS) signals. In the accuracy assessment of RO products observed by GNSS occultation sounder (GNOS), the maximum F2-layer electron density (NmF2) of GPS occultation and BDS occultation has a standard deviation (std) of less than 20% in comparison with that of ionosondes. The std of F-layer worst case ionospheric scintillation index ( $S4_{\text{max}}^F$ ) between Constellation Observing System for Meteorology, Ionosphere, and Climate and FY3/GNOS is less than 0.1. The above results prove the high precision of FY3 ionospheric RO products. The RO products have been applied to preliminary scientific research and applications, e.g., the process of main phase and recovery phase of magnetic storms revealed by the NmF2 observed by FY3C/GNOS, the premidnight dynamics of F-layer strong scintillation during magnetic storms revealed by GNOS scintillation data, the ionospheric perturbation driven by Tonga volcano eruption revealed by FY3/GNOS, applications of the RO data for the research of sporadic E layers, evaluation of International Reference Ionosphere model in statistics and ionospheric climatological characteristics, etc. With the successive network observation and continuous deployment of FY3 meteorological satellites, the continuous improvement of GNOS payload and the BDS system, massive high-precision ionospheric RO products will be developed and show more significant value.

In radio occultation (RO) data processing and data assimilation, the forward model (FM) is used to calculate bending angle (BA) from refractivity (N). The accuracy and precision of forward modeled BA are affected by refractivity profiles and FM methods, including Abel integral algorithms (direct, exp, exp_T, linear) and methods of interpolating refractivity during integral (log-cubic spline and log-linear). Experiment 1 compares these forward model methods by comparing the difference and relative difference (RD) of the experimental value (forward modeled ECMWF analysis) and the true value (BA of FY3D RO data). Results suggested that the exp with log-cubic spline (log-cubic) interpolation is the most accurate FM because it has better integral accuracy (less than 2%) to inputs, especially when the input is lower than an order of magnitude of 1 × 10−2 (that is, above 60 km). By contrast, the direct induced a 10% error, and the improvement of exp T to exp is limited. Experiment 2 simulated the exact errors of an FM (exp) based on inputs on different vertical resolutions. The inputs are refractivity profiles on model levels of three widely used analyses, including ECMWF 4Dvar analysis, final operational global analysis data (FNL), and ERA5. Results demonstrated that based on exp and log-cubic interpolation, BA on model level of ECMWF 4Dvar has the highest accuracy, whose RD is 0.5% between 0–35 km, 4% between 35–58 km, and 1.8% between 58–80 km. By contrast, the other two analyses have low accuracy. This paper paves the way to better understanding the FM, and simulation errors on model levels of three analyses can be a helpful FM error reference.

The development of global navigation satellite systems (GNSSs) and multi-system compatible radio occultation (RO) techniques provides favorable conditions and opportunities for increasing the number of occultation events and improving their spatiotemporal coverage. The performance of the multiple GNSS RO event number, spatiotemporal coverage, and uniformity need assessments by robust and functional approaches. Firstly, a simulation system of RO events, which took the orbit perturbations into account, was established, and the concepts of global coverage fraction and uniformity of RO events were defined. Secondly, numerical experiments were designed to analyze the GNSS RO performances of a single-receiving satellite and satellite constellations under the condition of using current multiple GNSSs as transmitting satellite systems, in which the Earth was divided into 400 × 400 km2 grids. Finally, the number, timeliness, global coverage fraction, and uniformity of GNSS RO events for a single-receiving satellite and receiving satellite constellations were numerically calculated and analyzed. The results showed that ➀ multiple GNSS integration improved the number of GNSS RO events and their global coverage for a single polar-orbit satellite significantly, e.g., the 24 h multiple GNSS RO event number was about 7.8 times that of the single GNSS system, BeiDou navigation satellite system-3, while the corresponding 24 h global coverage fraction increased nearly 3 times. ➁ In the multiple GNSS integration scenario, the constellation composed of 12 polar-orbit low-Earth-orbit satellites achieved 100% RO event global coverage fraction within 24 h, of which the RO detection capability was comparable to the 100 Spire weather satellites and global positioning system (GPS) RO system. ➂ More GNSS RO events of the polar-orbit constellations were distributed in the middle- and high-latitude zones. Therefore, multiple GNSS integration could increase the RO event number and global coverage significantly to benefit the global climate monitoring and global numerical weather prediction, and the polar-orbit constellations were more favorable to atmospheric detection in middle- and high-latitude regions.

The FY3C/GNOS launched in 2013 can only detect the scintillation of GPS navigation signals, while the FY3D satellite launched in 2017 supports the scintillation observation of BeiDou navigation satellite system (BDS) signals, thus enabling simultaneous detection of GPS and BDS ionospheric scintillation. This work presents a preliminary accuracy evaluation of BDS ionospheric scintillation observed by the global navigation satellite system (GNSS) occultation sounder (GNOS) onboard the FY3D satellite, to support long-term ionospheric scientific applications based on the BDS system. First, the F-layer worst-case (maximum) amplitude scintillation index ($S4_{\text{max}}^F$) of FY3D-BDS, FY3D-GPS, and COSMIC-GPS are, respectively, quality-controlled, and the spatial-temporal matching of $S4_{\text{max}}^F$ between FY3D-BDS and FY3D-GPS/COSMIC-GPS is performed. Then, based on the statistical deviation (std) of the $S4_{\text{max}}^F$ data pairs, the data accuracy of FY3D-BDS $S4_{\text{max}}^F$ relative to FY3D-GPS and COSMIC-GPS is obtained. The results show that the std of the $S4_{\text{max}}^F$ differences in data pairs between FY3D-BDS and FY3D-GPS/COSMIC-GPS is smaller than 0.1, which proves the high precision of BDS ionospheric scintillation detection of FY3D/GNOS. Meanwhile, the std of the $S4_{\text{max}}^F$ differences between FY3D-BDS and FY3D-GPS/COSMIC-GPS at nighttime is higher than that at daytime, and the std in the mid-latitude region is lower than that in the low-latitude and high-latitude regions.

With the development of computer technology and expanding environmental issues, machine learning has received more and more attention in the field of weather forecasting. Global Navigation Satellite System-Radio Occultation(GNSS-RO) technology is a kind of remote sensing technology. This investigation proposes an alternative to numerical weather forecasting model. The new method is based on machine learning utilizing GNSS-RO data to forecast the wind field in the Beijing-Tianjin-Hebei region of China. The dataset including temperature, humidity, pressure, wind speed and direction was obtained by numerical calculation in terms of historical monitoring data in Beijing-Tianjin-Hebei region. Then the models of wind fields forecasting based on machine learning were established with different neural network including Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN) and Deep Neural Networks (DNN). The prediction performance of different models was analyzed. The results demonstrate that LSTM and CNN have better performance on predicting the wind field than Deep Neural Networks. The wind speed error is about 1.4m/s, and the wind direction error is about 30°. Moreover, the time required for neural network to predict a new sample is about 1 second, which is only 0.2% of the prediction time compared with numerical model. Finally, the machine learning model can be used to predict the wind field effectively, with GNSS-RO data as the input in application. This paper pro-vides a new method in sight to use machine learning to forecast the regional wind field utilizing GNSS-RO data.