Zhiyu Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Zhiyu Zhang has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Fei Guo
Wuhan University
6 shared publications - Xiaohong Zhang
Chinese Antarctic Center of Surveying and Mapping, Wuhan University, Wuhan 430079, China
6 shared publications - Wentao Yang
School of Geodesy and Geomatics, Wuhan University, Wuhan, China
4 shared publications - Zheng Li
School of Geodesy and Geomatics, Hubei Luojia Laboratory, Wuhan University, Wuhan, China
4 shared publications - Yifan Zhu
School of Geodesy and Geomatics, Wuhan University, Wuhan, China
3 shared publications Dengkui Mei
Chinese Antarctic Center of Surveying and Mapping, Wuhan University, China
2 shared publications
Latest Publications
- Open Access
- 29 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19 (19391404) , 5447-5459
Evaluating the Performance of China’s Tianmu-1 GNSS-R Satellite Constellation Observations in Land and Ocean Remote Sensing Applications
China's Tianmu-1, equipped with the Global Navigation Satellite System Reflectometry (GNSS-R) payload, has formed a constellation of 22 small satellites since its batch launch in 2023. Since the TM-1 constellation has the largest number of GNSS-R satellites to date, TM-1 observations provide powerful samples from the Earth's surface. This study is the first to utilize ten months of TM-1 GNSS-R observations for land (soil moisture (SM), freeze-thaw (F/T), waterbody) and ocean (sea ice, sea surface winds, and significant wave height (SWH)) parameter retrievals. Compared with the reference satellite data, the root mean square error (RMSE) and correlation (R) of the TM-1 SM retrieval were 0.035 cm3/cm3 and 0.61, respectively. Compared with in-situ data, the RMSE and R of the TM-1 SM retrieval were 0.050 cm3/cm3 and 0.83, respectively. The accuracies of the TM-1 F/T retrievals were 96.5% and 92.0% compared with the reference satellite data and in-situ data, respectively. The accuracy of TM-1 water body detection exceeded 97% when compared to optical remote sensing products. TM-1 sea ice detection accuracy exceeds 98% compared to microwave radiometer products. The RMSE for TM-1 sea surface wind speed is 2.09 m/s with an R of 0.82, and the RMSE for SWH is 0.65 m with an R of 0.86. Additionally, the RMSE of TM-1 wind speed retrievals compared to NDBC buoy wind speeds is 1.84 m/s, with an R of 0.82. The RMSE of TM-1 SWH compared to the NDBC SWH is 0.46 m, with an R of 0.86. Both land and ocean remote sensing retrievals demonstrate the excellent performance of the TM-1 GNSS-R constellation. As a result, the TM-1 GNSS-R constellation holds promise as a valuable tool for Earth observation.
- 29 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19 (19391404) , 5447-5459
Open AccessEvaluating the Performance of China’s Tianmu-1 GNSS-R Satellite Constellation Observations in Land and Ocean Remote Sensing Applications
China's Tianmu-1, equipped with the Global Navigation Satellite System Reflectometry (GNSS-R) payload, has formed a constellation of 22 small satellites since its batch launch in 2023. Since the TM-1 constellation has the largest number of GNSS-R satellites to date, TM-1 observations provide powerful samples from the Earth's surface. This study is the first to utilize ten months of TM-1 GNSS-R observations for land (soil moisture (SM), freeze-thaw (F/T), waterbody) and ocean (sea ice, sea surface winds, and significant wave height (SWH)) parameter retrievals. Compared with the reference satellite data, the root mean square error (RMSE) and correlation (R) of the TM-1 SM retrieval were 0.035 cm3/cm3 and 0.61, respectively. Compared with in-situ data, the RMSE and R of the TM-1 SM retrieval were 0.050 cm3/cm3 and 0.83, respectively. The accuracies of the TM-1 F/T retrievals were 96.5% and 92.0% compared with the reference satellite data and in-situ data, respectively. The accuracy of TM-1 water body detection exceeded 97% when compared to optical remote sensing products. TM-1 sea ice detection accuracy exceeds 98% compared to microwave radiometer products. The RMSE for TM-1 sea surface wind speed is 2.09 m/s with an R of 0.82, and the RMSE for SWH is 0.65 m with an R of 0.86. Additionally, the RMSE of TM-1 wind speed retrievals compared to NDBC buoy wind speeds is 1.84 m/s, with an R of 0.82. The RMSE of TM-1 SWH compared to the NDBC SWH is 0.46 m, with an R of 0.86. Both land and ocean remote sensing retrievals demonstrate the excellent performance of the TM-1 GNSS-R constellation. As a result, the TM-1 GNSS-R constellation holds promise as a valuable tool for Earth observation.
- Open Access
- 18 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 18 (19391404) , 25039-25053
Daily 9-km Soil Moisture Retrieval With Reconstructed GNSS-R Observations
Spaceborne Global Navigation Satellite System-Reflectometry (GNSS-R) observations are effective for extensively monitoring soil moisture (SM). Reliance on gridded data from other remote sensing systems is the general way of GNSS-R SM retrieval. Increasing the grid size improves temporal coverage but decreases spatial resolution, which may mask changes in SM within the grid. Conversely, finer grids result in higher spatial resolution but less temporal resolution, leading to discontinuities in SM temporal dynamics. Therefore GNSS-R observation gridding cannot achieve high spatiotemporal resolutions simultaneously. To leverage the potential of GNSS-R technology in SM retrieval with a high spatiotemporal resolution, this study introduced a reconstruction method for spaceborne GNSS-R observations to generate gridded observations with both high temporal and spatial resolution. This method was used to build a pixel-by-pixel multivariate temporal fitting model that incorporates the characteristics of GNSS-R observations and the associated driving factors. Specifically, this study utilizes Cyclone GNSS (CYGNSS) sliding average observations to supplement daily trend observations and integrates ancillary data from other satellite missions as adjustments to daily variability, which generates spatiotemporally seamless CYGNSS observations. The reconstructed CYGNSS observations were then utilized to provide daily quasi-global SM retrievals on a 9-km grid. The root mean square error (RMSE) of the SM retrievals was 0.052 cm3/cm3, which are comparable to the accuracy of the SM retrievals conducted prior to reconstruction. Similarly, independent evaluations of local in situ sites demonstrated that the RMSE and correlation (R) of the SM retrieval were 0.051 cm3/cm3 and 0.77, respectively. The SM from the reconstructed CYGNSS observations also captured local wet and dry dynamics, which is consistent with the performance of the SM retrieved from the original CYGNSS observations. Notably, the temporal resolution was improved by 258% over the original CYGNSS observations at 9 km. Therefore, we argue that the method used in this study addresses the problem of mutual constraints between spatial and temporal resolutions in quasi-global CYGNSS SM retrieval without loss of SM retrieval accuracy. Furthermore, the observation reconstruction method developed in this study may be a promising reference for monitoring other geophysical parameters with high spatiotemporal resolution in the GNSS-R domain.
- 18 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 18 (19391404) , 25039-25053
Open AccessDaily 9-km Soil Moisture Retrieval With Reconstructed GNSS-R Observations
Spaceborne Global Navigation Satellite System-Reflectometry (GNSS-R) observations are effective for extensively monitoring soil moisture (SM). Reliance on gridded data from other remote sensing systems is the general way of GNSS-R SM retrieval. Increasing the grid size improves temporal coverage but decreases spatial resolution, which may mask changes in SM within the grid. Conversely, finer grids result in higher spatial resolution but less temporal resolution, leading to discontinuities in SM temporal dynamics. Therefore GNSS-R observation gridding cannot achieve high spatiotemporal resolutions simultaneously. To leverage the potential of GNSS-R technology in SM retrieval with a high spatiotemporal resolution, this study introduced a reconstruction method for spaceborne GNSS-R observations to generate gridded observations with both high temporal and spatial resolution. This method was used to build a pixel-by-pixel multivariate temporal fitting model that incorporates the characteristics of GNSS-R observations and the associated driving factors. Specifically, this study utilizes Cyclone GNSS (CYGNSS) sliding average observations to supplement daily trend observations and integrates ancillary data from other satellite missions as adjustments to daily variability, which generates spatiotemporally seamless CYGNSS observations. The reconstructed CYGNSS observations were then utilized to provide daily quasi-global SM retrievals on a 9-km grid. The root mean square error (RMSE) of the SM retrievals was 0.052 cm3/cm3, which are comparable to the accuracy of the SM retrievals conducted prior to reconstruction. Similarly, independent evaluations of local in situ sites demonstrated that the RMSE and correlation (R) of the SM retrieval were 0.051 cm3/cm3 and 0.77, respectively. The SM from the reconstructed CYGNSS observations also captured local wet and dry dynamics, which is consistent with the performance of the SM retrieved from the original CYGNSS observations. Notably, the temporal resolution was improved by 258% over the original CYGNSS observations at 9 km. Therefore, we argue that the method used in this study addresses the problem of mutual constraints between spatial and temporal resolutions in quasi-global CYGNSS SM retrieval without loss of SM retrieval accuracy. Furthermore, the observation reconstruction method developed in this study may be a promising reference for monitoring other geophysical parameters with high spatiotemporal resolution in the GNSS-R domain.
- 11 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 63 (01962892) , 1-10
An Algorithm for Downscaling SMAP Soil Moisture to 3 km Using CYGNSS Observations
Global soil moisture (SM) mapping at high spatial and temporal resolution contributes significantly to various hydrologic and meteorological researches. This work presents an algorithm for combining fine-resolution Global Navigation Satellite System Reflectometry (GNSS-R) observations from the Cyclone Global Navigation Satellite System (CYGNSS) and coarse-resolution SM estimates from the Soil Moisture Active Passive (SMAP) mission to estimate SM at 3 km resolution. In practice, the expression for downscaled 3 km SM is derived from the mathematical double-scale SM equations based on the linear assumption between SM and reflectivity, and the CYGNSS signal-to-noise ratio data are leveraged to compensate the differences in heterogeneity between the two scales. Experimental validation over 150 in-situ sites shows strong consistency between the SMAP/CYGNSS 3 km SM estimates and the in-situ measurements, with a median correlation coefficient of 0.806 and a median unbiased root-mean-square error of 0.038 cm3/cm3. The contributions of this work are twofold: 1) introducing the normalized signal-to-noise ratio to account for the deviations in double-scale coefficients, which depends on the variations in vegetation and surface roughness between 36 km and 3 km scales, without relying on any ongoing knowledge of ancillary data; and 2) achieving daily 3 km SM estimations at a quasi-global scale, and providing a new way for enhancing the temporal and spatial resolution of SMAP SM.
- 11 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 63 (01962892) , 1-10
An Algorithm for Downscaling SMAP Soil Moisture to 3 km Using CYGNSS Observations
Global soil moisture (SM) mapping at high spatial and temporal resolution contributes significantly to various hydrologic and meteorological researches. This work presents an algorithm for combining fine-resolution Global Navigation Satellite System Reflectometry (GNSS-R) observations from the Cyclone Global Navigation Satellite System (CYGNSS) and coarse-resolution SM estimates from the Soil Moisture Active Passive (SMAP) mission to estimate SM at 3 km resolution. In practice, the expression for downscaled 3 km SM is derived from the mathematical double-scale SM equations based on the linear assumption between SM and reflectivity, and the CYGNSS signal-to-noise ratio data are leveraged to compensate the differences in heterogeneity between the two scales. Experimental validation over 150 in-situ sites shows strong consistency between the SMAP/CYGNSS 3 km SM estimates and the in-situ measurements, with a median correlation coefficient of 0.806 and a median unbiased root-mean-square error of 0.038 cm3/cm3. The contributions of this work are twofold: 1) introducing the normalized signal-to-noise ratio to account for the deviations in double-scale coefficients, which depends on the variations in vegetation and surface roughness between 36 km and 3 km scales, without relying on any ongoing knowledge of ancillary data; and 2) achieving daily 3 km SM estimations at a quasi-global scale, and providing a new way for enhancing the temporal and spatial resolution of SMAP SM.
- Open Access
- 1 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 17 (19391404) , 3226-3235
Correction of the SMAP Sea Surface Brightness Temperature and Retrieval of Sea Surface Salinity Incorporating CYGNSS Observables
The correction of sea surface brightness temperature is crucial for improving the accuracy of sea surface salinity (SSS) retrieval by L-band microwave radiometer. However, the traditional method of correcting brightness temperature using only wind speed and significant wave height (SWH) is inadequate, as sea surface roughness is affected by multiple factors. The Global Navigation Satellite System Reflectometer (GNSS-R) observables, which directly respond to sea surface roughness, have been preliminarily validated in ground-based experiments for their potential to correct sea surface brightness temperature. Compared with ground-based GNSS-R, spaceborne GNSS-R has a wider coverage and can better support the brightness temperature correction of spaceborne L-band microwave radiometers. This article has preliminarily verified the correlation between cyclone GNSS (CYGNSS) observables and brightness temperature variations, and found that the incidence angle of the observable needs to be taken into account when retrieving SSS jointly with soil moisture active and passive (SMAP) and CYGNSS. A multilayer perceptron (MLP) model was established to assess the SSS retrieval performance of SMAP combined with different parameters. The results show that the retrieval performance based on the MLP model is better than that based on the geophysical model function model. Compared with joint wind speed and SWH, joint CYGNSS observables performs better in retrieving SSS. The root mean square error of retrieval salinity decreased from 0.58 to 0.46 psu, and the correlation coefficient (R) increased from 0.83 to 0.90. This provides reference for future joint retrieval of SSS using L-band microwave radiometers and spaceborne GNSS-R.
- 1 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 17 (19391404) , 3226-3235
Open AccessCorrection of the SMAP Sea Surface Brightness Temperature and Retrieval of Sea Surface Salinity Incorporating CYGNSS Observables
The correction of sea surface brightness temperature is crucial for improving the accuracy of sea surface salinity (SSS) retrieval by L-band microwave radiometer. However, the traditional method of correcting brightness temperature using only wind speed and significant wave height (SWH) is inadequate, as sea surface roughness is affected by multiple factors. The Global Navigation Satellite System Reflectometer (GNSS-R) observables, which directly respond to sea surface roughness, have been preliminarily validated in ground-based experiments for their potential to correct sea surface brightness temperature. Compared with ground-based GNSS-R, spaceborne GNSS-R has a wider coverage and can better support the brightness temperature correction of spaceborne L-band microwave radiometers. This article has preliminarily verified the correlation between cyclone GNSS (CYGNSS) observables and brightness temperature variations, and found that the incidence angle of the observable needs to be taken into account when retrieving SSS jointly with soil moisture active and passive (SMAP) and CYGNSS. A multilayer perceptron (MLP) model was established to assess the SSS retrieval performance of SMAP combined with different parameters. The results show that the retrieval performance based on the MLP model is better than that based on the geophysical model function model. Compared with joint wind speed and SWH, joint CYGNSS observables performs better in retrieving SSS. The root mean square error of retrieval salinity decreased from 0.58 to 0.46 psu, and the correlation coefficient (R) increased from 0.83 to 0.90. This provides reference for future joint retrieval of SSS using L-band microwave radiometers and spaceborne GNSS-R.
- Open Access
- 28 September 2018
- Published by MDPI AG in Remote Sensing
- Vol. 10 (10) , 1561
The Effects of Higher-Order Ionospheric Terms on GPS Tropospheric Delay and Gradient Estimates
Atmospheric delays, e.g., ionospheric delay and tropospheric delay, are the dominant error sources for the Global Navigation Satellite System (GNSS), especially for Precise Point Positioning (PPP). The common method for eliminating ionospheric delay is to form an ionosphere-free (IF) observable, which is a linear combination of observables on two frequencies such as GPS L1 and L2. As for the tropospheric delay, the dry component can be precisely corrected by empirical models, while the wet component is usually estimated as unknowns. However, the higher-order ionospheric (HOI) terms are not totally cancelled out in the (first-order) IF observable and as such, when not accounted for, they degrade the accuracy of other parameters. The impact of HOI corrections is well documented in the literature. This paper investigates the temporal effects of HOI terms on estimated tropospheric parameters, i.e., zenith tropospheric wet delay (ZWD) and north and east gradients. For this purpose, observations from over 100 stations with good global coverage were used considering various geographic and geophysical conditions. The results of numerical experiments show that HOI effects have a significant impact on the estimated tropospheric parameters, and the influence is dependent on location and time. The maximum differences of ZWD estimates reach over 20 mm during periods of activity such as solar storms and geomagnetic storms. Additionally, the north gradients are more likely to be affected by HOI effects compared with east gradients. In particular, the tropospheric gradient component is most affected for low latitude station during daytime. Additionally, the effects of bending errors and HOI terms on slant tropospheric delay at low elevation angles are much larger than those at high elevation angles.
- 28 September 2018
- Published by MDPI AG in Remote Sensing
- Vol. 10 (10) , 1561
Open AccessThe Effects of Higher-Order Ionospheric Terms on GPS Tropospheric Delay and Gradient Estimates
Atmospheric delays, e.g., ionospheric delay and tropospheric delay, are the dominant error sources for the Global Navigation Satellite System (GNSS), especially for Precise Point Positioning (PPP). The common method for eliminating ionospheric delay is to form an ionosphere-free (IF) observable, which is a linear combination of observables on two frequencies such as GPS L1 and L2. As for the tropospheric delay, the dry component can be precisely corrected by empirical models, while the wet component is usually estimated as unknowns. However, the higher-order ionospheric (HOI) terms are not totally cancelled out in the (first-order) IF observable and as such, when not accounted for, they degrade the accuracy of other parameters. The impact of HOI corrections is well documented in the literature. This paper investigates the temporal effects of HOI terms on estimated tropospheric parameters, i.e., zenith tropospheric wet delay (ZWD) and north and east gradients. For this purpose, observations from over 100 stations with good global coverage were used considering various geographic and geophysical conditions. The results of numerical experiments show that HOI effects have a significant impact on the estimated tropospheric parameters, and the influence is dependent on location and time. The maximum differences of ZWD estimates reach over 20 mm during periods of activity such as solar storms and geomagnetic storms. Additionally, the north gradients are more likely to be affected by HOI effects compared with east gradients. In particular, the tropospheric gradient component is most affected for low latitude station during daytime. Additionally, the effects of bending errors and HOI terms on slant tropospheric delay at low elevation angles are much larger than those at high elevation angles.