14 June 2022 in IEEE Transactions on Geoscience and Remote Sensing
Snow depth monitoring is meaningful for climate analysis, hydrological research, and snow disaster prevention. Global navigation satellite system-reflectometry (GNSS-R) technology uses the relationship between the modulation frequency of the signal-to-noise ratio (SNR) and reflector height to monitor snow depth. Existing research on single constellation has made good progress and is gradually developing toward multiconstellation combined inversion. Aiming at the accuracy of snow depth inversion, this article introduces the variational mode decomposition (VMD) algorithm with the characteristics of an adaptive high-pass filter to detrend the SNR data. The experimental results of KIRU station and P351 station show that the VMD algorithm is suitable for different constellations and has better signal separation effect. The snow depth inversion results for both stations are in high agreement with the in situ snow depths provided by the Swedish Meteorological and Hydrological Institute (SMHI) and the SNOTEL network. The root-mean-square error (RMSE) of the inversion results is reduced by 20%–40% compared to the least-squares fitting (LSF) algorithm, and the correlation coefficients are also greatly improved. Moreover, considering that there is no overlap between the climate station and the inversion area, this article introduces the maximum spectral amplitude as another reference data source and obtains basically consistent experimental conclusions. On this basis, the maximum spectral amplitude is used as the input variable of the entropy method, and the feasibility of the combination strategy is studied. The results show that the combined strategy reduces a little inversion error and improves the temporal resolution of snow depth monitoring. It is of great significance for more accurate and rapid monitoring of snow depth changes and disaster warnings and provides an important reference for further research on the GNSS-R technology.
22 July 2022 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
The signal-to-noise ratio (SNR) is important observations in global navigation satellite system-reflectometry (GNSS-R) technology. The oscillation frequency in the SNR arc is sensitive to different reflecting surfaces and can be used to build height model to track the variation of snow depth. However, it is difficult to obtain retrieval results with snow depth of zero in the actual snow depth retrieval experiments based on GNSS-R technology, which indicates that the classical model has nonnegligible retrieval errors in the snow-free state. This study aims to realize the detection of ground truth information before snow depth retrieval, i.e., classification of snow-free state and snow-covered state. Machine learning was introduced to achieve the aforementioned purpose and the SNR arc was used as the input data. Compared with the current mainstream topography correction algorithms, the algorithm proposed in this study does not rely on any priori ground measured data and has theoretical universality. The detection results can constrain the retrieval snow depth in the snow-free state and, thus, improve the retrieval accuracy. The experimental results for the 2014 seasonal snowpack at P351 station in Idaho, USA, show that the detection results obtained based on support vector machines agree well with the measured snow depth provided by the SNOTEL network, and the overall detection accuracy can reach about 96%. The daily snowpack state is determined by the majority of SNR arcs detected during the day and is only considered reliable if the percentage exceeds 75%. Only one day of the detection results was inaccurate and only 8 days (8/365) did not reach the set threshold of 75%. With the help of the detection results, the root-mean-square error of snow depth retrieval can be reduced from 20 cm in the classical algorithm to 15 cm, which results in a 25% improvement in retrieval accuracy. Moreover, this study broadens the application value of GNSS signals and provides a reference for the application of SNR in the detection field.
21 April 2023 in IEEE Transactions on Geoscience and Remote Sensing
Global Navigation Satellite System-Reflectometry (GNSS-R) is one of the main technologies used for sea ice remote sensing detection, and is based on the multipath interference effect of satellite signals. To improve the GNSS-R sea ice detection performance in terms of accuracy, robustness to noise, and data utilization, a linear discriminant analysis (LDA)-based method was proposed in this paper. Delay-Doppler maps (DDMs) collected from TechDemoSat-1 (TDS-1) were employed as input and classified into different types based on the signal-to noise ratio (SNR) related to the noise effect. For low-effect-noise DDMs, the LDA-based sea-ice detection method presented an accuracy of 95.03%, verifying the feasibility of LDA-based GNSS-R sea-ice detection. For the middle noise effect and high noise effect DDMs, the LDA-based method is more robust to noise effects than the convolutional neural network (CNN) method. Although the detection accuracy decreased when the SNR decreased or integral delay waveform average (IDWA) increased, the LDA-based method was more robust than the CNN-based one. The data utilization and melting period were also analyzed to account for variations in detection accuracy. The LDA-based method used 67.82% more data than previous experiments with threshold IDWA≤58210.32 and SNR>-17.48dB. The melting periods were analyzed based on the noise, SNR, surface reflectivity, and permittivity. When the status of sea ice changes, outliers of surface reflectivity appear, the average permittivity varies in [10, 60], and the detection accuracy decreases during the melting period of sea ice. The results show that the correlation coefficient with the National Oceanic and Atmospheric Administration (NOAA) data is up to 0.93, with different threshold IDWA or IDWA. The LDA-based method predicted results that greatly matched the sea ice distribution from the NOAA data.
8 April 2025 in IEEE Transactions on Geoscience and Remote Sensing
Sea ice thickness (SIT) retrieval using Global Navigation Satellite System-Reflectometry (GNSS-R) is a challenging problem in sea ice remote sensing, especially for SITs over 1 m, which is still in the blank stage. In this article, a seamless stacking-based retrieval method for SIT is proposed, which reduces the root mean square error (RMSE) for thicknesses below 1 m while ensuring the accuracy of SIT retrieval for thicknesses above 1 m. Principal component analysis (PCA) was used to extract delayed Doppler map (DDM) features, while the scattering coefficient and incidence angle were calculated using TechDemoSat-1 (TDS-1) data. Sea ice salinity and temperature were derived from Soil Moisture and Ocean Salinity (SMOS) data and used as inputs to the model along with other features. The performance of four machine learning (ML) algorithms—decision tree (DT), K-nearest neighbors (KNNs), support vector regression (SVR), and random forest (RF)—was compared, and the stacking model was constructed using these four algorithms as the base learner to improve performance. Validation using SMOS data for thicknesses up to 1 m showed that the stacking algorithm significantly improved retrieval accuracy, reducing the RMSE from 7 to 0.4 cm and improving the correlation coefficient (r) from 0.94 to 0.99. For thicknesses greater than 1 m, validation using Cryosat-2 data also showed strong performance. In addition, the effect of sea ice parameters on retrieval accuracy and sources of error was analyzed.
2 October 2024 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
The Global Navigation Satellite System Reflectometry (GNSS-R) technology has shown significant potential in retrieving snow depth using Signal-to-Noise Ratio (SNR) data. However, compared to traditional in-situ snow depth measurement techniques, we have observed that the accuracy and performance of GNSS-R can be significantly impacted under certain conditions, particularly when the elevation angle increases. This is due to the attenuation of the multipath effect, which is particularly evident during snow-free periods and under low snow conditions where snow depths are below 50 cm. To address these limitations, we propose a snow depth inversion method that integrates SNR signals with the Support Vector Regression (SVR) algorithm, utilizing SNR sequences as feature inputs. We conducted studies at stations P351 and P030, covering elevation angles ranging from 5° to 20°, 5° to 25°, and 5° to 30°. The experimental results show that the Root Mean Square Error (RMSE) at both stations decreased by 50% or more compared to traditional methods, demonstrating an improvement in inversion accuracy across different elevation angles. More importantly, the inversion accuracy of our method does not significantly lag behind that at lower elevation angles, indicating its excellent performance under challenging conditions. These findings highlight the contribution of our method in enhancing the accuracy of snow depth retrieval and its potential to drive further advancements in the field of GNSS-R snow depth inversion.
3 July 2023 in IEEE Geoscience and Remote Sensing Letters
The Global Navigation Satellite System interferometric reflectometry (GNSS-IR) technique based on signal-to-noise ratio (SNR) data is widely used for snow depth retrieval. Since snow depth retrieval in a snow-free state is very important for meteorological monitoring and since many corrections are post-processed to improve the retrieval accuracy, we propose a GNSS-IR snow depth retrieval model based on a back-propagation neural network optimized by a genetic algorithm to detect the snow state and predict snow depth using the frequency, amplitude and phase of the multipath oscillation term as input features. GPS data collected from the P351 station of the PBO network and measured snow depth from the SNOTEL network were used to conduct the experiments. The accuracy of daily snow state detection for the experimental station exceeded 96%. Combined with the snow state detection results for snow depth regression prediction, the experimental results show that the root mean square error of the snow depth retrieval results for P351 station is 12.09 cm. Compared with the traditional model, the retrieval accuracy is improved by 29.1%, and the correlation coefficient also reaches 0.97, indicating that the proposed snow depth retrieval model not only has high accuracy but also has strong stability. In this study, snow state detection is proposed to improve the retrieval accuracy in snow-free conditions, and the possibility of snow depth retrieval without antenna height is provided.