Shore-star sea surface wind wave observation data fusion method and system

Abstract

The embodiment of the invention provides a shore-star sea surface wind wave observation data fusion method and system. The method is applied to the technical field of data processing, and comprises the following steps: receiving a signal by adopting a Beidou satellite receiver, performing feature extraction according to Doppler frequency shift and time delay information, and establishing a wind wave feature data set; constructing a land reef distribution map based on geographic information, and performing interference identification and filtering on the data set; constructing a sea condition reference library through ERA5 and NCEP reanalysis data, and establishing and correcting a change mode; carrying out weight fusion on the effective signals, shore-based observation and reference data; carrying out wind wave field reconstruction on the comprehensive data to generate high-resolution data; and verifying and optimizing by utilizing independent observation to obtain an observation result. According to the method, effective fusion and refined reconstruction of the multi-source heterogeneous data are realized.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing, and particularly to a shore-satellite sea surface wind and wave observation data fusion method and system. Background Art

[0002] At present, there are mainly two types of methods for sea surface wind and wave observation: one is based on traditional shore-based observation equipment, such as wave radars, ocean weather stations, etc., which can obtain wind and wave data in the nearshore sea area in real time; the other is based on satellite remote sensing technology, such as scatterometers, altimeters, etc., which can obtain wind field and wave field information over a large area. At the same time, in order to obtain more sea condition information, numerical forecasts and reanalysis data are also widely used as supplementary data sources. These observation methods and data sources have their own characteristics and play important roles at different spatio-temporal scales.

[0003] However, the existing observation methods have some limitations: shore-based observation equipment is restricted by terrain and observation range and it is difficult to obtain data in the open sea area; satellite remote sensing is restricted by orbits and revisit periods and has limited spatio-temporal resolution; although numerical forecasts and reanalysis data have a wide coverage, their real-time performance and accuracy need to be further improved. Especially in the complex nearshore sea area, due to the influence of terrains such as land and reefs, it is difficult for a single observation method to accurately capture the change characteristics of sea surface wind and waves. In addition, the data obtained by different observation methods have problems such as uneven spatio-temporal distribution and inconsistent accuracy. How to effectively fuse multi-source observation data has become an urgent problem to be solved. Summary of the Invention

[0004] The present disclosure provides a shore-satellite sea surface wind and wave observation data fusion method and system, and the present disclosure realizes the effective fusion and refined reconstruction of multi-source heterogeneous data.

[0005] According to a first aspect of the present disclosure, there is provided a shore-satellite sea surface wind and wave observation data fusion method, including: receiving direct signals and sea surface reflection signals by using a Beidou satellite receiver, extracting features of the direct signals and the sea surface reflection signals according to Doppler frequency shift and time delay information, and establishing a sea surface wind and wave feature data set; constructing a land and reef distribution map of a detection area based on a geographic information system, performing interference identification and filtering on the sea surface wind and wave feature data set to obtain sea surface effective signal data; constructing a historical sea condition reference library through ERA5 and NCEP reanalysis data, establishing a sea condition change pattern according to the historical sea condition reference library, performing spatio-temporal correction processing on the sea condition change pattern to obtain sea surface state reference data; performing fusion processing on the sea surface effective signal data, shore-based observation data, and the sea surface state reference data according to weight allocation to obtain comprehensive observation data; performing wind and wave field reconstruction processing on the comprehensive observation data, and adjusting parameters through ocean dynamics constraints to generate high-resolution wind and wave field distribution data; using independent observation data to perform cross-validation and parameter optimization on the high-resolution wind and wave field distribution data to obtain sea surface wind and wave observation results.

[0006] According to a second aspect of the present disclosure, there is provided a shore-satellite sea surface wind and wave observation data fusion system, including:

[0007] An extraction module, configured to receive direct signals and sea surface reflection signals by using a Beidou satellite receiver, extract features of the direct signals and the sea surface reflection signals according to Doppler frequency shift and time delay information, and establish a sea surface wind and wave feature data set;

[0008] An identification module, configured to construct a land and reef distribution map of a detection area based on a geographic information system, perform interference identification and filtering on the sea surface wind and wave feature data set to obtain sea surface effective signal data;

[0009] A correction module, configured to construct a historical sea condition reference library through ERA5 and NCEP reanalysis data, establish a sea condition change pattern according to the historical sea condition reference library, perform spatio-temporal correction processing on the sea condition change pattern to obtain sea surface state reference data;

[0010] A fusion module, configured to perform fusion processing on the sea surface effective signal data, shore-based observation data, and the sea surface state reference data according to weight allocation to obtain comprehensive observation data;

[0011] A reconstruction module, configured to perform wind and wave field reconstruction processing on the comprehensive observation data, and adjust parameters through ocean dynamics constraints to generate high-resolution wind and wave field distribution data;

[0012] A verification module, configured to use independent observation data to perform cross-validation and parameter optimization on the high-resolution wind and wave field distribution data to obtain sea surface wind and wave observation results.

[0013] The present disclosure receives direct signals from Beidou satellites and sea surface reflection signals, extracts features based on Doppler frequency shift and time delay information, and establishes a dataset of sea surface wind and wave features, providing basic data support for subsequent processing; constructs a distribution map of land and reefs using a geographic information system, identifies and filters interference in the sea surface wind and wave feature dataset, effectively eliminates the influence of terrain factors, and improves the reliability of signal data; constructs a historical sea condition reference library through ERA5 and NCEP reanalysis data and performs spatio-temporal correction processing, providing reliable background field information for wind and wave field analysis; uses a weight allocation method to fuse multi-source data, makes full use of the advantages of different data sources, and improves the quality of comprehensive observation data; reconstructs the wind and wave field through ocean dynamics constraints, achieving a refined description of a high-resolution wind and wave field; uses independent observation data for cross-validation and parameter optimization, and establishes a quality control mechanism to ensure the accuracy and reliability of observation results. The entire technical solution solves problems such as uneven data and low accuracy in sea surface wind and wave observations through the fusion processing of multi-source data and dynamic constraints, and can still obtain accurate observation results under complex sea conditions.

[0014] It should be understood that the content described in the section of the invention content is not intended to limit the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In combination with the accompanying drawings and with reference to the following detailed description, the above and other features, advantages, and aspects of the embodiments of the present disclosure will become more apparent. The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. In the drawings, the same or similar reference numerals represent the same or similar elements, where:

[0016] Figure 1 shows a flowchart of a shore-satellite sea surface wind and wave observation data fusion method according to an embodiment of the present disclosure;

[0017] Figure 2 shows a block diagram of a shore-satellite sea surface wind and wave observation data fusion system according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are some, but not all, of the embodiments of the present disclosure. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present disclosure without creative efforts fall within the scope of protection of the present disclosure.

[0019] In addition, the term "and / or" in this text is merely a relational description of associated objects, indicating three possible relationships. For example, A and / or B can represent three cases: A exists alone, both A and B exist simultaneously, and B exists alone. In addition, the character " / " in this text generally indicates an "or" relationship between the associated objects before and after.

[0020] Figure 1 Fig. shows a schematic flowchart of a shore-satellite sea surface wind and wave observation data fusion method 100 in an embodiment of the present disclosure, as Figure 1 shown, the method 100 includes:

[0021] S110: Use a Beidou satellite receiver to receive direct signals and sea surface reflection signals, extract features of the direct signals and sea surface reflection signals based on Doppler frequency shift and delay information, and establish a sea surface wind and wave feature data set;

[0022] Optionally, obtain the phase difference between the direct signal and the sea surface reflection signal through a phase difference calculator, and calculate the sea surface roughness parameter according to the phase difference; perform power spectral density analysis on the direct signal and the sea surface reflection signal by a power spectrum analyzer to obtain signal power spectrum eigenvalues; perform noise reduction processing on the direct signal and the sea surface reflection signal based on wavelet transform to obtain denoised signal data; establish a signal quality evaluation index according to the sea surface roughness parameter, signal power spectrum eigenvalues, and denoised signal data; screen high-quality signals according to the signal quality evaluation index, and record the incident angle and azimuth angle parameters of the high-quality signals; integrate the characteristic parameters, incident angle, and azimuth angle parameters of the high-quality signals to generate a sea surface wind and wave feature data set.

[0023] Among them, in the detection of sea surface wind waves by Beidou satellites, the main content of signal processing includes the feature extraction of direct signals and sea surface reflection signals. The phase difference between the two signals is obtained through a phase difference calculator, and the phase difference directly reflects the roughness of the sea surface. The phase difference calculation adopts the phase difference calculation method of complex signals, and the phase difference is obtained by calculating the difference between the complex phase angles of the two signals. There is a corresponding relationship between the phase difference value and the sea surface roughness parameter. The rougher the sea surface, the greater the fluctuation of the phase difference value. By establishing the mapping relationship between the phase difference value and the sea surface roughness, the sea surface roughness parameter is obtained. Power spectrum analysis is an important means of extracting signal features. The power spectrum analyzer processes the direct signal and the sea surface reflection signal to obtain the energy distribution characteristics of the signal in the frequency domain. The power spectrum density analysis adopts the fast Fourier transform method to convert the time-domain signal to the frequency domain and calculate the power spectrum density function of the signal. The signal power spectrum characteristic values include information such as the main frequency component, bandwidth, and sideband characteristics of the signal, and these characteristic values reflect the modulation effect of the sea surface state on the signal. For the noise interference existing in the signal, wavelet transform is used for noise reduction processing. Wavelet transform has the characteristics of multi-resolution analysis and can separate signals and noise at different scales. In the noise reduction process, the signal is first wavelet decomposed to obtain wavelet coefficients at different scales, and then the wavelet coefficients are processed according to the threshold criterion to remove the noise components. Finally, the denoised signal data is obtained through wavelet reconstruction. The establishment of the signal quality evaluation index comprehensively considers the sea surface roughness parameter, the signal power spectrum characteristic value, and the denoised signal data. The evaluation index includes parameters in multiple dimensions such as signal-to-noise ratio, signal integrity, and phase stability. By setting the evaluation threshold, high-quality signals are selected, and the incident angle and azimuth angle parameters of these signals are recorded. The incident angle reflects the angle between the signal and the sea surface, and the azimuth angle represents the propagation direction of the signal. The characteristic parameters and geometric parameters of the high-quality signals are integrated to generate a sea surface wind wave characteristic data set. The data set contains multi-dimensional information such as the phase characteristics, power spectrum characteristics, and spatial geometric characteristics of the signal, providing a data basis for subsequent sea surface state analysis. For example, when a group of Beidou satellite signals is received, first calculate the phase difference between the direct signal and the reflected signal to obtain the phase difference sequence. Perform power spectrum analysis on these sequences to extract the frequency characteristics of the signal. Perform multi-scale decomposition on the signal through wavelet transform to remove the high-frequency noise components. According to the processed signal characteristics, calculate the signal quality evaluation index, and select the signal data with a high evaluation score. Finally, combine the spatial information such as the incident angle and azimuth angle of the signal to form a sea surface wind wave characteristic data set.

[0024] For example, when the Beidou satellite signal irradiates the sea surface, the reflected signal will be modulated by the sea surface wind and waves. Through phase difference calculation, it is found that the phase difference between the direct signal and the reflected signal shows periodic changes, and this change is related to the period of the sea surface waves. By performing power spectrum analysis on the signal, obvious Doppler frequency shifts can be observed in the frequency spectrum, and the magnitude of the frequency shift reflects the speed of the sea surface movement. After wavelet transform noise reduction processing, the time-frequency characteristics of the signal become clearer, and the characteristic information of the sea surface wind and waves is effectively extracted. By comprehensively evaluating the signal quality, signals that meet the requirements are selected for subsequent processing. The recording of the incident angle and azimuth angle provides a basis for determining the spatial distribution of the wind and wave field. This multi-dimensional signal feature extraction method makes full use of the phase, frequency, and spatial characteristics of the Beidou satellite signal. When there are obvious periodic changes found through phase difference calculation in a set of Beidou satellite signal data, this change has a good correspondence with the actually observed sea surface wave period. The power spectrum analysis results show that significant energy aggregation appears in a specific frequency range, and these energy distribution characteristics are directly related to the intensity of the sea surface wind and waves. After wavelet transform processing, the random noise in the signal is effectively suppressed, and the characteristics of the sea surface wind and waves become more prominent. According to the signal quality evaluation index, signal data with higher quality are screened out, and these data show good consistency in both the frequency domain and the time domain. The generated sea surface wind and wave characteristic data set not only contains various characteristic parameters of the signal but also records the spatial distribution information of the signal.

[0025] S120: Based on the geographic information system, construct a distribution map of land and reefs in the detection area, identify and filter the interference in the sea surface wind and wave characteristic data set to obtain sea surface effective signal data;

[0026] Optionally, extract the terrain data of the detection area from the geographic information system, calibrate the accuracy of the terrain data to generate electronic map data; extract the land boundary and reef positions of the detection area based on the electronic map data to construct a distribution map of land and reefs; perform spatial matching between the signal reflection points in the sea surface wind and wave characteristic data set and the distribution map of land and reefs to identify the interfered signals; perform hierarchical processing on the interfered signals according to the tide data to establish an interference characteristic database; perform multipath effect analysis on the sea surface wind and wave characteristic data set through three-dimensional scattering calculation and record the multipath interference data; remove the interference characteristic database and the multipath interference data from the sea surface wind and wave characteristic data set to obtain sea surface effective signal data.

[0027] Among them, obtain the original terrain data of the detection area from the geographic information system, and these data include information such as land elevation, coastline contour, and underwater terrain. Perform accuracy calibration on the original terrain data, and the calibration process mainly considers factors such as data acquisition error and coordinate system difference, and generates high-precision electronic map data through control point matching and coordinate transformation.

[0028] Feature extraction is carried out based on electronic map data, with a focus on identifying the land boundary line and the distribution location of reefs in the detection area. For the extraction of the land boundary line, edge detection and contour tracing methods are used. By analyzing the gradient change of elevation data, the demarcation line between land and sea is determined. The identification of reef positions is based on underwater terrain features. By setting elevation thresholds and performing morphological analysis, the reef areas exposed above or close to the water surface are marked. The extracted land boundary and reef position information are integrated to construct a land and reef distribution map. The key to signal interference identification lies in spatially matching the signal reflection points in the sea surface wind and wave feature dataset with the land and reef distribution map. The position of the reflection point is calculated through the geometric parameters of satellite signals, including information such as the signal incident point coordinates and reflection angle. The spatial matching process uses distance calculation and region judgment methods to determine whether each reflection point falls within the influence range of the land boundary or reef area, thereby identifying the signals interfered by land and reefs.

[0029] To further refine the interference analysis, tidal data is introduced to classify the interfered signals. Tidal data reflects the periodic change of the sea level, which directly affects the exposure degree of reefs and the actual position of the land boundary. According to the tidal height, the interfered signals are divided into different levels, such as complete interference, partial interference, etc. At the same time, characteristic information such as the time and intensity of interference occurrence is recorded to establish a system interference characteristic database. Multipath effect is another important source of interference. The signal propagation path is analyzed through three-dimensional scattering calculation. The scattering calculation takes into account factors such as terrain undulation and reflection surface characteristics, and simulates the signal propagation process in a complex environment. The calculation process includes the interaction between the incident wave and the reflection surface, the generation and propagation of scattered waves, etc. Finally, multipath interference data is obtained, which records parameters such as the delay and intensity of each reflection path. The interfered signals recorded in the interference characteristic database and the multipath interference data are removed from the original sea surface wind and wave feature dataset. The filtering process first marks all the data items identified as interference, and then decides whether to retain them according to the interference degree. For slightly interfered signals, they are retained after compensation processing; for severely interfered signals, they are directly removed. After this series of processing, the obtained sea surface effective signal data has high reliability.

[0030] For example, when conducting sea surface wind and wave observations, terrain data of the observation area is obtained, and through precision calibration processing, an electronic map with a resolution of 5 meters is generated. The land boundary and reef positions are extracted from the electronic map, and it is found that there are multiple reef groups in the observation area. When receiving Beidou satellite signals, the position information of the signal reflection points is calculated, and these positions are matched with the land and reef distribution maps. Combining the tidal data at that time, it is determined that the influence degree of some reef areas on the signal will change with the tide. Through three-dimensional scattering calculation, the multipath propagation paths caused by terrain undulation are identified. Finally, considering all interference factors comprehensively, the effective signal data that truly reflects the sea surface state is selected. This process demonstrates how to effectively remove the influence of environmental factors such as land and reefs on sea surface wind and wave observations through multi-dimensional data analysis and processing, and obtain accurate sea surface state information.

[0031] S130: Construct a historical sea condition reference library through ERA5 and NCEP reanalysis data, establish a sea condition change model based on the historical sea condition reference library, and perform spatio-temporal correction processing on the sea condition change model to obtain sea surface state reference data;

[0032] Optionally, perform data format unification processing on ERA5 and NCEP reanalysis data, extract sea condition correlation parameters, and generate standardized sea condition data; perform time series analysis on the standardized sea condition data, separate seasonal and diurnal variation characteristics, and form a sea condition time-varying characteristic sequence; perform Bayesian correction on the outliers in the sea condition time-varying characteristic sequence to obtain sea condition anomaly correction data; integrate and store the standardized sea condition data and the sea condition anomaly correction data to construct a historical sea condition reference library; extract the sea condition change law from the historical sea condition reference library to establish a sea condition change model; perform correction processing on the sea condition change model through spatio-temporal correlation analysis to obtain sea surface state reference data.

[0033] Among them, ERA5 and NCEP reanalysis data contain historical data of sea condition parameters such as sea surface wind field, wave height, and period. Unifying the data formats of these two data sources involves aligning spatio-temporal resolutions, unifying units, and coordinate system conversion. Key sea condition parameters are extracted from the unified data, including sea surface wind speed, wind direction, significant wave height, wave direction, etc., to generate standardized sea condition data. The standardization process uses the maximum-minimum normalization method to make the data of different physical quantities comparable. Perform time series analysis on the standardized sea condition data, and use the spectral analysis method to separate seasonal and diurnal variation characteristics. The seasonal variation characteristics are extracted by Fourier transform to obtain annual and semi-annual cycle components, and the diurnal variation characteristics are identified by wavelet analysis to identify 24-hour and its harmonic components. These time-varying characteristics form a sea condition time-varying characteristic sequence, reflecting the periodic change law of sea condition parameters.

[0034] Perform Bayesian correction on the outliers in the sea condition time-varying characteristic sequence, and the correction formula is:

[0035]

[0036] Among them: P(e|D) represents the posterior probability of parameter θ given the observed data D; P(D|θ) represents the likelihood function; P(θ) represents the prior distribution of parameter θ; P(D) represents the marginal likelihood.

[0037] The specific outlier correction model is:

[0038]

[0039] Among them: H(ω, t) represents the time-frequency distribution of sea state parameters; W i (ω) represents the weight function of the i-th frequency component; φ i (t) represents the time modulation function; α i represents the amplitude coefficient of the frequency component; β represents the residual term coefficient; R(ω, t) represents the random perturbation term; N represents the number of frequency components considered.

[0040] The sea state anomaly correction data obtained by Bayesian correction and the standardized sea state data are integrated and stored to construct a historical sea state reference library. The sea state change rules, including the evolution characteristics of the wind field and the wave propagation characteristics, are extracted from the reference library to establish a sea state change model. The spatio-temporal correlation analysis is carried out on the sea state change model, considering the spatial proximity and time continuity, to obtain the reference data of the sea surface state.

[0041] For example: The observation data of a certain sea area are processed. First, the 6-hour interval data of ERA5 and the 3-hour interval data of NCEP are unified to a 3-hour time resolution, and the spatial resolution is unified to 0.25 degrees. The parameters such as wind speed, wind direction, and significant wave height of this sea area are extracted and standardized. Through time series analysis, it is found that there are obvious seasonal variations in this sea area, with stronger wind waves in winter and weaker in summer, and significant diurnal variation characteristics at the same time. The Bayesian correction method is used to process the outliers, and the historical statistical characteristics and physical constraint conditions are considered in the correction process. The corrected data are integrated to establish a historical sea state reference library, and the evolution rules of the wind wave field are extracted from it to establish a sea state change model reflecting the characteristics of this sea area. Finally, through spatio-temporal correlation analysis, the model is revised to obtain the reference data reflecting the actual sea state.

[0042] S140: The sea surface effective signal data, shore-based observation data, and sea surface state reference data are fused according to the weight distribution to obtain the comprehensive observation data;

[0043] Optionally, calculate the signal data weight coefficient according to the signal-to-noise ratio and geometric factor of the sea surface effective signal data; extract the equipment performance index and environmental impact factor from the shore-based observation data to generate the shore-based data weight coefficient; based on the consistency analysis of the sea surface state reference data and the measured value, obtain the reference data weight coefficient; perform data interpolation on the sea surface effective signal data, shore-based observation data and sea surface state reference data through Kalman filtering to generate uniformly distributed data;

[0044] Perform data normalization based on the signal data weight coefficient, shore-based data weight coefficient and reference data weight coefficient to form weight fusion data; combine the weight fusion data with the uniformly distributed data to obtain the comprehensive observation data.

[0045] Among them, when calculating the weight of the sea surface effective signal data, mainly consider two parameters: the signal-to-noise ratio and the geometric factor. The signal-to-noise ratio reflects the quality of the signal and is calculated by the ratio of the signal power to the noise power. The geometric factor considers the spatial geometric relationships such as the incident angle and azimuth angle of the satellite signal. These two parameters jointly determine the reliability of the signal data, and then determine the signal data weight coefficient. The weight calculation of the shore-based observation data needs to consider the equipment performance index and environmental impact factor. The equipment performance index includes parameters such as instrument accuracy, stability, and sampling frequency, and the environmental impact factor includes external factors such as meteorological conditions and measurement location. Through the comprehensive evaluation of these parameters, generate the shore-based data weight coefficient reflecting the data quality.

[0046] The weight determination of the sea surface state reference data is based on the consistency analysis with the measured value. By calculating statistical indicators such as the correlation and mean square error between the reference data and the historical measured data, evaluate the accuracy of the reference data, so as to obtain the reference data weight coefficient. During the consistency analysis process, the matching relationship of the spatio-temporal scale also needs to be considered to ensure the rationality of the weight coefficient. For the problem of uneven spatio-temporal distribution existing in these three types of data, Kalman filtering is used for data interpolation. Kalman filtering is a recursive estimation algorithm that estimates the missing data through two steps: prediction and update. The prediction step predicts the state at the next moment based on the known data, and the update step corrects the prediction result according to the new observation data. In this way, a data set with uniform spatio-temporal distribution is generated.

[0047] After the weight coefficients of various types of data are determined, normalization processing is required to make the weights from different sources comparable. The normalization process takes into account the characteristics and reliability of various types of data to form unified weight fusion data. Finally, combine the weight fusion data with the uniformly distributed data to generate the final comprehensive observation data.

[0048] For example, in the observation of a certain sea area, the signal-to-noise ratio of Beidou satellite signals is obtained through spectral analysis. At the same time, geometric parameters such as the incident angle and azimuth angle of the signals are calculated, and the weight coefficients of the signal data are determined after comprehensive evaluation. The radar equipment of the shore-based observation station has fixed measurement accuracy and sampling interval, and the weight coefficient of the shore-based data is calculated in combination with the local meteorological conditions. The ERA5 and NCEP reanalysis data determine their weight coefficients by comparing with historical Bui data. These three types of data have uneven distributions in time and space, and are interpolated through Kalman filtering to obtain a gridded data set. Finally, data fusion is performed according to their respective weight coefficients to generate a comprehensive observation result reflecting the true sea conditions. This process demonstrates how to effectively fuse observation data from different sources and with different characteristics to obtain more accurate sea surface wind and wave observation information.

[0049] S150: Perform wind and wave field reconstruction processing on the comprehensive observation data, adjust parameters through ocean dynamics constraints, and generate high-resolution wind and wave field distribution data;

[0050] Optionally, conduct sea surface dynamic characteristic analysis on the comprehensive observation data, extract wind-wave coupling parameters, and obtain wind-wave initial field data; perform dynamic constraints on the wind-wave initial field data based on the energy balance equation to generate wind-wave constraint conditions; extract local characteristic structures from the comprehensive observation data to establish wind-wave field boundary conditions; couple and calculate the wind-wave constraint conditions and the wind-wave field boundary conditions to obtain wind-wave field iterative data; perform data interpolation processing on the wind-wave field iterative data to establish refined grid data; perform spatial distribution reconstruction according to the refined grid data to generate high-resolution wind and wave field distribution data.

[0051] Among them, the key to the analysis process is to extract wind-wave coupling parameters, including basic parameters such as wind speed, wind direction, wave height, and wave direction, as well as interaction parameters between wind and waves. Through the analysis of these parameters, initial field data reflecting the sea surface state is established, and these data serve as the basis for subsequent dynamic constraints. The energy balance equation is the core tool for performing dynamic constraints. This equation describes the processes of sea surface wave energy propagation, generation, dissipation, etc., and based on this equation, constraint processing is performed on the wind-wave initial field data. The constraint process takes into account multiple physical processes such as wind field input, wave interaction, and bottom friction dissipation to generate wind-wave constraint conditions that conform to physical laws.

[0052] The extraction of local feature structures is the key to establishing boundary conditions. Local features such as coastline morphology, water depth changes, and reef distributions are extracted from comprehensive observation data, and these features directly affect the propagation and deformation of waves. Feature extraction uses spatial filtering and morphological analysis methods to identify topographic features that have important impacts on the evolution of the wind-wave field and establish the boundary conditions of the wind-wave field. The coupled calculation of wind-wave constraint conditions and boundary conditions is an iterative process. During the calculation process, the constraint conditions are first applied to the entire calculation domain, and then corrections are made at the boundaries to ensure the continuity and physical rationality of the solution. The iterative calculation continues until the convergence conditions are met, and iterative data of the wind-wave field is obtained. The iterative data reflects the distribution characteristics of the wind-wave field under the dual effects of constraint conditions and boundary conditions.

[0053] Interpolating the iterative data of the wind-wave field is an important step to improve the resolution. The interpolation process uses two-dimensional spline interpolation method to increase the grid point density while maintaining the original data characteristics. Continuity and smoothness of physical quantities need to be considered during interpolation to avoid introducing false spatial structures during the refinement process. High-density refined grid data is established through interpolation. The reconstruction process is based on the refined grid data and uses spatial analysis methods to reconstruct the detailed distribution of the wind-wave field. Physical constraints such as terrain effects and energy conservation need to be considered during reconstruction to ensure that the generated high-resolution wind-wave field distribution data not only meets the observation constraints but also conforms to physical laws.

[0054] For example, when processing the comprehensive observation data of a certain sea area, it is first found through wind-wave feature analysis that there are obvious interactions between the wind field and the wave field in this area. The energy balance equation constraint shows that the wave energy mainly comes from the local wind field input, and part of the energy is lost through propagation and dissipation processes. Local feature analysis finds that there are many reefs and islands in this sea area, and these topographic features have a significant impact on wave propagation. During the iterative calculation process, the coupling of constraint conditions and boundary conditions causes local deformation of the wave field around the reefs. The interpolation process encrypts the original grid by 5 times to generate refined grid data. The final spatial distribution reconstruction clearly shows the spatial variation characteristics of the wind-wave field, especially the detailed features near complex terrains.

[0055] S160: Use independent observation data to cross-validate and optimize the parameters of the high-resolution wind-wave field distribution data to obtain the sea surface wind-wave observation results.

[0056] Optionally, divide the independent observation data by time period, verify the high-resolution wind and wave field distribution data in batches, and generate verification index data; conduct spatial consistency analysis based on the verification index data to construct an error distribution map; perform sensitivity analysis on the system parameters through the error distribution map to obtain parameter optimization indicators; adjust the signal processing threshold according to the parameter optimization indicators to obtain threshold optimization data; compare and process the threshold optimization data with the high-resolution wind and wave field distribution data to form parameter correction data; update the high-resolution wind and wave field distribution data according to the parameter correction data to obtain the sea surface wind and wave observation results.

[0057] Among them, classify and screen the independent observation data. Divide the independent observation data into multiple batches according to the time period, and each batch of data needs to be compared with the high-resolution wind and wave field distribution data of the corresponding time period to generate verification index data. The verification index data mainly includes the comparison results of elements such as wind speed, wind direction, wave height, and wave direction, and quantifies the verification results by calculating statistical parameters such as root mean square error and correlation coefficient. Spatial consistency analysis is an important means to evaluate the accuracy of high-resolution wind and wave field distribution data. Based on the verification index data, analyze the error distribution characteristics at different spatial positions, and focus on the spatial variation law of the errors. This analysis process needs to consider the influence of factors such as terrain and sea conditions to construct an error distribution map reflecting the spatial distribution of errors. The error distribution map intuitively shows the accuracy differences in the spatial reconstruction results of the wind and wave field.

[0058] Conduct sensitivity analysis on the error distribution map to study the relationship between system parameters and errors. During the sensitivity analysis process, it is necessary to adjust each key parameter and observe the influence degree of parameter changes on the error distribution. Through this analysis, parameter optimization indicators are obtained, and these indicators reflect the influence weights of different parameters on the result accuracy. Adjust the threshold of each link of signal processing according to the parameter optimization indicators. Threshold adjustment involves multiple processing steps such as signal filtering, feature extraction, and data fusion. The threshold adjustment of each step needs to consider its influence on the overall result. The adjusted thresholds constitute the threshold optimization data, and these data record the optimal parameter settings of each processing link.

[0059] Apply the threshold optimization data to the processing process of the high-resolution wind and wave field distribution data, and verify the effect of parameter adjustment through comparative analysis. During the comparison process, it is necessary to focus on the changes in the results before and after adjustment, especially the improvement in areas with large errors. The results of the comparative analysis form parameter correction data, and these data reflect the specific effects of parameter optimization. Update the high-resolution wind and wave field distribution data according to the parameter correction data. This process is actually a reprocessing process, using the optimized parameters to reprocess the data, and finally obtaining more accurate sea surface wind and wave observation results.

[0060] For example, in the observations of a certain offshore area, the independent observation data are divided into multiple verification batches at 6-hour intervals. Each batch contains the wind and wave element data of multiple observation stations, and these data are compared with the high-resolution wind and wave field distribution data in the same period. Through comparative analysis, it is found that the error in the area with dense reefs is relatively large, and this spatial distribution characteristic is clearly shown through the error distribution map. Further sensitivity analysis shows that the noise threshold and data fusion weight in signal processing have a significant impact on the results in these areas. Based on this discovery, the processing parameters in these areas are mainly adjusted, including improving the signal quality control standard, adjusting the data fusion weight, etc. The adjusted parameters are applied to the data processing process, significantly improving the observation accuracy in complex terrain areas. The entire optimization process demonstrates how to improve the accuracy of wind and wave field observations through systematic verification and parameter adjustment.

[0061] Figure 2 FIG. shows a block diagram of a shore-satellite sea surface wind and wave observation data fusion system 200 according to an embodiment of the present disclosure. As Figure 2 shown, the system 200 includes:

[0062] An extraction module 210, configured to receive a direct signal and a sea surface reflection signal by using a Beidou satellite receiver, extract features of the direct signal and the sea surface reflection signal according to Doppler frequency shift and time delay information, and establish a sea surface wind and wave feature data set;

[0063] An identification module 220, configured to construct a land and reef distribution map of a detection area based on a geographic information system, perform interference identification and filtering on the sea surface wind and wave feature data set, and obtain sea surface effective signal data;

[0064] A correction module 230, configured to construct a historical sea condition reference library through ERA5 and NCEP reanalysis data, establish a sea condition change pattern according to the historical sea condition reference library, perform spatio-temporal correction processing on the sea condition change pattern, and obtain sea surface state reference data;

[0065] A fusion module 240, configured to perform fusion processing on the sea surface effective signal data, shore-based observation data, and the sea surface state reference data according to weight distribution to obtain comprehensive observation data;

[0066] A reconstruction module 250, configured to perform wind and wave field reconstruction processing on the comprehensive observation data, adjust parameters through ocean dynamics constraints, and generate high-resolution wind and wave field distribution data;

[0067] A verification module 260, configured to perform cross-verification and parameter optimization on the high-resolution wind and wave field distribution data by using independent observation data to obtain sea surface wind and wave observation results.

[0068] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the described modules can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0069] The above specific implementation manners do not constitute a limitation on the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present disclosure shall be included within the protection scope of the present disclosure.

Claims

1. A shore-star sea surface wind and wave observation data fusion method, characterized in that Including: Receiving direct signals and sea surface reflection signals by using a Beidou satellite receiver, extracting features of the direct signals and the sea surface reflection signals according to Doppler frequency shift and time delay information, and establishing a sea surface wind and wave feature dataset; Constructing a land and reef distribution map of the detection area based on a geographic information system, identifying and filtering interferences in the sea surface wind and wave feature dataset to obtain sea surface effective signal data; Constructing a historical sea condition reference library through ERA5 and NCEP reanalysis data, establishing a sea condition change model according to the historical sea condition reference library, and performing spatio-temporal correction processing on the sea condition change model to obtain sea surface state reference data; Fusing the sea surface effective signal data, shore-based observation data, and the sea surface state reference data according to weight allocation to obtain comprehensive observation data; Performing wind and wave field reconstruction processing on the comprehensive observation data, adjusting parameters through ocean dynamics constraints, and generating high-resolution wind and wave field distribution data; Using independent observation data to perform cross-validation and parameter optimization on the high-resolution wind and wave field distribution data to obtain sea surface wind and wave observation results.

2. The shore-star sea surface wind wave observation data fusion method according to claim 1, wherein The step of receiving direct signals and sea surface reflection signals by using a Beidou satellite receiver, extracting features of the direct signals and the sea surface reflection signals according to Doppler frequency shift and time delay information, and establishing a sea surface wind and wave feature dataset includes: Obtaining the phase difference between the direct signal and the sea surface reflection signal through a phase difference calculator, and calculating the sea surface roughness parameter according to the phase difference; Performing power spectral density analysis on the direct signal and the sea surface reflection signal by a power spectrum analyzer to obtain signal power spectrum eigenvalue; Performing noise reduction processing on the direct signal and the sea surface reflection signal based on wavelet transform to obtain noise-reduced signal data; Establishing a signal quality evaluation index according to the sea surface roughness parameter, the signal power spectrum eigenvalue, and the noise-reduced signal data; Screening high-quality signals according to the signal quality evaluation index, and recording the incident angle and azimuth angle parameters of the high-quality signals; Integrating the characteristic parameters, the incident angle, and the azimuth angle parameters of the high-quality signals to generate the sea surface wind and wave feature dataset.

3. The shore-star sea surface wind and wave observation data fusion method according to claim 1, characterized in that The step of constructing a land and reef distribution map of the detection area based on a geographic information system, identifying and filtering interferences in the sea surface wind and wave feature dataset to obtain sea surface effective signal data includes: Extracting terrain data of the detection area from the geographic information system, calibrating the accuracy of the terrain data, and generating electronic map data; Extracting the land boundary and reef positions of the detection area based on the electronic map data to construct the land and reef distribution map; Performing spatial matching between the signal reflection points in the sea surface wind and wave feature dataset and the land and reef distribution map to identify interfered signals; Performing grading processing on the interfered signals according to tide data to establish an interference feature database; Performing multipath effect analysis on the sea surface wind and wave feature dataset through three-dimensional scattering calculation, and recording multipath interference data; Removing the interference feature database and the multipath interference data from the sea surface wind and wave feature dataset to obtain the sea surface effective signal data.

4. The shore-star sea surface wind and wave observation data fusion method according to claim 1, wherein Constructing a historical sea condition reference library through ERA5 and NCEP reanalysis data, establishing a sea condition change model based on the historical sea condition reference library, and performing spatio-temporal correction processing on the sea condition change model to obtain sea surface state reference data, including: Performing data format standardization processing on ERA5 and NCEP reanalysis data, extracting sea condition correlation parameters, and generating standardized sea condition data; Performing time series analysis on the standardized sea condition data, separating seasonal and diurnal variation characteristics, and forming a sea condition time-varying characteristic sequence; Performing Bayesian correction on the outliers in the sea condition time-varying characteristic sequence to obtain sea condition anomaly correction data; Integrating and storing the standardized sea condition data and the sea condition anomaly correction data to construct the historical sea condition reference library; Extracting the sea condition change law from the historical sea condition reference library to establish the sea condition change model; Performing correction processing on the sea condition change model through spatio-temporal correlation analysis to obtain the sea surface state reference data.

5. The shore-star sea surface wind and wave observation data fusion method according to claim 1, characterized in that Fusing the sea surface effective signal data, shore-based observation data, and the sea surface state reference data according to weight allocation to obtain comprehensive observation data, including: Calculating the signal data weight coefficient according to the signal-to-noise ratio and geometric factor of the sea surface effective signal data; Extracting equipment performance indicators and environmental impact factors from the shore-based observation data to generate shore-based data weight coefficients; Obtaining the reference data weight coefficient based on the consistency analysis between the sea surface state reference data and the measured values; Performing data interpolation on the sea surface effective signal data, the shore-based observation data, and the sea surface state reference data through Kalman filtering to generate uniformly distributed data; Normalizing the data according to the signal data weight coefficient, the shore-based data weight coefficient, and the reference data weight coefficient to form weight fusion data; Combining the weight fusion data with the uniformly distributed data to obtain the comprehensive observation data.

6. The shore-star sea surface wind and wave observation data fusion method according to claim 1, characterized in that Performing wind-wave field reconstruction processing on the comprehensive observation data and adjusting parameters through ocean dynamics constraints to generate high-resolution wind-wave field distribution data, including: Performing sea surface dynamic characteristic analysis on the comprehensive observation data, extracting wind-wave coupling parameters, and obtaining wind-wave initial field data; Performing dynamic constraints on the wind-wave initial field data based on the energy balance equation to generate wind-wave constraint conditions; Extracting local feature structures from the comprehensive observation data to establish wind-wave field boundary conditions; Performing coupled calculation on the wind-wave constraint conditions and the wind-wave field boundary conditions to obtain wind-wave field iteration data; Performing data interpolation processing on the wind-wave field iteration data to establish refined grid data; Performing spatial distribution reconstruction according to the refined grid data to generate the high-resolution wind-wave field distribution data.

7. The shore-star sea surface wind and wave observation data fusion method according to claim 1, wherein Using independent observation data to perform cross-validation and parameter optimization on the high-resolution wind-wave field distribution data to obtain sea surface wind-wave observation results, including: Dividing the independent observation data by time period, performing batch verification on the high-resolution wind-wave field distribution data, and generating verification index data; Performing spatial consistency analysis based on the verification index data to construct an error distribution map; Perform sensitivity analysis on system parameters through the error distribution map to obtain parameter optimization indicators; Adjust the signal processing threshold according to the parameter optimization indicators to obtain threshold optimization data; Compare and process the threshold optimization data with the high-resolution wind-wave field distribution data to form parameter correction data; Update the high-resolution wind-wave field distribution data according to the parameter correction data to obtain the sea surface wind-wave observation results.

8. An onshore-star sea surface wind and wave observation data fusion system for implementing the onshore-star sea surface wind and wave observation data fusion method according to any one of claims 1-7, characterized in that The shore-satellite sea surface wind-wave observation data fusion system includes: An extraction module, configured to receive direct signals and sea surface reflection signals using a Beidou satellite receiver, extract features of the direct signals and the sea surface reflection signals according to Doppler frequency shift and time delay information, and establish a sea surface wind-wave feature data set; An identification module, configured to construct a land and reef distribution map of the detection area based on a geographic information system, perform interference identification and filtering on the sea surface wind-wave feature data set to obtain sea surface effective signal data; A correction module, configured to construct a historical sea condition reference library through ERA5 and NCEP reanalysis data, establish a sea condition change model according to the historical sea condition reference library, and perform spatio-temporal correction processing on the sea condition change model to obtain sea surface state reference data; A fusion module, configured to fuse the sea surface effective signal data, shore-based observation data, and the sea surface state reference data according to weight distribution to obtain comprehensive observation data; A reconstruction module, configured to perform wind-wave field reconstruction processing on the comprehensive observation data, and adjust parameters through ocean dynamics constraints to generate high-resolution wind-wave field distribution data; A verification module, configured to perform cross-verification and parameter optimization on the high-resolution wind-wave field distribution data using independent observation data to obtain sea surface wind-wave observation results.

Images