124 results about "Sea-surface height" patented technology

The invention relates to a method for automatically identifying and tracking mesoscale ocean eddy, belonging to the technical field of information. The method is mainly used for automatically extracting and tracking and constructing the process of the mesoscale ocean eddy. The realization method comprises the following steps of: providing an improved method for automatically identifying the eddy based on SSH (Sea Surface Height) on the basis of researching a method for automatically identifying the mesoscale ocean eddy; extracting the state of the eddy and tracking the process of the eddy by using a tracking method based on the closest distance and the similarity; and constructing a time and space process database of the mesoscale ocean eddy with a time and space data model based on topology reconstruction. The algorithm disclosed by the invention is simpler and more flexible; furthermore, the eddy time and space process data model based on the topology reconstruction is combined; the automatic identifying and the process tracking of the eddy can be tightly combined; the extraction and time and space process reconstruction of the eddy can be realized automatically; and strong support is provided for well exploring the process of the eddy and internal detail evolution law therein.

The invention discloses an external calibration method of a satellite-borne interferometric synthetic aperture radar (InSAR) system. The external calibration method comprises that sea surface height data is measured by a satellite height gauge, and average sea surface height data is acquired after time varying influence factors are removed; sea surface height data measured by the satellite-borne InSAR system is acquired through sea images and uncalibrated interference parameters of preset time and space distribution acquired by the satellite-borne InSAR system; sea height variation brought about by the time varying influence factors except for the spreading effect under the condition of the preset time and space distribution is calculated in a simulation mode; time varying synthesized sea surface height data is acquired by the average sea surface height and the sea surface height variation brought about by time varying influence factors except for the spreading effect; and the sea surface height data measured by the satellite-borne InSAR system is corrected by taking the time varying synthesized sea surface height data as standard data, thereby realizing calibration of the satellite-borne InSAR system. According to the invention, an external calibration method based on a sea field is adopted, thereby being capable of carrying out long-term dynamic monitoring on slow variation elements in error of the satellite-borne InSAR system.

The invention discloses a sea surface error correction method for measuring the submarine topography on the basis of airborne laser radar. The method comprises the steps that three-dimensional sea surface data is generated according to sea return pulses, a space rectangular coordinate system taking a laser radar launching point M as an original point is established, and coordinates of a sea surface incidence point in the space rectangular coordinate system are calculated; a three-dimensional sea surface coordinate system taking the sea surface incidence point as a coordinate original point, an incidence angle alpha obtained when a laser is incident on sea water and a refraction angle beta are calculated, then the submarine depth h can be calculated according to the incidence angle alpha, the refraction angle beta and the incidence distance L' of refracted rays, the submarine depth h is corrected according to the sea surface height delta h to obtain a submarine depth correction value h1, and then submarine topography data can be formed. According to the sea surface error correction method, errors caused by sea heaving can be effectively corrected, the problem that the precision is poor when the submarine topography is measured by traditional airborne laser radar is solved, and precise measuring of the submarine topography is achieved.

The invention discloses a tsunami motion forecasting method based on multi-hierarchy interaction, which comprises the following steps: 1) generating hierarchical grids according to the position of an actual tsunami source, establishing corresponding relation between the grids and spherical coordinates, and establishing an interaction mechanism of the hierarchical grids; 2) carrying out sector and square combined dispersion on a tsunami motion equation, setting initial conditions and boundary conditions according to actual conditions, and solving a wave equation to compute flux change of tsunami motion; and 3) setting an initial sea level height field according to the height of a sea level, establishing an interaction mechanism of particles and the grids in the process of tsunami motion, establishing updating and iterating rules of the particles, and then, forecasting the height of tsunami through interaction of the particles and the grids. The method can quickly and vividly forecast the tsunami propagation phenomenon and solves the problems that the existing simulation method only pays attention to numerical simulation, the simulation result is not visual, the computing efficiency is too slow, and the like. The method has certain practical value in the fields of disaster prevention, tsunami early warning, virtual simulation etc.

The invention relates to a high-precision sea surface height extracting method of a radar altimeter through small incidence angle interference. A small incidence angle and short base line interference processing mode is adopted to carry out synchronous interference receiving on echo signals through a main channel and an auxiliary channel, pre-filtering, two-dimensional image processing, image registration, interference processing, multi-vision processing, phase-position denoising filtering, phase position unwinding, water recognizing, classifying, height conversion and geometric correcting processing are carried out on received data in sequence, and a water area height value is obtained. According to the method, the swath width is large, the temporal-spatial resolution is high, the method can be applied to observation of the ocean current, ocean tide, vortex and other meso and micro scale phenomena in the ocean and the observation of the coastal water, the sea ice shelf and the land hydrology.

The invention belongs to the crossing field of physical oceanography and computer graphics and image processing, and particularly relates to a mesoscale eddy tracing method based on a hybrid algorithm. The hybrid algorithm mainly comprises nearest neighbor search, deformation control based similarity match and delay logic. The method comprises the following steps: step one, for the nearest neighbor search, eddies in a search range are delineated according to global mesoscale eddy data recognized with an SSH (sea surface height) method; step two, for a deformation control based similarity match method, the eddies in the range and attributes of the eddies are subjected to similarity calculation of area, amplitude, kinetic energy, relative vorticity and Hausdorff distance, the eddy with the maximum similarity is selected as the position of the next eddy, and jump of an eddy path is avoided through combination of physical attributes and geometric attributes of the eddies; step three, the delay logic is adopted, search at multiple time points is considered, the eddies temporarily disappearing at certain time points are processed, and discontinuity of the eddy path is avoided, so that the purpose of multi-year long-term efficient tracing on the eddies is achieved.

The invention discloses a method for determining sea surface topographic structure based on barodynamics, which comprises the following steps: S1, performing the gross error detection on sea surface height data to obtain a gross error-free height data; S2, calculating a reference gravity field in satellite altimetry gravity field inversion; S3, based on the reference gravity field, deleting long-wave components in the sea surface topography from the gross error-free height data; S4, utilizing the gross error-free height data to separate the intermediate wave components in the sea surface topography; and S5, in the calculation of a geoid height by utilizing the reference gravity field, adding parameters which are acquired in the step of separating the intermediate wave components in the sea surface topography, and comparing the added result with a sea surface height acquired according to a satellite altimetry data so as to determine various frequency components in the sea surface height. By using the method, the reliable sea surface height data is acquired, thereby acquiring the accurate measured sea surface topographic structure data.

The invention discloses a method for inverting an ocean gravity field by using imaging type altimeter data. The method belongs to the field of ocean measurement, and comprises the steps of: determining plumb line deviation and inverting an ocean gravity field for a satellite height-finding technology; considering mutual restriction of precision and resolution of data of the traditional subsatellite point altimeter, and meanwhile, wherein the effective spatial resolution of the ocean gravity field is limit due to large distance between tracks; utilizing high-precision sub-satellite point altimeter and a high-resolution synchronous swath altimeter of an imaging type altimeter satellite to obtain sea surface height observation data, balancing joint intersection points of observation data of the sub-satellite point altimeter and the swath altimeter to improve the overall sea surface height precision after filtering out high-frequency noise; and then performing differential calculation to obtain rich plumb line deviation information and inverting the ocean gravity field. By adopting the method, the defect that the traditional sub-satellite point altimeter is insensitive to a plumb linedeviation east component can be overcome, and a high-resolution ocean gravity field with uniform precision in each direction can be inversed.

The invention relates to an underwater temperature field reconstruction method based on a self organizing neural network and an empirical orthogonal function. A self-organizing characteristic mappinggraph including multidimensional information including an empirical orthogonal function coefficient, position information, time information, sea surface height and sea surface temperature corresponding to a temperature profile is established, an Euclidean distance between known information and a self-organizing characteristic mapping unit is used to determine an optimal matching unit, and the empirical orthogonal function coefficient to inverse is obtained. A sea surface parameter and water body temperature profile characteristic mapping network is established based on a lot of data information, and nonlinear mapping from sea surface parameters to the water body profile can be realized. The underwater temperature field reconstruction method based on the self organizing neural network and the empirical orthogonal function is excellent in performance, high in stability, needless of knowing a power process in the marine site, easy to realize, low in computing complexity, and suitable forobtaining ocean environmental parameters of the key marine site in quasi-real time by using satellite remote sensing data, and only uses correlation among the ocean environmental parameters.

The invention provides a method and a system for sea ice thickness distribution estimation based on least square adjustment. The method comprises the following steps: carrying out satellite height measurement data processing, and extracting related parameters which comprise longitude, latitude, elevation, average sea surface height, orbit number, pulse peak value and pulse accumulation standard deviation; extracting a sea ice freeboard height: performing inter-ice waterway identification according to satellite echo signal related characteristic parameters to obtain the sea ice freeboard height; estimating the sea ice thickness: establishing a function model in a grid unit according to a conversion relationship from the sea ice freeboard to the sea ice thickness, establishing an observationequation, and solving by utilizing least square adjustment to obtain the corresponding sea ice thickness result; and performing resampling and filling to obtain a final sea ice thickness distributiondiagram. The method gets rid of a traditional experience valuing method, avoids uncertainty caused by conditions such as accumulated snow thickness, accumulated snow density and the like under special geographical positions and weather conditions of a sea ice area, and accordingly improves the reliability of sea ice thickness estimation.

The present invention discloses a method for the production of mean sea height products based on satellite altimeter data network function interpolation, which comprises the steps of: data preprocessing, determination of grid size and central point, grid interpolation, and refined interpolation. The method according to the present invention facilitates automatic grid segmentation on the computer, has obvious statistical characteristics, reflect the spatial distribution of the marine gravity field on the physical properties, wherein the marine gravity field is in a positive proportion to mean sea surface height, gives full consideration to the physical characteristics of sea level, and can improve mean sea height products of high accuracy and spatial resolution.

The invention discloses a device for measuring sea surface height based on time-delay modification. The device comprises a navigation receiver, a right-hand antenna, a left-hand antenna, a radio-frequency front end module, an intermediate frequency signal processing module and an inversion module. The navigation receiver is connected with the right-hand antenna, the intermediate frequency signal processing module and the inversion module, and the right-hand antenna and the left-hand antenna are independent with each other and connected with the radio-frequency front end module, the intermediate frequency signal processing module and the inversion module sequentially; a direct radio-frequency voltage signal transmitted from the right-hand antenna and a reflective radio-frequency voltage signal transmitted from the left-hand antenna are adopted as the input of the radio-frequency front end module, a direct digital intermediate frequency signal and the reflective digital intermediate frequency signal transmitted from the radio-frequency front end module are adopted as the input of the intermediate frequency signal processing module, time-delay doppler two-dimensional related power transmitted from the intermediate frequency signal processing module is adopted as the input of the inversion module, the direct radio-frequency voltage signal transmitted from the right-hand antenna is adopted as the input of the navigation receiver, and navigation information transmitted from the navigation receiver is adopted as the input of the intermediate frequency signal processing module and the inversion module.

The invention provides a novel method for inverting a sound velocity profile of a whole sea area by using sea surface height data, sea surface temperature data and feedback temperature data based on an empirical orthogonal function. The method comprises the following steps: S1, acquiring an Argo sound velocity profile and sea surface remote sensing data; s2, reconstructing the sound velocity profile based on an empirical orthogonal function; and s3, performing sound velocity profile inversion based on the sea surface parameters; the invention further provides a method for optimizing the regression relation by using the seawater layer temperature data with the poor inversion effect and application of the method. Compared with a previous inversion method based on EOF decomposition of a first mode, the method considers more modes, optimizes a sound velocity profile inversion method, improves the correlation between the sea surface data and each order of mode of the sound velocity profile, and improves the performance of the inversion method in each sea area. And a powerful basis is provided for applying inversion sound velocity profile data to the fields of underwater sound detection, marine environment monitoring, sonar performance evaluation and the like in the future.

The invention discloses a method and a system for elevation precision control and correction of island remote sensing mapping, and relates to the technical field of geodetic surveying engineering. The method comprises the steps of S1, extracting shorelines from island remote sensing images; S2, obtaining a first average sea surface height and a second average sea surface height, and performing datum conversion on the second average sea surface height; S3, calculating geoidal surface; S4, carrying out harmonic analysis to obtain a tidal harmonic constant; S5, calculating elevation of mean high-water springs; S6, calculating and removing systematic errors of the shorelines in the island remote sensing images; and S7, fitting the shorelines in the island remote sensing images with the systematic errors removed and the shorelines on the mean high-water springs. Correction of the shorelines in the island remote sensing images is realized by fitting the fitting the shorelines in the island remote sensing images with the systematic errors removed and the shorelines on the mean high-water springs, thereby increasing the elevation precision obtained by the island remote sensing mapping.

The invention provides an intelligent buoy networking simulation method and a system thereof for ocean phenomena, comprising: identification of ocean phenomena: establishing an ocean three-dimensionalthermohaline environment and an ocean three-dimensional fluid environment, and identifying the ocean phenomena by combining the sea surface height anomalies measured by the ocean altimeter accordingto the monitored flow field change information, the thermohaline degree profile information and the information including intelligent buoy drift characteristics during the movement of the intelligentbuoy; and networking simulation assembly of the intelligent buoy: inputting the required parameters according to the dynamic model of the intelligence buoy by regarding the identified ocean phenomenaas an object to obtain the total time and total energy consumption of the whole observation process, and respectively establishing a shortest time model and an optimal energy consumption model of theintelligent buoy networking simulation observation by the observation path, the shortest principle and the observation energy consumption mechanism of the intelligent buoy. The motion trajectory of the invention is more realistic and the network simulation is efficient, the ocean data obtained by the buoy is highly available, and the energy consumption during the simulation observation process islow.

The invention discloses a satellite radar altimetric waveform resetting method based on waveform derivatives. The method comprises the steps that a first-order derivative and a second-order derivativeof an echo waveform are deduced based on a theoretical echo model; a computational formula of a leading edge midpoint is obtained by use of a waveform leading edge slope maximum value obtained when the second-order derivative is equal to zero; a fitting mode is utilized to solve unknown parameters in the computational formula of the leading edge midpoint, and the position of the leading edge midpoint is determined; a reset distance corrected value is determined according to a preset tracking gate; and sea surface height reset through the resetting method is further calculated according to thereset distance corrected value and an altimetric error corrected value. Through the method, the quality of satellite altimetric data can be improved, the precision of the reset sea surface height ishigher than that obtained through a conventional resetting method, and the correlation between the reset sea surface height and geoidal surface height is obviously improved.

A method for forecasting the central fishery of squid in the North Pacific Ocean, including time-space scale setting, environmental factor setting, and establishment of a central fishery prediction model. The factor setting uses surface temperature (SST) as the main environmental factor, supplemented by two environmental factors, sea surface height (SSH) and chlorophyll a (Chl-a), and the environmental factors are divided into four situations when establishing the central fishery prediction model: According to the setting of time-space scale and environmental factors, a sample plan set of 24 situations is established; the central fishery prediction model adopts the classic error backpropagation BP neural network model, and the BP neural network model has a three-layer structure, namely the input layer and the hidden layer And the output layer, the input layer inputs the spatio-temporal factors and environmental factors of the fishery, and the output layer outputs CPUE or the fishery level index converted from CPUE for forecasting.

The invention provides a tide level measuring method based on SNR (signal to noise ratio) data of a continuously operating GNSS (global navigation satellite system) station. The method comprises the following steps: establishing an interference model for a direct signal of a shore-based GNSS station and a right-handed circular polarization reflected signal after reflection by sea surface; generating an SNR data file according to a standard RINEX file to obtain SNR data waveforms; performing fitting processing on the SNR data waveforms with the interference model to obtain the height of an effective reflecting surface of each GNSS satellite in each rising or falling process; performing summation on the height of the effective reflecting surface obtained in each rising or falling process of each GNSS satellite in a predetermined period and sea surface height obtained by a tide level measuring instrument, and taking an arithmetic mean value to obtain the mean height between the phase center of an GNSS antenna and the zero-meter sea surface; calculating the difference between the mean height of the phase center of the GNSS antenna away from the zero-meter sea surface and the height of the effective reflecting surface of each GNSS satellite in the rising or falling process to obtain the tide level. The tide level measuring method is simple, and measuring results are accurate.

The invention relates to an absolute calibration method for a satellite radar altimeter without a special calibration field and belongs to the technical field of calibration of satellite altimeters. According to the method, an absolute deviation of the satellite radar altimeter can be determined through processing tidal observation data and satellite altimetry data of a satisfactory oceanographic station, adding essential correction entries such as a geoid model, carrying out height datum unification, altimetry data space position interpolation and tidal observation data time interpolation, and finally, carrying out difference comparison on sea surface height of tidal observation and sea surface height of the satellite radar altimeter. Compared with the prior art, the method has the advantages that the special calibration field is not required, the existing oceanographic stations can have calibration capability by using existing business observation data of oceanographic stations of coasts or oceanic islands of our country and a regional accurate geoid, and thus, absolute measured deviations of the satellite altimeters at home and abroad can be acquired.

The invention discloses a method for utilizing a multifunctional buoy for tide monitoring. The method aims at achieving tide monitoring with a high-precision GPS as a basis and through inversion of the sequence of the sea surface height by means of buoy posture correction, abnormal point removal, curve fitting and other modes. By means of the method, the situation that the measurement range is influenced by the position limitation of a tide station in traditional tide measurement work can be avoided, the monitoring precision is effectively improved, and precise tide monitoring and forecastingcan be achieved as expected.

The invention discloses an underwater mobile channel simulation method. The method comprises the following steps: 1) setting environmental parameters and signal parameters of a three-dimensional underwater space communication scene, establishing a motion model of an acoustic source emission platform and a receiving platform and a sea surface height model; 2) real-timely sampling the state of the channel by calling the ray track model for many times to obtain a series of channel snapshots at different time, and calculating a time-varying channel impulse response according to the obtained snapshots; 3) dynamically filtering the transmit signal of the acoustic source transmitting platform in the time domain by using the time-varying channel impulse response to obtain the received signal. Themethod of the invention ensures the accuracy and the precision of the underwater mobile communication channel simulation to a certain extent, and at the same time, the complexity of the simulation realization can be effectively reduced, the memory requirement can be reduced, and the computer operation efficiency can be improved.

The invention discloses a construction method of a three-dimensional ocean current real-time rapid analysis system based on satellite remote sensing; the method comprises the steps: assimilating on-site observation temperature and salinity profile data through a three-dimensional temperature and salinity field obtained through the vertical inversion of real-time satellite observation sea surface temperature (SST), sea surface salinity (SSS) and sea surface height anomaly (SSHa), and obtaining a three-dimensional temperature and salinity analysis field; and carrying out earth current conversion correction on the three-dimensional ocean current reanalysis background field, observing a sea surface wind field by using a satellite to calculate a wind ocean current, carrying out wind ocean current correction on the three-dimensional ocean current reanalysis background field, and constructing a three-dimensional ocean current real-time rapid analysis system based on a visual platform. The system has the characteristics of miniaturization and rapid analysis, and can be loaded on a ship to provide real-time marine environment condition guarantee for the ship, including sea surface height, three-dimensional ocean current, temperature, salinity, sound field and the like.

The invention provides an anisotropic sea surface radar back scattering simulation method and system. The method comprises the steps of: observing a data sample according to a plurality of satellite radars, carrying out fitting calculation on parameters in a to-be-configured anisotropic sea surface radar back scattering model which is established on the basis of a small slope approximation method,so as to obtain a configured anisotropic sea surface radar back scattering model, in the model, expressing asymmetric influences, on radar back scattering coefficients, of along wind and dead wind through a deflection function, and expressing autocorrelation of a sea surface height by adoption of an index-like type function; and obtaining data of a to-be-simulated field, and calculating and outputting a target radar back scattering coefficient corresponding to the data of the to-be-simulated scene according to the configured anisotropic sea surface radar back scattering model. According to the method and system, the characteristics of anisotropic sea surfaces can be truly reflected, and the calculation precision and calculation efficiency are both reasonably considered, so that the methodand system have important application value in the aspects of satellite-borne radar parameter design, sea surface wind field inversion and the like.

The invention provides a data processing method and device for a multi-source satellite altimeter, equipment and a storage medium. The data processing method comprises the steps of performing data quality control on data of a to-be-processed satellite altimeter and a Jason-3 satellite altimeter; by taking the Jason-3 satellite altimeter as a reference, performing data unification on height measurement data of the to-be-processed satellite altimeter after data quality control; by taking the Jason-3 satellite altimeter as a reference, performing orbit error correction on the height measurement data of the to-be-processed satellite altimeter after data unification; and calculating sea surface height abnormal data of each satellite altimeter according to the height measurement data after orbit error correction. In the implementation process, before the sea surface height abnormal data of each satellite altimeter is calculated, data unification and orbit error correction of the height measurement data of the satellite altimeters are performed, and the sea surface height abnormal data is calculated according to the height measurement data after orbit error correction, so that the accuracy of a sea surface height abnormal result is greatly improved.

The invention relates to the technical field of vortex identification, and provides a mesoscale vortex identification method based on sea surface height anomaly and rotation speed, which comprises the following steps of: 1) judging possible air vortexes, and aiming at sea surface height anomaly data with the spatial resolution of 0.25 degree, determining a rotation speed; in a 1-degree * 1-degree longitude and latitude moving window, finding a minimum or maximum value of a sea surface height abnormal local, so as to judge the center of a possible air vortex or anti-air vortex; and 2) determining a sea surface height abnormal closed isoline, finding out the sea surface height abnormal closed isoline from the inside of each possible air vortex or anti-air vortex center at an amplification or reduction amplitude of 1cm, and determining the sea surface height abnormal closed isoline meeting a five-point condition. By recognition steps, the mesoscale vortex in the ocean can be effectively identified, the outer boundary range and the vortex type of the mesoscale vortex can be accurately determined through calculation, and an accurate guidance basis is provided for people to find, develop and utilize the mesoscale vortex.

A method for predicting central fishing grounds of Argentinian soft fish, including time-space scale setting, environmental factor setting, and establishment of a central fishing ground prediction model, which is characterized in that the time-space scale setting adopts three levels of spatial scales, and two levels of time scales of weeks and months; Environmental factors are set using surface temperature (SST) as the main environmental factor, supplemented by two environmental factors, sea surface height (SSH) and chlorophyll a (Chl-a), and the environmental factors are divided into four situations when establishing the central fishery prediction model : According to the setting of time-space scale and environmental factors, a set of sample plans for 24 situations is established; the central fishery prediction model adopts the classic error backpropagation BP neural network model, and the BP neural network model has a three-layer structure, namely the input layer and the hidden layer And the output layer, the input layer inputs the spatio-temporal factors and environmental factors of the fishery, and the output layer outputs CPUE or the fishery level index converted from CPUE for forecasting.

Embodiments of the invention provide a model rendering method and device, an electronic device and a storage medium. The method comprises the steps of obtaining preset sea surface height information and model height information corresponding to a to-be-rendered grid model; determining a target grid in a plurality of grids of the grid model by adopting the model height information and the sea surface height information; performing mapping mixing on the target grid to obtain target mapping information; and rendering the target grid by adopting the target mapping information. Through the embodiment of the invention, the rendering effect of material coverage is automatically generated according to the sea surface height, the situation that detail materials cannot be presented due to limitationof the model position in the scene is avoided, the target grid is rendered through mapping mixing, the method and device, the electronic device and the storage medium are simple and easy to use, andthe rendering efficiency is improved.

The invention discloses a method for recognizing tidal mixing frontal surface with satellite altimeter along-track data.A non-dimensional momentum equation is obtained through dimensional analysis according to satellite altimeter 20 Hz along-track data, on the sea surface, the second vertical gradient of tidal stress is neglected, the momentum equation obtained in step 1 is subjected to integral from the shore side A to the offshore side B of the frontal surface to obtain an sea surface height SSH, the obtained sea surface height is subjected to EMD decomposition to obtain a series of IMF components, an HHT method is applied to each IMF component, amplitude functions and phase functions of the IMF components are utilized to represent the obtained sea surface height, and therefore the sea surface height is converted to be an Hilbert spectrum equation, so that an Hilbert spectrum of the sea surface height is obtained.The HHT method is utilized to decompose the 20 Hz along-track data into the IMF components, and an SSH saltus step can be recognized clearly through the HHT method.

The present invention discloses a satellite borne double-side looking wide swath interference imaging height gauge base line inclination angle estimation method. The method comprises the steps of: setting a range and a step length of a base line inclination value according to a right-side looking base line inclination measurement value to obtain a plurality of base line inclination values; inverting the sea surface height in each base line inclination value to obtain a left-side looking sea surface height vector and a right-side looking sea surface height vector; making difference of the left-side looking sea surface height vector and the right-side looking sea surface height vector to obtain a left-side looking differential vector and a right-side looking differential vector; calculatingaverage values in the swathes of the left-side looking differential vector and the right-side looking differential vector to draw two differential curves; and taking any one differential curve to perform x-axis symmetrical reversal to obtain a reversal differential curve, and calculating a base line inclination value of an intersection point of the reversal differential curve and the other differential curve as a right-side looking base line inclination estimation value of a height gauge, wherein the left-side looking base line inclination estimation value is a value by subtracting the right-side looking base line inclination estimation value from [Pi].

The invention discloses a method for improving height measurement precision based on machine learning weighted average fusion feature extraction, which utilizes a global navigation satellite system reflectometer to perform sea surface height measurement and can improve observation and inversion of a mesoscale process of an ocean by increasing space coverage of ocean surface observation. In order to make up the defects of a traditional inversion method, a novel machine learning weighted average fusion feature extraction method with airborne time delay waveform data as input and sea surface height as output is constructed on the basis of a machine learning fusion model in combination with a feature extraction principle. Meanwhile, two features of HALF and DER which are sensitive to sea surface height change are constructed based on a time delay waveform data set, and the influence of feature sets of different information details on sea surface height inversion precision is analyzed. The inversion precision of the sea surface height can be effectively improved by adopting a novel machine learning weighted average fusion feature extraction method, and the precision is improved by about 61%.