A method and device for detecting the physical state of the ocean surface based on high-resolution satellite-borne GNSS-R complex waveform data

Abstract

A method and device for detecting high-resolution ocean surface physical state based on satellite-borne GNSS-R complex waveform data, the method comprising: acquiring complex waveform data; preprocessing the complex waveform data to obtain a complex waveform data sequence for detecting the ocean surface physical state, dividing the detection complex waveform data sequence into multiple complex waveform groups in chronological order, and determining the coherence coefficient between two consecutive complex waveforms in each complex waveform group; taking the complex waveform data sequence as a test set, selecting 20% of the data in the test set as a training set, determining the threshold value of the average coherence coefficient feature quantity according to the validation data and the training set using the histogram method, taking the threshold value of the average coherence coefficient feature quantity as the demarcation line for judging sea ice and seawater, and obtaining the surface physical state of the detection area determined by the multiple sampling points corresponding to each complex waveform group in the test set according to the threshold value and the average coherence coefficient of each complex waveform group in the test set.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of satellite remote sensing, in particular to a method and device for detecting the physical state of the ocean surface with high resolution based on complex waveform data of a satellite-borne GNSS-R. BACKGROUND

[0002] Sea ice is an important variable in global climate models and an important component of the high-latitude ocean. Sea ice covers most of the ocean surface in polar regions, and about 71% of the Earth's surface is covered by oceans, of which about 7% to 15% is covered by sea ice. Under the background of global warming, the sea ice coverage area in the polar regions of the north and south poles is decreasing, which is deeply affecting the natural ecological environment and human economic society in the polar regions and the world, and therefore, accurate and efficient acquisition of sea ice range and change information is of great significance to marine transportation, ocean resource development, global climate change and ecological system stability.

[0003] The initial sea ice monitoring mainly relies on in-situ measurement means such as icebreakers and buoys, which provide accurate sea ice information, but the coverage is limited, the measurement process is time-consuming and labor-intensive, and it is difficult to achieve long-term, large-scale and efficient dynamic monitoring. Satellite remote sensing originated in the early 1970s, mainly using visible, near-infrared and microwave (active / passive) sensors for earth observation, however, optical remote sensing (visible, near-infrared) is easily affected by cloud cover and light conditions, and cannot achieve all-weather sea ice monitoring, and microwave remote sensing is difficult to balance the spatial and temporal resolution. With the development of satellite technology, satellite-borne GNSS-R as a new emerging remote sensing means has the advantages of low cost, low power consumption, all-weather, all-weather, large scale, high spatial and temporal resolution, and as a new type of remote sensing technology, it shows great potential in the field of sea ice remote sensing. Therefore, the reflected signal collected by the satellite-borne GNSS-R task can realize sea ice detection.

[0004] Before this, a plurality of feature observations are usually extracted from the delay Doppler map (DDM map) collected by the satellite-borne GNSS-R task for sea ice detection. However, the along-track resolution of the satellite-borne GNSS-R task, i.e. the spatial distance between the centers of two consecutive GNSS-R data signal peak (SP) points, is affected by the 1-second incoherent averaging process in the DDM product generation process, resulting in an along-track resolution usually lower than 6 kilometers. This limitation means that the traditional DDM product is difficult to achieve high-resolution sea ice detection, and cannot meet the demand for accurate monitoring of the fine structure and dynamic changes of sea ice. SUMMARY

[0005] Therefore, the first aspect of the present application provides a method for detecting the physical state of the ocean surface with high resolution based on complex waveform data of a satellite-borne GNSS-R, the method comprising:

[0006] Obtaining complex waveform data obtained by cross-correlation processing of a reflected signal received by a space-borne GNSS-R and a local copy code, the complex waveform data including amplitude and phase information of the reflected signal obtained at a high sampling rate of 1 kHz, and the reflected signal being a satellite signal emitted by a target satellite and reflected by the earth's surface, the complex waveform data being capable of being detected at a controllable along-track resolution by selecting a coherent and incoherent number;

[0007] Preprocessing the complex waveform data to obtain a complex waveform data sequence for detecting a physical state of an ocean surface, wherein the complex waveform data sequence includes a plurality of complex waveforms, each complex waveform corresponding to a sampling point of the ocean surface;

[0008] Dividing the complex waveform data sequence into a plurality of complex waveform groups in time sequence, and determining a coherence coefficient between any two consecutive complex waveforms in each complex waveform group, wherein the number of complex waveforms in each complex waveform group ranges from 1 to 20, and each complex waveform corresponds to a time resolution of 1 ms;

[0009] According to all coherence coefficients of each complex waveform group, extracting an average coherence coefficient feature quantity of the complex waveform group with a time resolution of 20 ms, which quantifies the degree of phase change;

[0010] Taking the complex waveform data sequence as a test set, selecting 20% of the data in the test set as a training set, and determining a threshold value of the average coherence coefficient feature quantity according to the validation data and the training set by using a histogram method;

[0011] Taking the threshold value of the average coherence coefficient feature quantity as a demarcation line for judging sea ice and sea water, and obtaining a surface physical state of a detection area determined by a plurality of sampling points corresponding to each complex waveform group in the test set according to the threshold value and the average coherence coefficient of each complex waveform group in the test set.

[0012] According to an embodiment of the present application, the preprocessing method of the complex waveform data includes:

[0013] Obtaining initial complex waveform data;

[0014] Filtering the initial complex waveform data by using a land mask method to filter out complex waveforms corresponding to sampling points located on land or within a preset distance from land, to obtain filtered complex waveform data;

[0015] Evaluating the quality of the filtered complex waveform data by using a signal-to-noise ratio, and retaining filtered complex waveform data with a signal-to-noise ratio greater than a preset threshold value, to obtain a complex waveform data sequence for detecting a physical state of an ocean surface.

[0016] According to an embodiment of the present application, the coherence coefficient between two continuous complex waveforms is represented as follows:

[0017]

[0018] wherein, represents the coherence coefficient of the two continuous complex waveforms, t c represents the time interval of the two continuous complex waveforms, represents the complex waveform value at the complex waveform peak power at time t m represents the complex waveform value at the complex waveform peak power at time t m +t c represents the complex waveform value at the complex waveform peak power at time t represents the complex waveform value at the complex waveform peak power at time t m represents the complex waveform value at the complex waveform peak power at time t m +t c represents the complex waveform value at the complex waveform peak power at time t is the conjugate of

[0019] According to an embodiment of the present application, each complex waveform group comprises M complex waveforms arranged in time sequence;

[0020] According to all the coherence coefficients of each complex waveform group, the 20ms time resolution complex waveform group average coherence coefficient feature quantity quantifying the degree of phase variation is extracted, comprising:

[0021] The average coherence coefficient feature quantity of the complex waveform group is obtained by averaging all the coherence coefficients in the complex waveform group, and the average coherence coefficient feature quantity is represented as follows:

[0022]

[0023] wherein, represents the average coherence coefficient feature quantity.

[0024] According to an embodiment of the present application, the surface physical state comprises sea ice state and sea water state;

[0025] According to the threshold value and the average coherence coefficient of each complex waveform group in the test set, the physical state of the surface of the detection area determined by the plurality of sampling points corresponding to each complex waveform group in the test set is obtained, comprising:

[0026] In each complex waveform group in the test set, the average coherence coefficient feature quantity is compared with the average coherence coefficient feature quantity threshold value, and in the case that the average coherence coefficient feature quantity is greater than the average coherence coefficient feature quantity threshold value, the physical state avoided by the detection area determined by the M sampling points is sea ice state; in the case that the average coherence coefficient feature quantity is less than the average coherence coefficient feature quantity threshold value, the physical state of the surface of the detection area determined by the M positions is sea water state.

[0027] According to an embodiment of the present application, 20% of the data of the test set is selected as the training set, and the threshold of the average coherence coefficient feature quantity is determined according to the validation data and the training set by using the histogram method, including:

[0028] The training data set is grouped to obtain a plurality of complex waveform group samples, and the average coherence coefficient feature quantity of each group is calculated;

[0029] The validation data that is spatiotemporally matched with each complex waveform group sample in the training set is obtained, and the validation data represents the actual physical state of the target area surface corresponding to the complex waveform group sample;

[0030] According to the actual physical state given by the validation data, a histogram of the average coherence coefficient feature quantity of the sea ice and the sea surface is drawn, and the intersection of the sea ice and the sea histogram is determined by using the histogram method, and the intersection is the threshold of the average coherence coefficient feature quantity.

[0031] According to an embodiment of the present application, the preset threshold is 3db, and the preset distance is 10 kilometers.

[0032] As a second aspect of the present application, a device for detecting the physical state of the ocean surface based on the complex waveform data of the satellite-borne GNSS-R is also provided, which is used to implement the method as described above, and the device comprises:

[0033] An acquisition module is configured to acquire the complex waveform data, which is obtained by cross-correlation processing of the reflected signal received by the satellite-borne GNSS-R and the local copied code, and includes the amplitude and phase information of the reflected signal acquired at a high sampling rate of 1 kHz, and the along-track resolution of the detection can be controlled by selecting the coherent and incoherent times; the reflected signal is the satellite signal emitted by the target satellite and reflected by the earth surface;

[0034] A preprocessing module is configured to preprocess the complex waveform data to obtain a complex waveform data sequence used for detecting the physical state of the ocean surface; wherein the complex waveform data sequence includes a plurality of complex waveforms, and each complex waveform corresponds to a sampling point of the ocean surface;

[0035] A grouping module is configured to group the detection complex waveform data sequence into a plurality of complex waveform groups in time sequence, and determine the coherence coefficient between any two continuous complex waveforms in each complex waveform group, wherein the number of complex waveforms in each complex waveform group ranges from 1 to 20, and each complex waveform corresponds to a time resolution of 1 ms,

[0036] A first determination module is configured to extract the average coherence coefficient feature quantity of the complex waveform group with a 20ms time resolution of the quantized phase variation degree according to all the coherence coefficients of each complex waveform group;

[0037] The second determining module is adapted to take the complex waveform data sequence as a test set, take 20% of data of the test set as a training set, and determine a threshold of the average coherence coefficient feature quantity according to the verification data and the training set by using a histogram method;

[0038] The detecting module is adapted to take the threshold of the average coherence coefficient feature quantity as a demarcation line for distinguishing sea ice and sea water, and is adapted to obtain a surface physical state of a detection area determined by a plurality of sampling points corresponding to each complex waveform group in the test set according to the threshold and the average coherence coefficient of each complex waveform group in the test set.

[0039] The complex waveform data (CWF data) sequence according to the embodiment of the present application is obtained according to a TDS-1CWF product, which is obtained according to a cross-correlation processing of a reflection signal received by a TDS-1 and a local copy code, and has a sampling rate of 1 kHz. Since the number of complex waveforms included in each complex waveform group ranges from 1 to 20, the method in the embodiment of the present application is equivalent to detecting the ocean surface at a sampling rate of 1-20 ms. Therefore, the method in the embodiment of the present application has a high resolution, and provides a possibility for detecting the ice-water state in a small-scale area. BRIEF DESCRIPTION OF DRAWINGS

[0040] The above and other objects, features and advantages of the present application will become more apparent from the following description of the embodiments of the present application taken with reference to the accompanying drawings, in which:

[0041] Figure 1 A flow chart of a method for detecting a physical state of an ocean surface at a high resolution based on space-borne GNSS-R complex waveform data according to an embodiment of the present application is shown;

[0042] Figure 2 A flow chart of a method for detecting a physical state of an ocean surface at a high resolution based on space-borne GNSS-R complex waveform data according to another embodiment of the present application is shown;

[0043] Figure 3 A flow chart of a method for detecting a physical state of an ocean surface at a high resolution based on space-borne GNSS-R complex waveform data according to another embodiment of the present application is shown;

[0044] Figure 4A A CWF track of an ice channel according to an embodiment of the present application is shown;

[0045] Figure 4B A CWF track in Figure 4A corresponding to a MODIS surface reflectivity and Figure 4A MCC feature quantity obtained by a track point at the CWF track shown in

[0046] Figure 4C It shows Figure 4B The correspondence between MODIS surface reflectance and MCC characteristic quantity after scale unification;

[0047] Figure 4D It shows Figure 4C Scatter plot comparing surface reflectance and MCC characteristic values ​​in [the data].

[0048] Figure 5A The CWF trajectory of a sea ice region provided according to an embodiment of the present invention is shown;

[0049] Figure 5B It shows Figure 5A MODIS surface reflectance corresponding to local CWF trajectories and Figure 5A MCC feature quantities obtained from trajectory points at local CWF trajectories;

[0050] Figure 5C It shows Figure 5B The correspondence between MODIS surface reflectance and MCC characteristic quantity after scale unification;

[0051] Figure 5D It shows Figure 5C Scatter plot comparing surface reflectance and MCC characteristic values ​​in [the data].

[0052] Figure 6 A schematic diagram of an apparatus for detecting the physical state of the ocean surface according to an embodiment of the present invention is shown. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0054] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0055] All terms used herein, including technical and scientific terms, have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0056] In the case of using expressions similar to "at least one of A, B, and C", etc., it will be understood that the meaning is that the item is at least one of A, B, or C, etc. In the case of using expressions similar to "at least one of A, B, or C", etc., it will be understood that the meaning is that the item is at least one of A, B, or C, etc.

[0057] It should also be noted that the directional terms mentioned in the embodiments, such as "upper", "lower", "front", "back", "left", "right", etc., are only the directions of the drawings, and are not intended to limit the scope of protection of the present application. Throughout the drawings, the same elements are represented by the same or similar reference numerals. When it is possible to cause confusion in the understanding of the present application, the conventional structure or configuration will be omitted.

[0058] In the embodiments of the present application, the collection, updating, analysis, processing, use, transmission, provision, invention, storage, etc. of the data involved (for example, including but not limited to user personal information) comply with the relevant legal regulations, are used for legal purposes, and do not violate public order and good customs. In particular, necessary measures are taken for user personal information to prevent illegal access to user personal information data and to maintain user personal information security and network security.

[0059] Figure 1 A flowchart of a method for high-resolution detection of ocean surface physical state based on spaceborne GNSS-R complex waveform data according to an embodiment of the present application is shown.

[0060] As shown in Figure 1 The method for high-resolution detection of ocean surface physical state based on spaceborne GNSS-R complex waveform data includes operations S1-S6.

[0061] In operation S1, complex waveform data is obtained, which is obtained by cross-correlation processing of a reflected signal received by a spaceborne GNSS-R and a local copy code. The complex waveform data includes amplitude and phase information of the reflected signal obtained at a high sampling rate of 1 kHz. By selecting the number of coherent and incoherent times, the along-track resolution of the detection can be controlled. The reflected signal is a satellite signal emitted by a target satellite and reflected by the earth's surface.

[0062] According to an embodiment of the present invention, the cross-correlation process can be expressed as the following formula:

[0063]

[0064] In equation (1), It is a complex waveform of the reflected signal, where t0 represents the start time of the cross-correlation, represents the time of each complex waveform, τ is the waveform delay, and T is the time delay. I The coherent integration time is 1 ms; C(·) is the PRN code generated locally by the receiver, s r It is the reflected signal received by the spaceborne GNSS-R, ψ C The code phase, f, of the locally generated copy code at the receiver r and φ r It refers to the frequency and phase of the carrier locally generated by the receiver. The detected complex waveform data sequence includes the in-phase and quadrature components of the reflected signal, providing amplitude and phase information of the reflected signal at a high sampling rate of 1 kHz. By selecting the number of coherent and incoherent iterations, track-resolution controllable detection can be performed.

[0065] In operation S2, the complex waveform data is preprocessed to obtain a complex waveform data sequence for detecting the physical state of the ocean surface; wherein, the complex waveform data sequence includes multiple complex waveforms, each of which corresponds to a sampling point on the ocean surface.

[0066] In operation S3, the sequence of detected complex waveform data is divided into multiple complex waveform groups according to the timing, and the coherence coefficient (MCC) between any two consecutive complex waveforms in each complex waveform group is determined; wherein, the number of complex waveforms in each complex waveform group ranges from 1 to 20, and each complex waveform corresponds to a time resolution of 1 ms.

[0067] In operation S4, based on all the coherence coefficients of each complex waveform group, the average coherence coefficient (MCC) feature of the complex waveform group with a 20ms time resolution is extracted to quantify the degree of phase change.

[0068] In operation S5, the complex waveform data sequence is used as the test set, and 20% of the data in the test set is selected as the training set. The threshold of the average coherence coefficient feature is determined based on the validation data and the training set using the histogram method.

[0069] In operation S6, the threshold of the average coherence coefficient characteristic is used as the boundary between sea ice and seawater. Based on the threshold and the average coherence coefficient of each complex waveform group in the test set, the surface physical state of the detection area is obtained by multiple sampling points corresponding to each complex waveform group in the test set.

[0070] The complex waveform data (CWF data) sequence according to the embodiment of the present application is obtained according to a TDS-1 CWF product, which is obtained by processing the received reflection signal according to TDS-1 and the local copy code by cross-correlation, and the sampling rate is 1 kHz. Since the number of complex waveforms included in each complex waveform group ranges from 1 to 20, the method in the embodiment of the present application is equivalent to being able to detect the ocean surface at a sampling rate of 1-20 ms. Therefore, the method of the embodiment of the present application has high resolution, which provides the possibility of detecting the ice-water state of small-scale areas.

[0071] Figure 2 A flowchart of a method for high-resolution detection of the physical state of the ocean surface based on spaceborne GNSS-R complex waveform data according to an embodiment of the present application is shown.

[0072] In combination Figures 1-2 According to the embodiment of the present application, the determination process of the complex waveform data can include a TDS-1 CWF product data preparation process and a data selection process. First, the TDS-1 CWF product data preparation process: the embodiment of the present application processes and describes all CWF tracks in a certain time period of the TDS-1 CWF product (for example, 659 CWF tracks from December 2014 to March 2, 2019), each track contains complex waveforms and related metadata of direct and reflected signals from the target satellite (i.e. a given GPS satellite or a given GNSS satellite). Each track contains about 130s of data sequence, with a sampling rate of 1 ms, so each track contains about 130000 complex waveforms. The complex waveform is composed of 500 time delays, with a resolution of 1 / 16 chips (about 18.3169 m). Second, the data selection process, before detection, first determine the spatio-temporal range for sea ice detection, extract the CWF data sequence and corresponding metadata within the detection range, and select the complex waveform data from all tracks obtained within the above time period according to the determined spatio-temporal range.

[0073] In operation S2, the determination method of the complex waveform data sequence includes the following steps.

[0074] Firstly, data filtering. Specifically, the land mask method is used to filter the complex waveform data to filter out the complex waveforms corresponding to the sampling points located on the land and within a preset distance (for example, 10 kilometers) from the land, to obtain filtered complex waveform data. The sea ice detection only focuses on the CWF data of the sea surface, and the land mask method is used to filter the initial complex waveform data sequence to filter out the complex waveforms corresponding to the sampling points located on the land and within a preset distance from the land. Secondly, the quality of the filtered complex waveform data is evaluated by using the signal-to-noise ratio, and the filtered complex waveform data with a signal-to-noise ratio greater than a preset threshold is retained to obtain a complex waveform data sequence for detecting the physical state of the sea surface. According to an embodiment of the present application, since the effective information of part of the CWF data is completely overwhelmed by noise, the signal-to-noise ratio (SNR) is selected as a quality control condition to evaluate the CWF quality, and the complex waveform data with a low quality is removed. Only the complex waveform data with an SNR value greater than a preset threshold (for example, 3 dB) is retained. Finally, a detection complex waveform data sequence is obtained, and the selected detection complex waveform data sequence is also called a test data set.

[0075] According to an embodiment of the present application, by preparing TDS-1 CWF product data, selecting data, and filtering data, the processing efficiency is improved, the influence of various noises is reduced, and the data quality and the reliability of the sea surface state detection result are ensured.

[0076] According to an embodiment of the present application, the physical state of the sea surface includes the sea ice state and the sea water state, and there is a significant difference in the phase change of the satellite signal reflected by the sea ice and the sea water. The embodiment of the present application determines the sea ice and sea water state by extracting the average coherence coefficient (i.e., the MCC characteristic quantity) and selecting a suitable MCC characteristic quantity threshold as the dividing line for judging the sea ice and sea water, and further realizes the determination of the physical state of the sea surface. The following will be described in detail.

[0077] According to another embodiment of the present application, the detection of the physical state of the sea surface is realized by extracting the average coherence coefficient (i.e., the MCC characteristic quantity). Specifically, in operation S3, the detection complex waveform data sequence is divided into a plurality of complex waveform groups in time sequence, the coherence coefficient between two continuous complex waveforms in each complex waveform group is determined, and for any two adjacent complex waveforms, for example, the sampling times corresponding to the mth complex waveform and the m+1th complex waveform are t m and t m+1 , the time interval between t m and t m+1 is t c . The coherence coefficient between the mth complex waveform and the m+1th complex waveform (i.e., the coherence coefficient between two continuous complex waveforms) is represented as follows:

[0078]

[0079] wherein, wherein, denotes the coherence coefficient of the two consecutive complex waveforms, t c denotes the time interval of the two consecutive complex waveforms, denotes the complex waveform value at the complex waveform peak power at time t m denotes the complex waveform value at the complex waveform peak power at time t denotes the complex waveform value at the complex waveform peak power at time t m denotes the complex waveform value at the complex waveform peak power at time t c denotes the complex waveform value at the complex waveform peak power at time t is the conjugate of .

[0080] According to the embodiments of the present application, each complex waveform group comprises M complex waveforms arranged in time sequence. Operation S4 specifically comprises averaging all coherence coefficients in the complex waveform group to obtain the average coherence coefficient feature quantity of the complex waveform group, which is represented as follows:

[0081]

[0082] wherein, wherein, denotes the average coherence coefficient feature quantity.

[0083] According to the embodiments of the present application, operation S5 specifically comprises the following operations. First, the training data set is grouped, and the average coherence coefficient feature quantity of each group is calculated. Second, the verification data that matches the space-time of each complex waveform group sample in the training set is obtained, and the verification data represents the actual physical state of the target region surface corresponding to the complex waveform group sample. Third, according to the actual physical state given by the verification data, a histogram of the average coherence coefficient feature quantity of the sea ice and the sea surface is drawn, and the intersection of the sea ice and sea water histograms is determined by using the histogram method, which is the average coherence coefficient feature quantity threshold.

[0084] According to the embodiments of the present application, the CWF data sequence obtained from the satellite signals reflected by sea ice and sea water has obvious phase change difference, and the MCC feature quantity quantifies the phase difference between adjacent complex waveforms. By comparing the MCC feature quantity with the MCC feature quantity threshold, the surface physical state of the detection region can be obtained. The MCC feature quantity threshold is determined by using the histogram method, and the feature quantity threshold is the intersection of the sea ice and sea water histograms. Essentially, the MCC feature quantity measures the degree of phase change of the complex waveform, and the MCC feature quantity obtained from sea ice is greater than the feature quantity threshold, and the MCC feature quantity obtained from sea water is less than the feature quantity threshold, which is expected to be higher than the MCC feature quantity obtained from sea water, i.e.

[0085]

[0086] In the above formula, CWF para represents the CWF characteristic quantity, para thres represents the CWF characteristic quantity threshold. For example, the MCC characteristic quantity corresponding to the sea ice state is greater than the MCC characteristic quantity, and the MCC characteristic quantity corresponding to the seawater state is less than the MCC characteristic quantity.

[0087] According to the embodiments of the present application, the embodiments of the present application also include selecting data capable of representing the state of the ocean surface, such as the OSISAF SIC-SH and NSIDC MASIE-NH sea ice product data sets, to evaluate the reliability and accuracy of the method provided by the embodiments of the present application. The embodiments of the present application establish the correspondence between the ice-water state of the verification data and the ice-water state determined by the MCC characteristic quantity, compare the sea ice detection result of the MCC characteristic quantity with the verification data, and evaluate the detection accuracy of the method provided by the embodiments of the present application for the ocean surface.

[0088] Figure 3 A flowchart of a method for detecting the physical state of the ocean surface with high resolution based on the complex waveform data of the spaceborne GNSS-R according to another embodiment of the present application is shown.

[0089] As Figure 3 shown, according to the embodiments of the present application, the above method further includes the following operations.

[0090] First, in the case where the surface physical state of the detection area determined by the plurality of positions corresponding to the complex waveform group of the test set is the sea ice state, the above method further includes obtaining the image of the detection area, that is, obtaining the MODIS composite image. Second, the TDS-1 CWF data needs to be preprocessed to obtain the CWF data track. According to the MODIS composite image, it is determined whether the CWF data track passes through the ice channel. Finally, the MCC characteristic quantity of the CWF data track point passing through the ice channel is extracted (that is, the CWFMCC characteristic quantity extraction), and the consistency (that is, the correlation determination) of the MCC characteristic quantity with the MODIS surface reflectivity data extracted from the MODIS MOD09GQ product is determined.

[0091] According to the embodiments of the present application, the surface of the ice channel is smoother than the sea ice, and the Fresnel reflection coefficient of water is much larger than that of ice. Therefore, the CWF characteristic quantity obtained at the ice channel is greater than the CWF characteristic quantity at the sea ice. According to the MODIS composite image, the CWF track passing through the ice channel in the spatiotemporally matched and cloud-free area is selected, and the MCC characteristic quantity of the CWF track is extracted. The surface reflectivity of the CWF track point extracted from the MODIS MOD09GQ product can be used to evaluate the ice channel identification capability of the CWF data.

[0092] For the convenience of those skilled in the art to understand and implement the present application, the present application will be further described in detail below in conjunction with the drawings and specific embodiments. It should be understood that the embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application.

[0093] All TDS-1CWF tracks in a certain period of time (659 tracks from December 2014 to March 2019) are selected for processing and illustration, each track contains complex waveforms and related metadata of direct and reflected signals from a given GPS satellite. The OSISAF SIC-SH and NSIDC MASIE-NH sea ice product datasets are selected as verification products to evaluate the reliability and accuracy of spaceborne GNSS-R technology for sea ice detection.

[0094] The present application is based on a high-resolution sea ice detection method based on spaceborne GNSS-Reflectometry (GNSS-R) complex waveform (CWF) products, which consists of four main parts: CWF product acquisition and preprocessing, CWF data high-resolution sea ice detection scheme construction, CWF data sea ice detection performance evaluation, and CWF data ice channel identification application. The following specific embodiments are listed and combined with Figure 3 The process of the method for detecting the physical state of the ocean surface based on the spaceborne GNSS-R complex waveform data provided by the embodiments of the present application is described in detail, and the present application comprises the following steps:

[0095] Step one, data acquisition and preparation. Download TDS-1CWF products (659 tracks from December 2014 to March 2019), and extract the required CWF data and related metadata from them, including time, GNSS satellite, GNSS-R satellite (or spaceborne GNSS-R), mirror reflection point position and speed, etc. Download OSISAF SIC-SH and NSIDC MASIE-NH sea ice product datasets.

[0096] Step two, TDS-1CWF data selection and filtering. Select the DS-1CWF data that matches the time and spatial position of the verification product; select SNR to evaluate CWF quality, and remove DS-1CWF data with SNR values less than the preset threshold; use land mask to filter CWF data collected within 10 kilometers from land. The data after filtering is the determined complex waveform data sequence for detecting the physical state of the ocean surface.

[0097] Step three, data training and test data set construction. Analyze the data to determine the complex waveform data sequence of the probe, wherein the southern hemisphere has a total of 29025 complex waveform data, and the northern hemisphere has a total of 343438 complex waveform data. Training and test data sets are established for the north and south hemispheres respectively, and the training data set accounts for 20% of the overall data, and the test data set is all the analysis data.

[0098] Step four, feature extraction. The complex waveform data sequence is divided into multiple complex waveform groups, the number of complex waveforms in each complex waveform group ranges from 20, and the coherence coefficient between two consecutive complex waveforms in each complex waveform group is calculated. According to all the coherence coefficients of each complex waveform group, the average coherence coefficient (MCC) feature quantity of 20ms time resolution quantifying the phase change degree is extracted, and the phase change degree of TDS-1CWF data is quantified.

[0099] Step five, sea ice and seawater state determination. The NSIDC MASIE-NH and OSISAF SIC-SH data are used as verification data, the MCC feature quantity threshold is determined by the training set and the verification data, and the MCC feature quantity threshold is used as the dividing line for determining the sea ice and seawater. The sea ice and seawater state of the test set is determined.

[0100] Step six, sea ice detection performance evaluation. Establish the spatio-temporal matching relationship between the ice-water state of the verification data and the feature quantity of each complex waveform group of the test set. The time matching standard is that the collection time of TDS-1CWF and the verification data is the same day, and the spatial position matching standard is that the distance between CWF SP point and verification data grid point is the shortest. The sea ice detection results of the test set are compared with the ice-water state of the corresponding verification data, and the sea ice detection performance of CWF is evaluated.

[0101] Step seven, ice channel identification application of CWF data. Determine the CWF trajectory through the ice channel by MODIS composite image. Taking the CWF trajectory obtained in a certain region in a certain period of time as an example, the sensitivity of CWF feature quantity to ice channel is qualitatively analyzed and observed, the 20ms MCC feature quantity of the CWF trajectory is extracted, and the surface reflectivity data of MODIS MOD09GQ band 1 at the corresponding position is compared. The consistency of the two is compared, and the identification ability of CWF feature quantity to ice channel is evaluated through further quantitative analysis.

[0102] Figure 4A The CWF trajectory of the ice channel is shown.

[0103] Figure 4B The CWF trajectory in Figure 4A The CWF trajectory in Figure 4A The MCC feature quantity obtained by the trajectory point at the CWF trajectory shown in

[0104] Figure 4C shows the MODIS surface reflectance and the MCC characteristic quantity in Figure 4B after the scale of the two is unified.

[0105] Figure 4D shows the scatter plot of the surface reflectance and the MCC characteristic quantity in Figure 4C .

[0106] Figure 4A is the CWF trajectory with MOD09GQ band 1 image as the base map. The MODIS surface reflectance at the ice channel and the MCC characteristic quantity obtained by the trajectory point at the ice channel are shown in Figure 4B . As shown in Figures 4A-4D , the MODIS surface reflectance at the ice channel has a strong correlation with the MCC characteristic quantity. Subtracting the average value of the MODIS surface reflectance at the ice channel and the MCC characteristic quantity at the ice channel respectively and normalizing to the scale of 0-1, the corresponding relationship between the surface reflectance data and the CWF characteristic quantity is established. It can be seen that the MCC characteristic quantity and the surface reflectance change trend are basically the same. The scatter plot is calculated, and the regression slope k of the MODIS surface reflectance and the MCC characteristic quantity is 0.9456, the correlation coefficient is 95.33%, and the deviation is 0.0045.

[0107] Figure 5A shows the local CWF trajectory in the sea ice area according to the embodiment of the present application. Figure 5B shows the MODIS surface reflectance and the MCC characteristic quantity corresponding to the local CWF trajectory in Figure 5A . Figure 5A shows the MCC characteristic quantity obtained by the trajectory point at the local CWF trajectory in Figure 5A . Among them,

[0108] Figure 5C shows the corresponding relationship between the MODIS surface reflectance and the MCC characteristic quantity in Figure 5B after the scale of the two is unified. Figure 5D shows the scatter plot of the surface reflectance and the MCC characteristic quantity in Figure 5C .

[0109] As shown in Figures 5C-5D , if only the sea ice area is considered, the regression slope k of the MODIS surface reflectance and the MCC characteristic quantity is 0.7765, the correlation coefficient is 78.42%, and the deviation is 0.0045.

[0110] As a second aspect of the present application, a device for detecting the physical state of the sea surface is also provided, which is used to realize the above method.

[0111] Figure 6 Fig. 1 shows a schematic diagram of a device for detecting the physical state of the sea surface according to an embodiment of the present application.

[0112] As shown in Figure 6 the device comprises an acquisition module 601, a preprocessing module 602, a grouping module 603, a first determination module 604, a second determination module 605, and a detection module 606.

[0113] The acquisition module 601 is configured to acquire complex waveform data, which is obtained by cross-correlation processing of a reflected signal received by a space-borne GNSS-R and a local copy code, and includes amplitude and phase information of the reflected signal acquired at a high sampling rate of 1 kHz. The coherent and incoherent times can be selected by the user, and the along-track resolution of the detection can be controlled. The reflected signal is a satellite signal emitted by a target satellite and reflected by the earth's surface. The preprocessing module 602 is configured to preprocess the complex waveform data to obtain a complex waveform data sequence for detecting the physical state of the sea surface. The complex waveform data sequence includes a plurality of complex waveforms, and each complex waveform corresponds to a sampling point on the sea surface. The grouping module 603 is configured to group the complex waveform data sequence into a plurality of complex waveform groups in time sequence, and determine a coherence coefficient between any two consecutive complex waveforms in each complex waveform group. The number of complex waveforms in each complex waveform group ranges from 1 to 20, and each complex waveform corresponds to a time resolution of 1 ms. The first determination module 604 is configured to extract an average coherence coefficient feature quantity of a complex waveform group with a time resolution of 20 ms, which quantifies the degree of phase change, according to all coherence coefficients of each complex waveform group. The second determination module 605 is configured to use the complex waveform data sequence as a test set, select 20% of the data in the test set as a training set, and determine a threshold value of the average coherence coefficient feature quantity according to the validation data and the training set using a histogram method. The detection module 606 is configured to use the threshold value of the average coherence coefficient feature quantity as a judgment line for the sea ice-seawater boundary, and is configured to obtain the surface physical state of a detection area determined by a plurality of sampling points corresponding to each complex waveform group in the test set according to the threshold value and the average coherence coefficient of each complex waveform group in the test set.

[0114] The high-resolution sea ice detection method based on the complex waveform (CWF) product of the spaceborne GNSS reflectometry (GNSS-R) technology according to the application provides a new data source for high-resolution detection of sea ice by using the CWF data of the TDS-1 satellite without increasing additional cost, and supplements the deficiency of the existing sea ice detection data; the correlation coefficient characteristic quantity extracted from the 1ms CWF data is averaged for 20 times to obtain the MCC characteristic quantity, the sea ice can be detected at a sampling rate of 20ms, and the along-track resolution is increased by 50 times compared with the DDM data with a sampling rate of 1s; in addition, the high-resolution sea ice detection capability of the CWF data provides a possibility for detecting the ice-water state in a small-scale area, and can be used for identifying narrow ice-water channels.

[0115] The embodiments of the application are described above. However, these embodiments are only for the purpose of illustration, and are not intended to limit the scope of the application. Although each embodiment is described above, this does not mean that the measures in each embodiment cannot be used advantageously in combination. The scope of the application is defined by the appended claims and their equivalents. Without departing from the scope of the application, those skilled in the art can make various substitutions and modifications, which should fall within the scope of the application.

Claims

1. A method for high-resolution detection of the physical state of the ocean surface based on spaceborne GNSS-R complex waveform data, the method comprising: Complex waveform data is acquired by cross-correlation processing of the reflected signal received by the spaceborne GNSS-R and the local copy code. The complex waveform data includes the amplitude and phase information of the reflected signal obtained at a high sampling rate of 1 kHz. By selecting the number of coherent and incoherent steps, the detection along the orbit with controllable resolution is performed. The reflected signal is the satellite signal emitted by the target satellite reflected from the Earth's surface. The complex waveform data is preprocessed to obtain a complex waveform data sequence for detecting the physical state of the ocean surface; wherein the complex waveform data sequence includes multiple complex waveforms, each complex waveform corresponding to a sampling point on the ocean surface; The detection complex waveform data sequence is divided into multiple complex waveform groups according to the time sequence, and the coherence coefficient between any two consecutive complex waveforms in each complex waveform group is determined; wherein, the number of complex waveforms in each complex waveform group ranges from 1 to 20, and each complex waveform corresponds to a time resolution of 1 ms; Based on all the coherence coefficients of each complex waveform group, the average coherence coefficient feature of the complex waveform group with a 20 ms time resolution is extracted to quantify the degree of phase change. The complex waveform data sequence is used as the test set, and 20% of the data in the test set is selected as the training set. The threshold of the average coherence coefficient feature is determined based on the validation data and the training set using the histogram method. The threshold of the average coherence coefficient characteristic is used as the boundary between sea ice and seawater. Based on the threshold and the average coherence coefficient of each complex waveform group in the test set, the surface physical state of the detection area is obtained by multiple sampling points corresponding to each complex waveform group in the test set.

2. The method according to claim 1, wherein, Preprocessing methods for complex waveform data include: The complex waveform data is filtered using a land masking method to remove complex waveforms corresponding to sampling points located on land or within a preset distance from land, thus obtaining filtered complex waveform data. The signal-to-noise ratio (SNR) is used to evaluate the quality of filtered complex waveform data. Filtered complex waveform data with an SNR greater than a preset threshold are retained to obtain a complex waveform data sequence for detecting the physical state of the ocean surface.

3. The method according to claim 1, wherein, The coherence coefficient between two consecutive complex waveforms is expressed as follows: ， in, This represents the coherence coefficient of the two consecutive complex waveforms. It represents the time interval between two consecutive complex waveforms. express The complex waveform value corresponding to the peak power of the complex waveform at time t. express The complex waveform value corresponding to the peak power of the complex waveform at time t. for . conjugate.

4. The method according to claim 3, wherein, Each complex waveform group consists of M complex waveforms arranged in time sequence; Based on all coherence coefficients of each complex waveform group, the average coherence coefficient feature of the complex waveform group with a 20 ms time resolution, quantizing the degree of phase change, is extracted, including: The average coherence coefficient characteristic of the complex waveform group is obtained by averaging all coherence coefficients within the group. This average coherence coefficient characteristic is expressed as follows: ， in, This represents the characteristic quantity of the average coherence coefficient.

5. The method according to claim 4, wherein, The surface physical state includes sea ice state and seawater state; Based on the threshold and the average coherence coefficient of each complex waveform group in the test set, the physical state of the probe area surface determined by multiple sampling points corresponding to each complex waveform group in the test set is obtained, including: In each complex waveform group in the test set, the average coherence coefficient feature is compared with the threshold of the average coherence coefficient feature. If the average coherence coefficient feature is greater than the threshold of the average coherence coefficient feature, the physical state of the detection area determined by the M sampling points is the sea ice state; if the average coherence coefficient feature is less than the threshold of the average coherence coefficient feature, the physical state of the surface of the detection area determined by the M locations is the seawater state.

6. The method according to claim 3, wherein, Selecting 20% ​​of the test set as the training set, and using the histogram method, determine the threshold for the average coherence coefficient features based on the validation data and the training set, including: The training dataset is divided into groups to obtain multiple complex waveform groups, and the average coherence coefficient feature of each group is calculated. Obtain validation data that is spatiotemporally matched with each complex waveform group sample in the training set. The validation data characterizes the actual physical state of the target region surface corresponding to the complex waveform group sample. Based on the actual physical state given by the verification data, a histogram of the average coherence coefficient characteristic of sea ice and seawater surface is plotted. The intersection point of the sea ice and seawater histograms is determined by the histogram method, and the intersection point is the threshold of the average coherence coefficient characteristic.

7. The method according to claim 2, wherein, The preset threshold is 3dB, and the preset distance is 10 kilometers.

8. A device for high-resolution detection of the physical state of the ocean surface based on spaceborne GNSS-R complex waveform data, used to implement the method as described in any one of claims 1 to 7, the device comprising: The acquisition module acquires complex waveform data, which is obtained by cross-correlation processing of the reflected signal received by the spaceborne GNSS-R and the local copy code. The complex waveform data includes the amplitude and phase information of the reflected signal acquired at a high sampling rate of 1 kHz. By selecting the number of coherent and incoherent steps, the detection with controllable resolution along the orbit can be performed. The reflected signal is a satellite signal emitted by the target satellite reflected from the Earth's surface. The preprocessing module is suitable for preprocessing the complex waveform data to obtain a complex waveform data sequence for detecting the physical state of the ocean surface; wherein the complex waveform data sequence includes multiple complex waveforms, each complex waveform corresponding to a sampling point on the ocean surface; The grouping module is suitable for dividing a sequence of probed complex waveform data into multiple complex waveform groups according to time sequence, and determining the coherence coefficient between any two consecutive complex waveforms in each complex waveform group. The number of complex waveforms in each group ranges from 1 to 20, and each complex waveform corresponds to a time resolution of 1 ms. The first determining module is suitable for extracting the average coherence coefficient feature of the complex waveform group with a 20 ms time resolution, based on all coherence coefficients of each complex waveform group; The second determining module is suitable for using the complex waveform data sequence as a test set, selecting 20% ​​of the data in the test set as a training set, and using the histogram method to determine the threshold of the average coherence coefficient feature based on the verification data and the training set. The detection module is suitable for using the threshold of the average coherence coefficient characteristic quantity as the boundary between sea ice and seawater, and is also suitable for obtaining the surface physical state of the detection area determined by multiple sampling points corresponding to each complex waveform group in the test set based on the threshold and the average coherence coefficient of each complex waveform group in the test set.

Images