GNSS water level inversion model

Abstract

The invention discloses a GNSS (Global Navigation Satellite System) water level inversion model, which belongs to the technical field of water disaster prevention and marine environment monitoring, and comprises the following steps: S1, acquiring original data: an SNR (Signal to Noise Ratio) observation value and a GNSS observation value; s2, setting parameters: determining cut-off elevation angle information and azimuth angle information according to the surrounding environment of the shore-based GNSS monitoring station; s3, direct reflection signal separation: selecting IMF components containing main water level information to perform water level inversion and addition output; and S4, solving h by utilizing spectral analysis: analyzing the IMF component segment containing the water level information by utilizing a Lomb-Scargle algorithm to solve a main frequency f, and converting according to a formula to obtain a required vertical reflection distance h. The GNSS water level inversion model has the advantages of global coverage, no need of manual intervention, economy, effectiveness and simplicity and convenience in maintenance, and is suitable for monitoring tides, rivers, lakes and sea level; the accuracy of water level inversion can be improved, and key water level information is prevented from being lost in traditional denoising.

Description

Technical Field

[0001] The present invention relates to the technical field of water disaster prevention and control and marine environment monitoring, and in particular to a GNSS water level inversion model. Background Art

[0002] Sea level rise poses a significant threat to people and property in coastal areas. Therefore, it is crucial to accurately and effectively monitor sea level changes. The conventional methods used to measure coastal sea level height (SLH) are tide gauges and satellite altimetry. However, measurements obtained from tide gauges are affected by vertical land motion (VLM), which makes it challenging to directly calculate absolute sea level changes.

[0003] At present, the traditional method of measuring sea surface height is mainly to build tide gauge stations for measurement. Tide gauge stations are observation stations with tide gauges set up at selected coasts for long-term observation of ocean tide changes, and they are guaranteed to be able to conduct joint measurements with the geoid network. Usually, multiple tide gauge stations will be built in a sea area for network observation, but as fixed stations, this measurement method has poor flexibility and high construction costs. The observation range of a single station is limited, resulting in low spatial resolution of the observation data.

[0004] With the development of technology, satellite-borne radar altimeters can measure sea surface heights worldwide, making up for the shortcomings of conventional measurement methods that cannot measure distant sea areas and further improving the spatial resolution of sea surface observations. However, because it requires satellites as a platform, the observation cost of this method is relatively high, and it is affected by the number of satellites, satellite operating cycles, and orbital inclination. The data time resolution of this method is low and it is difficult to apply to small and medium-scale measurements. In addition, buoys are also used to monitor sea level changes and sea surface-related parameters, but this method has high deployment and maintenance costs, so its distribution points are relatively small. For inland lakes, rivers, and artificial reservoirs, the traditional method of obtaining water level information is mainly water level gauges and ultrasonic water level meters. The combination of the two to monitor water level information enables conventional waters to achieve automated long-term measurement and data recording, but this method is easily affected by the surrounding environment and extreme weather, and cannot guarantee the reliability and accuracy of data all day and all weather.

[0005] In recent years, with the rapid development of GNSS global navigation satellite system, the continuous enrichment of navigation satellite resources and the continuous improvement of the breadth and depth of application, some scholars have proposed to give full play to the all-weather, all-day, high-precision and low-cost characteristics of GNSS, and use navigation and positioning satellites, reflecting surfaces and receivers to build shore-based GNSS bistatic radar. By extracting and processing the reflected signals, the Global Navigation Satellite System Reflection Signal Remote Sensing Technology (Global Navigative Satellite System-Reflection, GNSS-R) that can realize the measurement of various reflecting surfaces such as water surface, soil, snow, etc. has gradually attracted the attention of researchers at home and abroad. On the one hand, after decades of development, the GNSS global navigation and positioning system currently includes multiple satellite navigation systems such as GPS, BDS, GLONASS, Galileo, hundreds of navigation and positioning satellites, and dozens of signal frequency bands. The research data of GNSS-R technology has also gradually shifted from the GPS single system to availability research based on multiple systems and multiple frequencies such as BDS and Galileo. On the other hand, in order to monitor the vertical movement of water level measurement sites or the deformation of water storage structures such as dams, GNSS observation systems are gradually being installed near sites and water areas such as reservoir dams. If the existing GNSS measurement stations are fully utilized, the utilization rate of GNSS data can be further improved without increasing the investment cost, and its application field can be expanded. At the same time, it can effectively make up for the low spatiotemporal resolution of existing water level measurement methods, which is of great significance for water level monitoring of dams and lakes.

[0006] Satellites in orbit of the Global Navigation Satellite System continuously transmit radio signals to the surface of the Earth. These signals can be used not only for navigation, positioning and timing, but also for ground-based remote sensing, known as the GNSS reflectometry (GNSS-R) method. Since Martin-Neira first proposed the feasibility of using GPS reflected signals for ocean height measurement, GNSS-R has received increasing attention in the field of GNSS and remote sensing. It has been confirmed and verified for measuring sea level, soil moisture and other fields. It is well known that monitoring and understanding the changes in sea level are of great significance to ecological and environmental protection and economic construction. Moreover, some studies have confirmed that GNSS-R can be used to monitor sea level from shore-based, airborne and satellite-based platforms. Therefore, GNSS-R technology, with its advantages of all-weather, rich signal sources and high temporal and spatial resolution, will become one of the important means of future marine surveys and monitoring. As a type of GNSS-R measurement, the interferogram technique (IPT) is proposed to observe water level. The extremely simple IPT, also known as GNSS interferometric reflectometry (GNSS-IR), is superior to the dual-antenna GNSS-R technology. GNSS-IR requires only a single antenna to receive a mixed signal of the direct signal and the reflected signal caused by the multipath effect; the vertical height between the antenna and the reflecting surface is then derived from the mixed signal. To explain further, the synthetic signal causes the observed signal-to-noise ratio (SNR) data to oscillate, and its frequency can be used to invert the height. The key to GNSS infrared height measurement is to extract the oscillation frequency from the SNR data. When extracting the SNR oscillation frequency, the sine value of the satellite elevation angle is the independent variable, resulting in uneven sampling of the SNR data. In 2013, Larson et al. proposed using the Lomb-Scargle periodogram (LSP) method to extract the SNR oscillation frequency and verify the efficiency of the method. They collected data from the GTGU station of the Onsala Space Observatory in Sweden and the SC02 station of the Friday Harbor Plate Boundary Observatory in the United States, with inversion accuracies of 5cm and 10cm, respectively. In 2016, Strandberg et al. proposed inverse modeling of SNR data to extract the oscillation frequency, and used data from the GTGU station and SPBY station in Tasmania, Australia to provide accuracies of 1.4cm and 3.1cm, respectively.

[0007] From the perspective of signal processing, many researchers have adopted various signal processing methods to improve the accuracy of GNSS-R retrieval. In order to obtain interferometric measurements of GNSS reflected signals under SNR, the direct signal component needs to be removed first, which can be done by fitting and subsequently subtracting a low-order polynomial. However, in addition to the direct and reflected signal components, SNR data also contains noise, and removing low-order polynomials may also inadvertently filter out the target signal. Larson et al. (2013a) studied the interference pattern of the reflected signal displayed in the GNSS direct signal and signal-to-noise ratio (SNR) data, and derived the vertical distance from the reflecting surface to the antenna phase center, also known as the reflector height. They proposed a new GNSS-R tide gauge system equipped with only RHCP receivers. The oscillations observed in the SNR data are the result of interference between the direct signal and the reflected signal. By applying a low-order polynomial fit to the SNR data, the direct signal can be isolated, leaving only the reflected signal, that is, the detrended SNR data. Then, the receiver height can be inferred from the relationship between the dominant frequency of the detrended SNR data and the specific wavelength of the signal. With the continuous development of the application of GNSS-R in SLH monitoring, more and more studies are devoted to improving the accuracy and stability of SLH inversion through error correction and various signal processing strategies. Larson et al. proposed a method to consider the dynamic correction of the sea surface, that is, the change of SLH within the observation interval, which is related to the rate of change of SLH and the satellite position. The oscillation in the SNR data is considered to be the contribution of the reflected signal. Some studies are committed to improving the quality and accuracy of oscillation extraction in SNR data. Empirical Mode Decomposition (EMD) is used to extract smoother and less noisy reflected signals from the original SNR data. Song Minfeng et al. (2021) applied GNSS-IR technology to achieve high-precision monitoring of reservoir water level changes, and obtained results with an error of no more than 2 cm compared with the measured water level value. GNSS remote sensing technology has broad development prospects in my country. It can not only help my country's construction of a strong marine country, but also be efficiently applied to watershed ecological protection and high-quality development. Summary of the invention

[0008] The purpose of the present invention is to provide a GNSS water level inversion model, which has the advantages of global coverage, no need for human intervention, cost-effectiveness and easy maintenance, and is suitable for tidal, river, lake and sea level monitoring; it can improve the accuracy of water level inversion and avoid the loss of key water level information by traditional denoising.

[0009] To achieve the above object, the present invention provides a GNSS water level inversion model, including the following process:

[0010] S1. Get the original data: extract the SNR observation value received by the receiver: use a single antenna receiver to receive GNSS observation values. The data received by the satellite is of NMEA type. These original NMEA data include azimuth and elevation.

[0011] S2. Setting parameters: Determine the cut-off elevation angle information and azimuth angle information according to the surrounding environment of the shore-based GNSS monitoring station; the elevation angle is selected based on the analysis of the dSNR value, and the azimuth angle needs to be selected based on the Fresnel reflection area of ​​the satellite;

[0012] S3, separation of direct reflection signals: using the improved EMD detrending method to decompose the SNR data obtained from the calm water surface to obtain different IMF components, selecting the IMF components containing water level information for water level inversion, removing the IMF component segments containing interference, and adding and outputting multiple IMF component segments containing multiple water level information;

[0013] S4. Solve h using spectrum analysis: Use the Lomb-Scargle algorithm to analyze the IMF component segments containing water level information, solve the main frequency f, and convert it according to the formula to obtain the required vertical reflection distance h, so as to map the change of water level.

[0014] Preferably, in S3, the method of using SNR data to invert the water surface height utilizes the characteristics of the signal-to-noise ratio. Where P signal is the signal power, P noise is the noise power;

[0015] It can be seen that noise power and signal strength power affect SNR. Assume that the direct signal is parallel to the partial reflected signal that has not reached the water surface. Assume that the height of the GNSS receiver antenna from the water surface is h, the elevation angle of the satellite is θ, and R' represents the mirror point of the receiver on the water surface. According to the spatial model, the spatial relationship between the height h of the receiver antenna phase center from the water surface and the path difference Δd between the direct signal and the reflected signal is expressed as: Δd = 2hsinθ.

[0016] Preferably, during the propagation of GNSS signals, part of the signal will be reflected on the surface of the earth, forming a reflection signal of a coherent component and an irregular scattered component; since the coherent component has the same frequency as the direct signal, interference will occur when a single receiver is used to receive data; the power of the coherent component is related to the roughness of the surface of the reflecting medium: the smoother the surface, the higher the coherent component power obtained; when the receiver is placed by the river or the sea, the signal reflected by the water surface interferes with the direct signal at the receiver antenna to form a composite signal S L (n), where L refers to the composite signal of the Lth satellite and n represents the time series index of the pseudo-random code.

[0017] Preferably, the S L (n) is expressed as:

[0018]

[0019] in:

[0020]

[0021]

[0022] In the above formula, A L represents the amplitude of the composite signal received by the receiver. represents the amplitude of the direct signal, represents the amplitude of the reflected signal; D represents the navigation message, and ω(n) represents the noise; Represents the pseudo-random code of the Lth satellite; represents the phase information of the reflected signal from the Lth satellite, T c represents the period information of the pseudo-random code, n represents the time series index of the pseudo-random code, θ′ represents the comprehensive phase, represents the carrier phase of the direct signal, represents the phase angle of the composite signal received by the receiver, f L Represents the main frequency of the Lth satellite.

[0023] Preferably, when λ represents the wavelength of the carrier, according to the spatial relationship, the additional phase of the reflected signal from the Lth satellite relative to the direct signal is:

[0024]

[0025] In the sea surface height inversion model based on signal-to-noise ratio data, the amplitude of the interference signal is used to extract important information to invert the water surface and sea surface height; the relationship between the amplitude and the signal power is:

[0026]

[0027] Among them, it can be seen It has a great impact on the whole formula. The required signal type can be extracted by simply removing the trend from the signal-to-noise ratio data and performing fitting errors.

[0028] Preferably, the SNR of the interference signal is r The formula is:

[0029]

[0030] make t = sinθ; the detrended dSNR is expressed as:

[0031]

[0032] Among them, f contains the height information h, so the Lomb-Scargle Fourier transform method is used to analyze the dSNR sequence to obtain

[0033] Therefore, the present invention adopts the above-mentioned GNSS water level inversion model, which has the advantages of global coverage, no need for human intervention, cost-effectiveness and easy maintenance, and is suitable for tidal, river, lake and sea level monitoring; it can improve the accuracy of water level inversion and avoid the loss of key water level information by traditional denoising.

[0034] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a feedback flow chart of an embodiment of a GNSS water level inversion model of the present invention;

[0036] Figure 2 It is a schematic diagram of a satellite Fresnel reflection area of ​​an embodiment of a GNSS water level inversion model of the present invention;

[0037] Figure 3 It is a conventional algorithm detrended calm water surface inversion height result diagram of an embodiment of a GNSS water level inversion model of the present invention;

[0038] Figure 4 This is an improved EMD detrended calm water surface inversion height result diagram of an embodiment of a GNSS water level inversion model of the present invention;

[0039] Figure 5 It is a schematic diagram of water surface height measurement in an embodiment of a GNSS water level inversion model of the present invention. DETAILED DESCRIPTION

[0040] The technical solution of the present invention is further described below through the accompanying drawings and embodiments.

[0041] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0042] Embodiment 1

[0043] like Figure 1 As shown, the present invention provides a GNSS water level inversion model, including the following process:

[0044] S1. Get the original data: extract the SNR observation value received by the receiver: use a single antenna receiver to receive GNSS observation values. The data received by the satellite is of NMEA type. These original NMEA data include azimuth and elevation.

[0045] S2. Set parameters: Determine the cut-off elevation angle information and azimuth angle information according to the surrounding environment of the shore-based GNSS monitoring station (i.e., the direction and distance to the water body); the elevation angle can be selected based on the analysis of the dSNR value, such as Figure 2 As shown, the azimuth needs to be selected according to the Fresnel reflection zone of the satellite;

[0046] S3, direct reflection signal separation: Figure 3 to Figure 4 As shown, the traditional water surface height inversion is carried out when the water surface fluctuation is relatively large, because the multipath effect will be better. In such an environment, the quadratic polynomial is generally used to remove the trend, but the inversion accuracy obtained by using the improved EMD detrending method for calm water surface is higher. Especially for calm water surface, the SNR data is generally very stable, which is not conducive to the inversion height. If denoising is performed in advance, more water level data information will be removed. Therefore, the improved EMD detrending method can avoid this situation. At the same time, according to the formula for SNR, it can be seen that filtering can be reduced at the antenna port to enhance the noise.

[0047] Therefore, the SNR data obtained from the calm water surface are decomposed using the improved EMD detrending method to obtain different IMF components. The IMF components containing the main water level information are selected for water level inversion, and the IMF component segments containing more interference are removed. Multiple IMF component segments containing multiple water level information are added and output to improve the inversion accuracy for the calm water surface.

[0048] S4. Solve h using spectrum analysis: The Lomb-Scargle algorithm can be used to analyze the IMF component segments containing water level information, solve the main frequency f, and convert it according to the formula to obtain the required vertical reflection distance h, so as to map the change of water level.

[0049] Principle: The SNR-based GNSS inversion of water surface height mainly uses the information of signal-to-noise ratio for inversion. When the satellite hits the water surface, interference will be formed, which will have a certain impact on the quality of the received satellite signal-to-noise ratio. However, in order not to affect the positioning accuracy and speed, many manufacturers will add filtering to the antenna when producing GNSS antennas to filter out noise and improve the signal-to-noise ratio. As for the inversion of water surface height, the higher the signal-to-noise ratio, the better or the lower the signal-to-noise ratio, the better. The GNSS antennas produced by manufacturers to improve the signal-to-noise ratio are not suitable for tide level or water surface height inversion. The inversion of water surface height requires a moderate noise and signal ratio to invert the tide level in order to achieve better results. The method of using SNR data to invert water surface height utilizes the characteristics of the signal-to-noise ratio. The signal-to-noise ratio formula is as follows: Where P signal is the signal power, P noise is the noise power;

[0050] It can be seen that noise power and signal strength power affect SNR, and the requirement for the signal-to-noise ratio for water surface height inversion based on SNR is that the fluctuation is large and the change law and multipath effect can be better captured. Since the distance between the GNSS satellite and the water surface is much greater than the distance between the GNSS receiver and the water surface, it can be assumed that the direct signal and the partial reflected signal that does not reach the water surface are parallel.

[0051] like Figure 5 As shown, assuming that the height of the GNSS receiver antenna from the water surface is h, the elevation angle of the satellite is θ, and R' represents the mirror point of the receiver on the water surface; according to the spatial model, the spatial relationship between the height h of the receiver antenna phase center from the water surface and the path difference Δd between the direct signal and the reflected signal is expressed as: Δd = 2hsinθ.

[0052] During the propagation of GNSS signals, part of the signal will be reflected on the earth's surface, forming a reflection signal of coherent components and irregular scattered components; since the coherent component has the same frequency as the direct signal, interference will occur when a single receiver is used to receive data. In addition, the power of the coherent component is related to the roughness of the surface of the reflecting medium: the smoother the surface, the higher the coherent component power obtained; when the receiver is placed by the river or the sea, the signal reflected by the water surface interferes with the direct signal at the receiver antenna to form a composite signal S L (n), where L refers to the composite signal of the Lth satellite and n represents the time series index of the pseudo-random code.

[0053] S L (n) is expressed as:

[0054]

[0055] in:

[0056]

[0057] In the above formula, A L represents the amplitude of the composite signal received by the receiver. represents the amplitude of the direct signal, represents the amplitude of the reflected signal; D represents the navigation message, and ω(n) represents the noise; Represents the pseudo-random code of the Lth satellite; represents the phase information of the reflected signal from the Lth satellite, T c represents the period information of the pseudo-random code, n represents the time series index of the pseudo-random code, θ′ represents the comprehensive phase, represents the carrier phase of the direct signal, represents the phase angle of the composite signal received by the receiver, f L Represents the main frequency of the Lth satellite.

[0058] When λ represents the wavelength of the carrier, according to the spatial relationship, the additional phase of the reflected signal from the Lth satellite relative to the direct signal is:

[0059]

[0060] In the sea surface height inversion model based on signal-to-noise ratio data, the amplitude of the interference signal is mainly used to extract important information for the inversion of water surface and sea surface height; and the relationship between the amplitude and the signal power is:

[0061]

[0062] Among them, it can be seen It has a huge impact on the entire formula. The required signal type can be extracted by simply removing the trend from the signal-to-noise ratio data and performing fitting errors.

[0063] Through the above steps, the SNR including the interference signal can be obtained r The formula is:

[0064]

[0065] make When t = sinθ, the detrended dSNR is expressed as:

[0066]

[0067] Among them, f contains the height information h, so the Lomb-Scargle Fourier transform method is used to analyze the dSNR sequence to obtain

[0068] Therefore, the present invention adopts the above-mentioned GNSS water level inversion model, which has the advantages of global coverage, no need for human intervention, cost-effectiveness and easy maintenance, and is suitable for tidal, river, lake and sea level monitoring; it can improve the accuracy of water level inversion and avoid the loss of key water level information by traditional denoising.

[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A GNSS water level inversion model, characterized in that: The following processes are included: S1. Get the original data: extract the SNR observation value received by the receiver: use a single antenna receiver to receive GNSS observation values. The data received by the satellite is of NMEA type. These original NMEA data include azimuth and elevation. S2. Setting parameters: Determine the cut-off elevation angle information and azimuth angle information according to the surrounding environment of the shore-based GNSS monitoring station; the elevation angle is selected based on the analysis of the dSNR value, and the azimuth angle needs to be selected based on the Fresnel reflection area of ​​the satellite; S3, separation of direct reflection signals: using the improved EMD detrending method to decompose the SNR data obtained from the calm water surface to obtain different IMF components, selecting the IMF components containing water level information for water level inversion, removing the IMF component segments containing interference, and adding and outputting multiple IMF component segments containing multiple water level information; S4. Solve h using spectrum analysis: Use the Lomb-Scargle algorithm to analyze the IMF component segments containing water level information, solve the main frequency f, and convert it according to the formula to obtain the required vertical reflection distance h, so as to map the change of water level.

2. A GNSS water level inversion model according to claim 1, characterized in that: In S3, the method of using SNR data to invert the water surface height utilizes the characteristics of the signal-to-noise ratio. Where P signal is the signal power, P noise is the noise power; It can be seen that noise power and signal strength power affect SNR. Assume that the direct signal is parallel to the partial reflected signal that has not reached the water surface. Assume that the height of the GNSS receiver antenna from the water surface is h, the elevation angle of the satellite is θ, and R' represents the mirror point of the receiver on the water surface. According to the spatial model, the spatial relationship between the height h of the receiver antenna phase center from the water surface and the path difference Δd between the direct signal and the reflected signal is expressed as: Δd = 2hsinθ.

3. A GNSS water level inversion model according to claim 1, characterized in that: During the propagation of GNSS signals, part of the signal will be reflected on the surface of the earth, forming a reflection signal of coherent components and irregular scattered components. Since the coherent component has the same frequency as the direct signal, interference will occur when a single receiver is used to receive data. The power of the coherent component is related to the roughness of the surface of the reflecting medium: the smoother the surface, the higher the coherent component power obtained. When the receiver is placed by the river or the sea, the signal reflected by the water surface interferes with the direct signal at the receiver antenna to form a composite signal S L (n), where L refers to the composite signal of the Lth satellite and n represents the time series index of the pseudo-random code.

4. A GNSS water level inversion model according to claim 3, characterized in that: The S L (n) is expressed as: in: In the above formula, A L represents the amplitude of the composite signal received by the receiver. represents the amplitude of the direct signal, represents the amplitude of the reflected signal; D represents the navigation message, and ω(n) represents the noise; Represents the pseudo-random code of the Lth satellite; represents the phase information of the reflected signal from the Lth satellite, T c represents the period information of the pseudo-random code, n represents the time series index of the pseudo-random code, θ' represents the comprehensive phase, represents the carrier phase of the direct signal, represents the phase angle of the composite signal received by the receiver, f L Represents the main frequency of the Lth satellite.

5. A GNSS water level inversion model according to claim 4, characterized in that: When λ represents the wavelength of the carrier, according to the spatial relationship, the additional phase of the reflected signal from the Lth satellite relative to the direct signal is: In the sea surface height inversion model based on signal-to-noise ratio data, the amplitude of the interference signal is used to extract important information to invert the water surface and sea surface height; the relationship between the amplitude and the signal power is: Among them, it can be seen It has a great impact on the whole formula. The required signal type can be extracted by simply removing the trend from the signal-to-noise ratio data and performing fitting errors.

6. A GNSS water level inversion model according to claim 5, characterized in that: The SNR of the interference signal r The formula is: make The detrended dSNR is expressed as: Among them, f contains the height information h, so the Lomb-Scargle Fourier transform method is used to analyze the dSNR sequence to obtain

Images