Assimilation method for improving temperature and salt analysis precision based on submarine pressure data

Abstract

An assimilation method for improving thermohaline analysis precision based on seabed pressure data comprises the steps that various observation data needed for improving thermohaline analysis precision assimilation are obtained, the observation data including temperature, salinity, sea surface temperature, sea surface salinity and seabed pressure are in an increment form relative to a background field, and sea surface height data are sea surface height anomalies; constructing an objective function of a space-time four-dimensional multi-scale data assimilation framework for improving temperature and salt analysis precision assimilation; setting dynamic constraint conditions of a space-time four-dimensional multi-scale data assimilation framework for improving thermohaline analysis precision assimilation; and inputting prepared observation data, carrying out joint data assimilation, and reconstructing a thermohaline analysis field. According to the method, the vertical structural precision of the temperature and salt field can be enhanced, the data quality and reliability are improved, and the overall precision of ocean state analysis is improved, so that the prediction precision of ocean disasters is improved, and the prediction and early warning of the ocean disasters are improved.

Description

Technical Field

[0001] The present invention relates to an assimilation method for improving the accuracy of temperature and salinity analysis. In particular, it relates to an assimilation method for improving the accuracy of temperature and salinity analysis based on seafloor pressure data. Background Art

[0002] Satellite observation data has the characteristics of global coverage, high timeliness, data richness, and multi-variable nature. By combining satellite remote sensing data with other observation data, the accuracy of ocean numerical models can be significantly improved. Currently, satellite remote sensing technology has entered a new stage of high precision and diversification, but most of the satellite observation data used in ocean multi-variable assimilation schemes only involve sea surface temperature, sea surface salinity, sea surface height, etc.

[0003] The GRACE (Gravity Recovery and Climate Experiment) satellite is a pair of satellites operating in low-earth orbit, mainly used for accurately measuring changes in the Earth's gravitational field. By analyzing GRACE satellite data, high-resolution changes in the gravitational field of various parts of the Earth can be obtained, and then the mass changes of water bodies and corresponding ocean thermal expansion, tidal fluctuations, etc. can be inferred. The gravity change data provided by the GRACE satellite provides clues to the mass changes at the ocean bottom, and the changes in seafloor pressure are closely related to the mass of ocean water bodies. Ocean mass changes usually include factors such as water density changes, temperature, and salinity. Therefore, it is considered to introduce the seafloor pressure data inverted from GRACE satellite gravity field data into the ocean data assimilation system to support the analysis of variables such as temperature and salinity. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide, in order to overcome the deficiencies of the prior art, an assimilation method for improving the accuracy of temperature and salinity analysis based on seafloor pressure data, which adjusts the vertical structure of the temperature and salinity fields according to seafloor pressure data, makes the temperature and salinity analysis fields more reasonable and credible, can obtain a seafloor pressure field with higher resolution and accuracy, and helps to analyze complex ocean dynamic processes such as ocean circulation, deep water body flow, and tidal action.

[0005] The technical solution adopted by the present invention is: an assimilation method for improving the accuracy of temperature and salinity analysis based on seafloor pressure data, including,

[0006] 1) Obtain various observation data required for assimilating to improve the accuracy of temperature and salinity analysis, including the observation data of temperature, salinity, sea surface temperature, sea surface salinity, and seafloor pressure in the form of increments relative to the background field, and the sea surface height data is the sea surface height anomaly;

[0007] 2) Construct the objective function of the four-dimensional spatio-temporal multi-scale data assimilation framework for assimilating to improve the accuracy of temperature and salinity analysis;

[0008] 3) Set the dynamic constraint conditions for the spatio-temporal four-dimensional multi-scale data assimilation framework to improve the accuracy of temperature and salinity analysis assimilation;

[0009] 4) Input the prepared observation data for joint data assimilation to reconstruct the temperature and salinity analysis field. An assimilation method for improving the accuracy of temperature and salinity analysis based on seafloor pressure data of the present invention has the following beneficial effects:

[0010] 1. It can enhance the vertical structure accuracy of the temperature and salinity fields

[0011] Seafloor pressure can provide unique information on the deep ocean state, especially in areas with greater water depth or complex seafloor topography. By combining and assimilating with in-situ observation data, it can effectively improve the prediction accuracy of ocean models for deep ocean circulation, temperature and salinity fields, etc. In particular, seafloor pressure reflects the change of seawater density, and the density change is closely related to temperature, salinity, and ocean current. Joint assimilation can help optimize the vertical temperature and salinity structure.

[0012] 2. It can improve data quality and reliability

[0013] Seafloor pressure data is scarce in some areas (such as the deep sea, remote sea areas, etc.), while in-situ observation data may be limited by measurement errors or insufficient spatial coverage in some cases. Joint assimilation can effectively improve the robustness of the system by combining the two types of data and make up for the deficiencies of a single data source. Through the fusion of multi-source data, it can improve the overall accuracy of ocean state analysis and ensure the reliability of the model under different regions and conditions.

[0014] 3. It can improve ocean disaster prediction and early warning

[0015] The change of seafloor pressure is closely related to ocean disasters such as submarine earthquakes and submarine landslides. At present, the acquisition and processing of GRACE satellite data are often offline and cannot provide real-time dynamic change information of seafloor pressure. However, joint assimilation can obtain seafloor pressure data with relatively high spatio-temporal resolution, thereby improving the prediction accuracy of ocean disasters. Description of the Drawings

[0016] Figure 1 is a flowchart of an assimilation method for improving the accuracy of temperature and salinity analysis based on seafloor pressure data of the present invention;

[0017] Figure 2 is a vertical distribution diagram of the root mean square error of temperature for three different assimilation schemes;

[0018] Figure 3 is a vertical distribution diagram of the root mean square error of salinity for three different assimilation schemes. Detailed Embodiment

[0019] The following will make a detailed description of an assimilation method for improving the accuracy of temperature and salinity analysis based on seabed pressure data of the present invention in combination with embodiments and drawings.

[0020] An assimilation method for improving the accuracy of temperature and salinity analysis based on seabed pressure data of the present invention is mainly based on a four-dimensional spatio-temporal multi-grid variational analysis method, introducing the joint assimilation of seabed pressure data and in-situ observation data, so as to effectively supplement the deficiencies of in-situ observation data and accurately adjust the vertical distribution of temperature and salinity in the ocean model, especially in areas rich in eddy information.

[0021] An assimilation method for improving the accuracy of temperature and salinity analysis based on seabed pressure data of the present invention specifically relates to a method of indirectly establishing the relationship between seabed pressure and temperature and salinity variables based on the natural physical relationship between seabed pressure and seawater density and sea surface height, so as to adjust the vertical structure of temperature and salinity by introducing seabed pressure data during the assimilation process. This method is mainly applied to the reanalysis of ocean observation data. It can make up for the lack of on-site information by introducing seabed pressure data in the case of sparse in-situ observation data, effectively reflect the water body state, and the finally generated analysis field not only conforms to the laws of ocean physics but also matches the actual observation data, thus improving the model's forecasting ability for deep water body movement, climate change, and ocean environmental changes.

[0022] An assimilation method for improving the accuracy of temperature and salinity analysis based on seabed pressure data of the present invention is an assimilation scheme for jointly assimilating seabed pressure and in-situ observations. On the one hand, it adjusts the vertical structure of the temperature and salinity field according to the seabed pressure data, making the temperature and salinity analysis field more reasonable and credible; on the other hand, the seabed pressure field with higher resolution and accuracy obtained by this assimilation scheme helps to analyze complex ocean dynamics processes such as ocean circulation, deep water body flow, and tidal action.

[0023] As Figure 1 shown, an assimilation method for improving the accuracy of temperature and salinity analysis based on seabed pressure data of the present invention includes:

[0024] 1) Obtain various observation data required for assimilating to improve the accuracy of temperature and salinity analysis, including the observation data of temperature, salinity, sea surface temperature, sea surface salinity, and seabed pressure in the form of increments relative to the background field, and the sea surface height data is the sea surface height anomaly;

[0025] Specifically, it includes systematically collecting satellite remote sensing observations of sea surface height anomaly, sea surface temperature, sea surface salinity, and gravity anomaly, as well as in-situ profile observations of temperature and salinity including those from Argo (Array for Real-time Geostrophic Oceanography), CTD (Conductivity-Temperature-Depth profiler), and XBT (Expendable Bathythermograph), and processing the in-situ profile observations of temperature and salinity into the form of increments relative to the background field temperature and salinity; calculating the seawater density using climatological temperature and salinity data, and thus retrieving the bottom pressure information from satellite remote sensing gravity anomaly data.

[0026] The retrieval of the bottom pressure information from satellite remote sensing gravity anomaly data adopts the method of retrieving the bottom pressure using the GRACE time-variable gravity field model established by Wahr et al., and the specific form is as follows:

[0027]

[0028] where θ and λ are the geocentric co-latitude and geocentric longitude respectively; a is the average radius of the Earth; g is the value of the average gravitational acceleration; ρ E is the average density of the Earth; ΔOBP(θ,λ) is the equivalent water column height of the mass change; l and m are the degree and order respectively; k l is the load Love number; is the normalized associated Legendre function; and are the changes in the spherical harmonic coefficients; during the retrieval, the spherical harmonic coefficients (GSM) of the geoid model and the spherical harmonic coefficients (GAD) corresponding to the atmosphere-ocean model in the same time period are used to calculate according to Equation (1) respectively, and then the calculation results of the two are added together to obtain the change in the bottom pressure.

[0029] 2) Construct the objective function of the spatio-temporal four-dimensional multi-scale data assimilation framework for improving the accuracy of temperature and salinity analysis and assimilation;

[0030] Under the three-dimensional variational analysis framework, the multi-grid technique is used, that is, variational analysis is carried out respectively from the coarse grid to the fine grid, and the large-scale, meso-scale to small and medium-scale information in the ocean numerical model and the observation data are successively optimally combined. The objective function of constructing the spatio-temporal four-dimensional multi-scale data assimilation framework for improving the accuracy of temperature and salinity analysis and assimilation is as follows:

[0031]

[0032] where n is the multiplicity of the analysis grid; and are control variables, namely the temperature increment and salinity increment of the n-th level analysis grid, both including four dimensions: meridional, zonal, vertical and time; and are the smoothing matrices of temperature and salinity, which are the integrals over the entire space of the square of the Laplacian operator of the control variables themselves; H T () is the temperature observation projection operator; H S () is the salinity observation projection operator; H SST () is the sea surface temperature observation projection operator; H SSS () is the sea surface salinity observation projection operator; H SSHa_PBT () is the sea surface height anomaly and sea floor pressure observation projection operator; The column vectors and represent the temperature background field and salinity background field; The column vector represents the temperature observation field; represents the salinity observation field; represents the sea surface temperature observation field; represents the sea surface salinity observation field; is the sea surface height anomaly observation increment; is the sea floor pressure observation increment; is the temperature observation field error covariance matrix of the n-th level analysis grid; is the salinity observation field error covariance matrix of the n-th level analysis grid; is the sea surface temperature observation field error covariance matrix of the n-th level analysis grid; is the sea surface salinity observation field error covariance matrix of the n-th level analysis grid; is the sea surface height anomaly and sea floor pressure observation field error covariance matrix of the n-th level analysis grid; p(·) is the sea floor pressure increment obtained by integrating the temperature and salinity analysis results and the sea surface height anomaly observation increment.

[0033] 3) Set the dynamic constraint conditions of the spatio-temporal four-dimensional multi-scale data assimilation framework for improving the accuracy of temperature and salinity analysis; The dynamic constraint conditions of the spatio-temporal four-dimensional multi-scale data assimilation framework for improving the accuracy of temperature and salinity analysis include the dynamic constraint conditions in two cases: the vertical height coordinate volume conservation mode and the vertical pressure coordinate mass conservation mode, specifically as follows:

[0034] (1) The dynamic constraint conditions in the vertical height coordinate volume conservation mode:

[0035]

[0036] Among them, p obs and p b are respectively the satellite remote sensing observation field and the model integration background field of the sea floor pressure anomaly; η obs and ηb are respectively the satellite remote sensing observation field of sea surface height anomaly and the background field of model integration; H is the water depth; then Δp obs and Δη obs are respectively the increment of bottom pressure observation and the increment of sea surface height anomaly observation; ρ[T(z), S(z), z] is the seawater equation of state; T(z) is the observed temperature value at water depth z; S(z) is the observed salinity value at water depth z; T b (z) is the background temperature value at water depth z; S b (z) is the background salinity value at water depth z; g is the acceleration of gravity; assume that the water column of seawater from the sea surface to the seabed at a certain location can be divided into K layers, the first layer is at the sea surface, and the Kth layer is at the seabed, T k 、S k 、z k and Δz k represent the observed temperature value, salinity value, vertical position and thickness of the kth layer; represents the background temperature and background salinity values of the kth layer; thus, the final form of the right - hand side of equation (3) is the operator p(·) of the bottom pressure increment obtained by integrating the temperature and salinity analysis results and the sea surface height anomaly observation increment of the objective function (equation (2)) of the four - dimensional multi - scale data assimilation framework in space and time;

[0037] (2) In the vertical pressure coordinate mass conservation model:

[0038]

[0039] where η obs and η b are respectively the satellite remote sensing observation field of sea surface height anomaly and the background field of model integration; g is the acceleration of gravity; p a is the sea surface atmospheric pressure; P obs and P b are respectively the satellite remote sensing observation field of bottom pressure and the background field of model integration; ρ[T(p), S(p), p] is the seawater equation of state, T(p) is the observed temperature value at underwater pressure p; S(p) is the observed salinity value at underwater pressure p, T b (p) is the background temperature value at underwater pressure p; S b (p) is the background salinity value at underwater pressure p; assume that the water column of seawater from the sea surface to the seabed at a certain location can be divided into K layers, the first layer is at the sea surface, and the Kth layer is at the seabed, T k 、S k 、p k and Δp k represent the observed temperature value, salinity value, vertical pressure position and pressure grid thickness of the kth layer; It represents the temperature background and salinity background values of the k-th layer. Thus, the final form of the right side of Equation (4) is the operator p(·) of the seabed pressure increment obtained by integrating the temperature and salinity analysis results and the sea surface height anomaly observation increment of the objective function (Equation (2)) of the four-dimensional spatio-temporal multi-scale data assimilation framework.

[0040] 4) Input the prepared observation data and perform joint data assimilation to reconstruct the temperature and salinity analysis fields. It includes:

[0041] Input the various observation data obtained in step 1) into the four-dimensional spatio-temporal multi-scale data assimilation framework for improving the accuracy of temperature and salinity analysis established, and perform data assimilation.

[0042] Such as Figure 2 、 Figure 3 shown, taking the application to the volume conservation model of the vertical height coordinate as an example, Figure 2 is the vertical distribution map of the root mean square error of temperature for three different assimilation schemes, Figure 3 is the vertical distribution map of the root mean square error of salinity for three different assimilation schemes. Compared with the other two assimilation schemes of only assimilating the in-situ observations of temperature and salinity and only assimilating the satellite remote sensing sea surface temperature, sea surface salinity and sea surface height observations, this joint assimilation scheme can effectively reduce the analysis errors of temperature and salinity.

Claims

1. An assimilation method for improving the accuracy of temperature-salinity analysis based on seafloor pressure data, characterized in that: include, 1) Obtain various observation data required for improving the accuracy of temperature and salinity analysis, including temperature, salinity, sea surface temperature, sea surface salinity and seabed pressure in the form of increments relative to the background field, and sea surface height data in the form of sea surface height anomalies; 2) Construct the objective function of the spatiotemporal four-dimensional multi-scale data assimilation framework to improve the accuracy of temperature and salinity analysis; 3) Setting dynamic constraints for the spatiotemporal four-dimensional multi-scale data assimilation framework to improve the accuracy of temperature and salinity analysis; 4) Input the prepared observation data, perform joint data assimilation, and reconstruct the temperature-salinity analysis field.

2. The assimilation method for improving the accuracy of temperature-salinity analysis based on seafloor pressure data according to claim 1, characterized in that: Step 1) includes: systematically collecting satellite remote sensing sea surface height anomaly, sea surface temperature, sea surface salinity and gravity anomaly observation data, as well as temperature and salinity field profile observation data including Argo, CTD, and XBT, and processing the temperature and salinity field profile observation data into an incremental form relative to the background field temperature and salinity; using climatological temperature and salinity data to calculate seawater density, thereby inverting seabed pressure information from satellite remote sensing gravity anomaly data.

3. The assimilation method for improving the accuracy of temperature-salinity analysis based on seafloor pressure data according to claim 2, characterized in that: The inversion of seafloor pressure information from satellite remote sensing gravity anomaly data is a method for inverting seafloor pressure using the GRACE time-varying gravity field model established by Wahr et al. The specific form is as follows: Among them, θ and λ are the geocentric colatitude and geocentric longitude respectively; a is the average radius of the earth; g is the average gravitational acceleration value; ρ E is the average density of the Earth; ΔOBP(θ,λ) is the equivalent water column height of mass change; l and m are the order and degree respectively; k l is the load Love number; is the normalized associated Legendre function; and is the change in spherical harmonic coefficients; during inversion, the spherical harmonic potential coefficients of the geoid model and the spherical harmonic potential coefficients corresponding to the atmosphere-ocean model in the same time period are used to calculate according to formula (1), and then the results of the two calculations are added together to obtain the change in seabed pressure.

4. The assimilation method for improving the accuracy of temperature-salinity analysis based on seafloor pressure data according to claim 1, characterized in that: The objective function of constructing a spatiotemporal four-dimensional multiscale data assimilation framework suitable for improving the accuracy of temperature and salinity analysis described in step 2) is as follows: Where n is the analysis grid multiplicity; and are the control variables, which are the temperature increment and salinity increment of the nth analysis grid, respectively, and both contain four dimensions: longitude, latitude, vertical and time; and is the smoothing matrix of temperature and salinity, which is the integral of the square of the Laplace operator of the control variable itself in the whole space; H T () is the temperature observation projection operator; H S () is the salinity observation projection operator; H SST () is the sea surface temperature observation projection operator; H SSS () is the sea surface salinity observation projection operator; H SSHa_PBT () is the projection operator of sea surface height anomaly and seabed pressure observation; column vector and Represents the temperature background field and salinity background field; column vector represents the temperature observation field; represents the salinity observation field; represents the sea surface temperature observation field; represents the sea surface salinity observation site; is the observed increment of sea surface height anomaly; is the observed increment of seafloor pressure; is the temperature observation field error covariance matrix of the nth analysis grid; is the error covariance matrix of the salinity observation field of the nth analysis grid; is the error covariance matrix of the sea surface temperature observation field of the nth analysis grid; is the error covariance matrix of the sea surface salinity observation field of the nth analysis grid; is the error covariance matrix of the sea surface height anomaly and the seabed pressure observation field of the nth analysis grid; p(·) is the seabed pressure increment obtained by integrating the temperature and salinity analysis results and the sea surface height anomaly observation increment.

5. The assimilation method for improving the accuracy of temperature-salinity analysis based on seafloor pressure data according to claim 1, characterized in that: The dynamic constraint conditions of the spatiotemporal four-dimensional multi-scale data assimilation framework for improving the accuracy of temperature and salinity analysis described in step 3) include dynamic constraint conditions used in the vertical height coordinate volume conservation mode and the vertical pressure coordinate mass conservation mode, which are as follows: (1) Dynamic constraints under the vertical height coordinate volume conservation mode: Among them, p obs and p b are the satellite remote sensing observation field and the model integrated background field of seafloor pressure anomaly; η obs and η b are the satellite remote sensing observation field of sea surface height anomaly and the model integrated background field respectively; H is the water depth; then Δp obs and Δη obs are the observed increments of seafloor pressure and sea surface height anomaly respectively; ρ[T(z),S(z),z] is the seawater state equation; T(z) is the observed temperature value at water depth z; S(z) is the observed salinity value at water depth z; T b (z) is the temperature background value when the water depth is z; S b (z) is the salinity background value when the water depth is z; g is the gravitational acceleration; Assume that the water column from the sea surface to the sea bottom at a certain location can be divided into K layers, the first layer is on the sea surface, the Kth layer is on the sea bottom, T k , S k 、z k and Δz k represents the temperature observation, salinity observation, vertical position and thickness of the kth layer; represents the temperature background and salinity background value of the kth layer; therefore, the final form of the right side of the equation (3) is the operator p(·) of the seafloor pressure increment obtained by integrating the temperature and salinity analysis results and the sea surface height anomaly observation increment of the objective function of the spatiotemporal four-dimensional multiscale data assimilation framework (Equation (2)); (2) In the vertical pressure coordinate mass conservation mode: Among them, η obs and η b are the satellite remote sensing observation field of sea surface height anomaly and the model integrated background field; g is the gravitational acceleration; p is the a is the atmospheric pressure at sea surface; P obs and P b are the satellite remote sensing observation field and the model integrated background field of the seafloor pressure respectively; ρ[T(p),S(p),p] is the seawater state equation, T(p) is the temperature observation value when the underwater pressure is p; S(p) is the salinity observation value when the underwater pressure is p, T b (p) is the temperature background value when the underwater pressure is p; S b (p) is the salinity background value when the underwater pressure is p; Assume that the water column from the sea surface to the sea bottom at a certain location can be divided into K layers, the first layer is at the sea surface, the Kth layer is at the sea bottom, T k , S k 、p k and Δp k represents the temperature observation value, salinity observation value, vertical pressure position and pressure grid thickness of the kth layer; represents the temperature background and salinity background value of the kth layer. Therefore, the final form of the right side of the equation (4) is the operator p(·) of the seafloor pressure increment obtained by integrating the temperature and salinity analysis results and the sea surface height anomaly observation increment of the objective function (Equation (2)) of the spatiotemporal four-dimensional multi-scale data assimilation framework.

6. The assimilation method for improving the accuracy of temperature-salinity analysis based on seafloor pressure data according to claim 1, characterized in that: Step 4) includes: inputting various observation data obtained in step 1) into the established spatiotemporal four-dimensional multi-scale data assimilation framework for improving the accuracy of temperature and salinity analysis, and performing data assimilation.

Images