Marine environment forecasting method and system based on energy balance and data fusion
Patent Information
- Application Number
- CN202511780840.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-28
- Publication Date
- 2026-03-06
Smart Images
Figure CN121614731A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of marine environmental forecasting technology, and specifically relates to a marine environmental forecasting method and system based on energy balance and data fusion. Background Technology
[0002] Marine environmental forecasting is an important field of marine science and engineering. Its core lies in simulating the coupling mechanism of air-sea interaction and dynamic processes such as waves and ocean currents through physical numerical models.
[0003] Traditional physical models are insufficient in characterizing multi-scale energy transfer processes, especially in high-resolution regions and extreme weather events where forecast accuracy declines. At the same time, multi-source observation data from satellites, buoys, drones, and other sources have not been fully and effectively integrated into the existing forecasting system, resulting in the underutilization of data value.
[0004] Current mainstream technical solutions primarily rely on numerical simulations based on physical laws, describing ocean dynamic processes by constructing governing equations (such as hydrodynamic equations) and parameterization schemes. To utilize observational data, some advanced solutions introduce data assimilation techniques (such as variational methods or ensemble Kalman filtering) to attempt to integrate observations into the model's initial field or perform state corrections, thereby improving the forecast starting point. Furthermore, some studies have explored using data-driven machine learning models alone, directly learning mapping relationships from historical data for prediction.
[0005] However, existing technical solutions have significant limitations. On the one hand, pure physical models are computationally expensive, and the parameterization of subgrid-scale processes is uncertain, leading to large deviations in the simulation of energy transfer processes in high-resolution simulations and extreme events (such as typhoons). On the other hand, although data assimilation techniques can partially fuse data, their assimilation efficiency is limited by the observation network layout and assimilation window, making it difficult to dynamically optimize the weights of multi-source, multi-frequency data, especially lacking the real-time correction capability for high-frequency observation data under extreme weather conditions. Pure machine learning models, on the other hand, lack physical constraints, have poor interpretability, and are prone to producing unreasonable results in extrapolation and long-term forecasts. Summary of the Invention
[0006] The purpose of this invention is to provide a marine environment forecasting method and system based on energy balance and data fusion to solve the above-mentioned problems.
[0007] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a marine environment forecasting method based on energy balance and data fusion, comprising: Collect satellite remote sensing data, field observation data, and historical reanalysis data, and preprocess the data to generate a spatially continuous dataset; By constructing regional energy balance equations and coupling ocean dynamics and thermodynamics, a physical process model of energy balance is obtained. Real-time observation data is dynamically integrated into the physical process model of energy balance to correct state variables, and the weight of high-frequency data is increased through a weighting strategy under extreme weather conditions; at the same time, machine learning models are combined to achieve a complementary integration of data-driven and physical processes. By optimizing the coupling parameters between waves and ocean currents, a multi-scale nested simulation framework is established, boundary conditions are dynamically adjusted, and prediction results are output.
[0008] Furthermore, the collected satellite remote sensing data, field observation data, and historical reanalysis data include: Satellite remote sensing data: Acquire remote sensing observation data on wind speed, wave height, sea surface temperature (SST), and chlorophyll concentration; Field observation data: Integrating observation data from buoys, sonar, and UAVs for local refinement and dynamic correction; Historical reanalysis data: Introduce reanalysis data from ERA5 or CMEMS to construct long-term time series for model training and validation.
[0009] Furthermore, the preprocessing of the data to generate a spatially continuous dataset includes: Outliers are detected and removed; statistical methods and machine learning are used to improve data quality; multi-scale interpolation methods are used to generate spatially continuous datasets. Fourier transform is performed on the time series data to separate high-frequency and low-frequency components; low-frequency seasonal features and high-frequency sudden event information are extracted to provide input for subsequent modeling.
[0010] Furthermore, the construction of the partitioned energy balance equation, coupled with ocean dynamics and thermodynamics, yields a physical process model of energy balance, including: A partitioned energy balance model, coupled with ocean dynamics and thermodynamics, is used to describe the energy input, transfer, and dissipation processes: Input energy: driven by wind stress, absorbed by solar radiation; Energy transfer: interaction between waves and ocean currents, and the transfer of turbulence at the sea surface; Energy dissipation: bottom friction, whitening wave dissipation. Observational data is used to verify the local energy closure and calibrate the parameters of the energy balance equation. For extreme events such as typhoons and storm surges, the energy distribution and dissipation during the high energy input phase are simulated, and the parameterization scheme is optimized.
[0011] Furthermore, the dynamic integration of real-time observation data into the physical process model of energy balance, correction of state variables, and enhancement of high-frequency data weights through a weighting strategy under extreme weather conditions, combined with machine learning models, achieves a complementary fusion of data-driven and physical processes, including: Ensemble Kalman filtering is used to integrate real-time observation data into the physical model and dynamically correct the model state variables. In extreme weather conditions, the influence weight of high-frequency observation data is increased through weighted assimilation. To address the need for localized precision, machine learning models are used to supplement nonlinear features that are difficult to capture by conventional models; the data-driven model takes energy balance simulation results and observational data as inputs and outputs corrected key variables.
[0012] Furthermore, by optimizing the coupling parameters between waves and ocean currents, a multi-scale nested simulation framework is established, boundary conditions are dynamically adjusted, and prediction results are output, including: The parameterization in the wave-current coupling process is optimized to simulate the dynamic changes in wave propagation paths and local ocean current distribution within a high-resolution region; a regional nested model is established, incorporating high-resolution sub-models within a large-scale model to finely characterize local features; and the nested boundary conditions are dynamically adjusted to ensure consistency of multi-scale simulation results.
[0013] Furthermore, the accuracy of the wind speed, wave height, and ocean current variables output by the model is evaluated using independent observation data, and the correlation coefficient, root mean square error, and observation bias are calculated. For the energy balance model, the closure of the energy input, transfer, and dissipation processes is verified. Sensitivity analysis is performed on the key parameters of the energy balance model to determine the main influencing factors of the forecast results. The robustness of the model is tested for different meteorological scenarios.
[0014] Secondly, the present invention provides a marine environment forecasting system based on energy balance and data fusion, comprising: The data acquisition module is used to collect satellite remote sensing data, field observation data, and historical reanalysis data, and to preprocess the data to generate a spatial continuous dataset. The model building module is used to construct the regional energy balance equations, couple ocean dynamics and thermodynamics processes, and obtain a physical process model of energy balance. The fusion module is used to dynamically integrate real-time observation data into the physical process model of energy balance, correct state variables, and increase the weight of high-frequency data through a weighting strategy under extreme weather conditions; at the same time, it combines machine learning models to achieve complementary integration of data-driven and physical processes. The output module is used to establish a multi-scale nested simulation framework by optimizing the coupling parameters between waves and ocean currents, dynamically adjusting boundary conditions, and outputting prediction results.
[0015] Thirdly, the present invention provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the marine environment forecasting method based on energy balance and data fusion.
[0016] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the marine environment forecasting method based on energy balance and data fusion.
[0017] Compared with the prior art, the present invention has the following technical effects: This invention uses the principle of energy balance to scientifically describe complex ocean dynamic processes and improve the physical consistency of models.
[0018] This invention utilizes multi-source data supplementation mode input and correction output to fully explore the potential of observation data.
[0019] This invention supports multi-scale, high-resolution simulations from global to local levels, meeting the needs of various scenarios.
[0020] This invention integrates real-time updates, dynamic correction, and user customization functions, providing an efficient tool for scientific research and practical applications.
[0021] This method, by deeply integrating physical modeling with data-driven technology, provides a novel, accurate, and practical solution for marine environmental forecasting, and has broad application prospects. Attached Figure Description
[0022] Figure 1 This is a flowchart of the present invention. Detailed Implementation
[0023] The present invention will be further described below with reference to the accompanying drawings: Example 1, please refer to Figure 1 This invention provides a marine environment forecasting method based on energy balance and data fusion, comprising: Collect satellite remote sensing data, field observation data, and historical reanalysis data, and preprocess the data to generate a spatially continuous dataset; By constructing regional energy balance equations and coupling ocean dynamics and thermodynamics, a physical process model of energy balance is obtained. Real-time observation data is dynamically integrated into the physical process model of energy balance to correct state variables, and the weight of high-frequency data is increased through a weighting strategy under extreme weather conditions; at the same time, machine learning models are combined to achieve a complementary integration of data-driven and physical processes. By optimizing the coupling parameters between waves and ocean currents, a multi-scale nested simulation framework is established, boundary conditions are dynamically adjusted, and prediction results are output.
[0024] This invention addresses the issues of insufficient data coverage and low data quality in traditional methods through multi-source data fusion and preprocessing, providing a more comprehensive and cleaner input foundation for the model. The introduction of an energy balance physics model enhances the physical interpretability of the forecast, particularly accurately characterizing energy dynamics during extreme events. By combining data assimilation with machine learning, model biases are dynamically corrected, improving the adaptability of real-time forecasts. Coupled optimization and multi-scale simulation ensure seamless integration from global to local perspectives, and the validation mechanism ensures the method remains stable under various ocean conditions. Overall, this method not only improves the accuracy of short-term forecasts but also supports long-term trend analysis and customized applications, providing more reliable technical support for marine disaster prevention and resource management.
[0025] Example 2: This invention provides a marine environment forecasting method based on energy balance and data fusion, comprising: Step 1: Multi-source data acquisition and preprocessing 1. Data Acquisition: Satellite data: Acquire remote sensing observation data such as wind speed, wave height, sea surface temperature (SST), and chlorophyll concentration to ensure wide-area coverage.
[0026] Field data: Integrating observation data from buoys, sonar, and drones for local refinement and dynamic correction.
[0027] Historical data: Introduce reanalysis data (such as ERA5, CMEMS) to construct long-term time series for model training and validation.
[0028] 2. Data cleaning and interpolation: Detect and remove outliers, and improve data quality using statistical methods (such as box plots) and machine learning methods (such as the Isolation Forest algorithm).
[0029] Spatial continuity datasets are generated using multi-scale interpolation methods such as Kriging interpolation.
[0030] 3. Time-frequency decomposition: Perform Fourier transform on time series data to separate high-frequency and low-frequency components.
[0031] Extract seasonal features (low frequency) and sudden events (high frequency) information to provide input for subsequent modeling.
[0032] Step 2: Modeling the physical processes based on energy balance 1. Construct the energy balance equation: A partitioned energy balance model is used, coupling ocean dynamics (waves, currents) and thermodynamics (evaporation, radiation) processes, to describe the energy input, transfer, and dissipation processes: Input energy: such as wind stress drive, solar radiation absorption.
[0033] Energy transfer: such as the interaction between waves and ocean currents, and the transfer of energy through sea surface turbulence.
[0034] Energy dissipation: such as bottom friction and whitening wave dissipation.
[0035] 2. Verification of local energy closure: Use observational data to verify the local energy closure and calibrate the parameters of the energy balance equation (such as wave attenuation coefficient and heat flux).
[0036] 3. Analysis of extreme event processes: For extreme events such as typhoons and storm surges, the focus is on simulating the energy distribution and dissipation during the high-energy input phase, and optimizing the parameterization scheme.
[0037] Step 3: Fusion of multi-source data and physical model 1. Data assimilation techniques: Ensemble Kalman Filtering (EnKF) is used to integrate real-time observation data into the physical model and dynamically correct model state variables (such as wind speed and sea surface temperature).
[0038] In extreme weather conditions, the influence weight of high-frequency observation data (such as buoys) can be increased through weighted assimilation. 2. Data-driven supplement: To address the need for localized refinement, machine learning models (such as random forests and deep neural networks) are used to supplement nonlinear features that are difficult to capture by conventional patterns.
[0039] The data-driven model takes energy balance simulation results and observational data as inputs and outputs corrected key variables.
[0040] Step 4: Coupling Optimization Between Variables and Multi-Scale Simulation 1. Wave-current coupling optimization: Optimize the parameterization scheme in the wave-current coupling process, such as the feedback effect of wave-induced surface mixing on ocean currents.
[0041] Simulate the dynamic changes in wave propagation paths and local ocean current distribution within a high-resolution region.
[0042] 2. Multi-scale simulation framework: Establish a region nested model, nesting high-resolution sub-models within a large-scale model, to finely characterize local features.
[0043] Dynamically adjust nested boundary conditions to ensure consistency of multi-scale simulation results.
[0044] Step 5: Model Validation and Performance Evaluation 1. Multivariate validation: The accuracy of variables such as wind speed, wave height, and ocean current output by the model is evaluated using independent observation data, and the correlation coefficient, root mean square error (RMSE), and observation bias are calculated.
[0045] For the energy balance model, the closure of the energy input, transfer, and dissipation processes is verified.
[0046] 2. Sensitivity analysis: Sensitivity analysis was conducted on key parameters of the energy balance model (such as wind stress coefficient) to determine the main influencing factors of the forecast results.
[0047] The robustness of the model was tested for different meteorological scenarios (such as typhoons and calm sea conditions).
[0048] Step 6: Application and Extension 1. Real-time forecasting and dynamic adjustment: Develop a real-time forecasting platform that dynamically updates output results by combining physical models and observational data.
[0049] It supports user-defined requirements (such as wave height forecasts accurate to a specific location).
[0050] 2. Cross-domain applications: It is applied to marine disaster prevention (such as storm surge warning), marine energy development (such as wind farm site selection and operation) and ecological environment protection (such as oil spill monitoring).
[0051] 3. Intelligent expansion: Reinforcement learning techniques are introduced to optimize pattern parameters and computational efficiency.
[0052] Develop a cloud-based intelligent system to provide online access and interaction services for multiple users.
[0053] This invention uses the principle of energy balance to scientifically describe complex ocean dynamic processes and improve the physical consistency of models.
[0054] This invention utilizes multi-source data supplementation mode input and correction output to fully explore the potential of observation data.
[0055] This invention supports multi-scale, high-resolution simulations from global to local levels, meeting the needs of various scenarios.
[0056] This invention integrates real-time updates, dynamic correction, and user customization functions, providing an efficient tool for scientific research and practical applications.
[0057] This method, by deeply integrating physical modeling with data-driven technology, provides a novel, accurate, and practical solution for marine environmental forecasting, and has broad application prospects.
[0058] This invention constructs a forecasting framework that accurately describes multi-scale energy transfer processes in the ocean through a deep integration of physical mechanisms and data-driven approaches. Its theoretical foundation is the energy balance principle, which establishes a physically constrained model by quantifying the entire process of energy input (e.g., wind stress, solar radiation), transfer (e.g., wave-current interaction), and dissipation (e.g., bottom friction). Simultaneously, it utilizes multi-source observational data (satellites, buoys, etc.) to dynamically correct and supplement the model. In particular, it employs data assimilation techniques (e.g., ensemble Kalman filter, EnKF) and machine learning methods to compensate for the shortcomings of purely physical models in characterizing nonlinear features and responding to extreme events, thereby achieving high-precision, adaptive forecasts from global to local levels and from normal to extreme events.
[0059] The specific methods for achieving the above principles involve four key steps. First, multi-source data acquisition and preprocessing are performed, integrating satellite remote sensing, on-site observations (buoys, UAVs), and historical reanalysis data (such as ERA5 and CMEMS). High-quality, continuous datasets are generated through data cleaning (using machine learning algorithms such as Isolation Forest to remove outliers) and multi-scale interpolation (such as Kriging interpolation). Seasonal and sudden event characteristics are extracted using time-frequency decomposition (such as Fourier transform). Second, a physical model based on energy balance is constructed, establishing regional energy balance equations, coupling dynamic (waves, ocean currents) and thermodynamic (evaporation, radiation) processes, and using observational data to verify local energy closure, calibrating key parameters (such as wave attenuation coefficients), and optimizing parameterization schemes for extreme events such as typhoons. Third, a data assimilation and machine learning fusion strategy is adopted. EnKF is used to dynamically integrate real-time observational data into the physical model to correct state variables (such as sea surface temperature), and high-frequency data is weighted under extreme weather conditions. Simultaneously, machine learning models such as random forests and deep neural networks are trained, using the physical model output and observational data as inputs to supplement nonlinear features that are difficult for the physical model to capture. Finally, variable coupling optimization and multi-scale simulation were performed to optimize the coupling parameterization between waves and ocean currents (such as surface mixing effect), establish a regional nested model and dynamically adjust the boundary conditions, and then verify the multivariate accuracy (calculate root mean square error RMSE, etc.) through independent observation data and conduct sensitivity analysis to ensure the robustness of the model under different scenarios.
[0060] This invention, through the systematic implementation of the aforementioned principles and methods, has achieved significant technical results. Regarding forecast accuracy, the energy balance model provides a solid physical foundation, and combined with dynamic correction from multi-source data, it can more accurately simulate complex ocean dynamic processes, especially in extreme events such as typhoons and storm surges, significantly improving the characterization of energy distribution and dissipation. In terms of adaptability and practicality, the multi-scale simulation framework achieves seamless integration from large-scale to local high-resolution regions. Combined with real-time data assimilation and machine learning supplementation, the forecast system can dynamically adapt to different sea states and support user-defined refined forecast requirements (such as wave height at specific points). Ultimately, this method not only improves the accuracy and reliability of short-term marine environmental forecasts but also provides a more efficient and reliable decision support tool for practical applications such as marine disaster prevention and mitigation, energy development, and ecological protection.
[0061] In another embodiment of the present invention, a marine environment forecasting system based on energy balance and data fusion is provided, which can be used to implement the above-mentioned marine environment forecasting method based on energy balance and data fusion. Specifically, the system includes: The data acquisition module is used to collect satellite remote sensing data, field observation data, and historical reanalysis data, and to preprocess the data to generate a spatial continuous dataset. The model building module is used to construct the regional energy balance equations, couple ocean dynamics and thermodynamics processes, and obtain a physical process model of energy balance. The fusion module is used to dynamically integrate real-time observation data into the physical process model of energy balance, correct state variables, and increase the weight of high-frequency data through a weighting strategy under extreme weather conditions; at the same time, it combines machine learning models to achieve complementary integration of data-driven and physical processes. The output module is used to establish a multi-scale nested simulation framework by optimizing the coupling parameters between waves and ocean currents, dynamically adjusting boundary conditions, and outputting prediction results.
[0062] The module division in this embodiment of the invention is illustrative and represents only one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional modules in the various embodiments of the invention can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.
[0063] In another embodiment of the present invention, a computer device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to achieve a corresponding method flow or corresponding function. The processor described in this embodiment of the present invention can be used for the operation of a marine environmental forecasting method based on energy balance and data fusion.
[0064] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the marine environment forecasting method based on energy balance and data fusion in the above embodiments.
[0065] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0066] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0067] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0068] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.
Claims
1. A method for marine environmental forecasting based on energy balance and data fusion, characterized in that, The application relates to a method for predicting extreme weather events, and belongs to the field of meteorological prediction. Satellite remote sensing data, field observation data and historical reanalysis data are collected, and the data is preprocessed to generate a spatially continuous data set; A partition energy balance equation is constructed, and ocean dynamics and thermodynamics are coupled to obtain a physical process model of energy balance; Real-time observation data is dynamically integrated into the physical process model of energy balance to correct the state variables, and the weight of high-frequency data is increased under extreme weather through a weighting strategy; meanwhile, a machine learning model is combined to realize complementary fusion of data driving and physical processes; A multi-scale nested simulation framework is established by optimizing the coupling parameters between waves and ocean currents, and the boundary conditions are dynamically adjusted to output the prediction results.
2. The energy balance and data fusion based marine environmental forecasting method according to claim 1, characterized in that, The satellite remote sensing data, field observation data and historical reanalysis data include: Satellite remote sensing data: wind speed, wave height, sea surface temperature (SST) and chlorophyll concentration remote sensing observation data are obtained; Field observation data: buoy, sonar and unmanned aerial vehicle observation data are integrated for local refinement and dynamic correction; Historical reanalysis data: ERA5 or CMEMS reanalysis data are introduced to construct a long-term time series for model training and verification.
3. The energy balance and data fusion based marine environmental forecasting method according to claim 1, wherein, The data is preprocessed to generate a spatially continuous data set, including: Abnormal values are detected and removed, and statistical methods and machine learning are used to improve data quality; a multi-scale interpolation method is used to generate a spatially continuous data set; Fourier transform is performed on the time series data to separate high-frequency and low-frequency components; the seasonal characteristics of the low-frequency components and the burst event information of the high-frequency components are extracted to provide input for subsequent modeling.
4. The energy balance and data fusion based marine environmental forecasting method according to claim 1, wherein, The partition energy balance equation is constructed, and ocean dynamics and thermodynamics are coupled to obtain a physical process model of energy balance, including: A partition energy balance model is used to couple ocean dynamics and thermodynamics to describe the input, transfer and dissipation processes of energy: Input energy: wind stress driving, solar radiation absorption; Energy transfer: wave-current interaction, sea surface turbulent transfer; Dissipated energy: bottom friction, white wave dissipation; The local energy closure condition is verified by using observation data to calibrate the parameters of the energy balance equation; for extreme events such as typhoons and storm surges, the energy distribution and dissipation in the high-energy input stage are simulated, and the parameterization scheme is optimized.
5. The energy balance and data fusion based marine environmental forecasting method according to claim 1, wherein, Real-time observation data is dynamically integrated into the physical process model of energy balance to correct the state variables, and the weight of high-frequency data is increased under extreme weather through a weighting strategy; a machine learning model is combined to realize complementary fusion of data driving and physical processes, including: Ensemble Kalman filtering is used to integrate real-time observation data into the physical model to dynamically correct the model state variables; under extreme weather conditions, the influence weight of high-frequency observation data is increased by weighted assimilation; For local refinement requirements, a machine learning model is used to supplement nonlinear characteristics that are difficult for the model to capture; the data-driven model inputs come from energy balance simulation results and observation data, and the output is the corrected key variables.
6. The energy balance and data fusion based marine environmental forecasting method according to claim 1, wherein, The coupling parameters between waves and ocean currents are optimized to establish a multi-scale nested simulation framework, and the boundary conditions are dynamically adjusted to output the prediction results. Optimize the parameterization in the wave-current coupling process, simulate the dynamic changes of wave propagation paths and local ocean current distribution in high-resolution areas; Establish a regional nested model, nest a high-resolution sub-model in a large-scale model, to finely depict local features; Dynamically adjust the boundary conditions of the nested model to ensure the consistency of the multi-scale simulation results.
7. The energy balance and data fusion based marine environmental forecasting method according to claim 6, characterized in that, Use independent observation data to evaluate the accuracy of the model output of wind speed, wave height, and sea current variables, calculate the correlation coefficient, root mean square error, and observation bias; For the energy balance model, verify the closure of the energy input, transfer, and dissipation processes; Perform sensitivity analysis on the key parameters of the energy balance model to determine the main influencing factors of the prediction results; Test the robustness of the model for different weather scenarios.
8. A marine environment forecasting system based on energy balance and data fusion, characterized in that, It includes: A data acquisition module for acquiring satellite remote sensing data, field observation data, and historical reanalysis data, and preprocessing the data to generate a spatially continuous data set; A model construction module for constructing a partition energy balance equation, coupling ocean dynamic processes and thermodynamic processes, and obtaining an energy balance physical process model; A fusion module for dynamically integrating real-time observation data into the energy balance physical process model, correcting state variables, and increasing the weight of high-frequency data under extreme weather conditions through a weighted strategy; At the same time, combine machine learning models to achieve complementary fusion of data-driven and physical processes; An output module for optimizing the coupling parameters between waves and ocean currents, establishing a multi-scale nested simulation framework, dynamically adjusting the boundary conditions, and outputting the prediction results.
9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to realize the steps of the ocean environment prediction method based on energy balance and data fusion according to any one of claims 1 to 7.
10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program is executed by the processor to realize the steps of the ocean environment prediction method based on energy balance and data fusion according to any one of claims 1 to 7.