Method and system for optimizing COARE model based on sea ice concentration

Abstract

The invention relates to the technical field of meteorological model optimization and polar region ocean data processing, in particular to a method and system for optimizing a COARE model based on sea ice density, and the method comprises the steps: obtaining the historical actual measurement data of the polar region ocean, calculating the real roughness and the model roughness outputted by the COARE model, and determining the roughness difference value; dividing a wind speed interval, and fitting and establishing a correction function of a difference value between the sea ice concentration and the roughness; and introducing the sea ice concentration into the COARE model, judging the interval according to the wind speed, calling a corresponding function to correct the roughness of the model, and outputting the ocean-atmosphere flux conforming to the polar region sea ice condition to complete the model optimization. According to the method, the sea ice concentration is introduced, and the wind speed partition correction relation is established, so that the roughness of the polar region under the sea ice coverage condition is dynamically corrected by the COARE model, the limitation of the original model in the aspects of wind speed applicability and polar region adaptability is broken through, and the accuracy and practicability of polar region ocean-atmosphere flux calculation are remarkably improved.

Description

Technical Field

[0001] This invention relates to the fields of meteorological model optimization and polar ocean data processing, and in particular to a method and system for optimizing the COARE model based on sea ice concentration. Background Technology

[0002] Energy and momentum exchange between the ocean and atmosphere is a crucial factor driving climate development and a key area of ​​international ocean and atmospheric research. Due to the difficulty in acquiring ocean data, tropical ocean coupling experiments were conducted as early as the 1970s, leading to the development of the COARE model. However, when the existing COARE model is directly applied to polar oceans, the estimation results of ocean-atmosphere fluxes show significant deviations, severely limiting the effectiveness and applicability of the COARE model in polar ocean scientific research and climate prediction. Therefore, the existing COARE model suffers from technical defects such as inaccurate calculations and limited applicability under polar ocean conditions, necessitating urgent improvement.

[0003] Therefore, there is an urgent need to provide a technical solution to address the above problems. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method and system for optimizing the COARE model based on sea ice concentration.

[0005] Firstly, this invention provides a method for optimizing the COARE model based on sea ice concentration, the technical solution of which is as follows:

[0006] Based on historical measured data collected in the polar ocean within a specified time period, for each observation time, the true roughness representing the actual sea surface condition is obtained, and the meteorological parameters required by the standard COARE model defined in the historical measured data of that observation time are input into the COARE model to obtain the uncorrected model roughness calculated internally by the model, and the roughness difference value is determined based on the true roughness and the model roughness.

[0007] According to the preset wind speed zoning strategy, the wind speed data in the historical measured data is divided into multiple wind speed intervals. Based on the sea ice concentration observation value in the historical measured data and the data subset in each wind speed interval, a correction function is established to associate the sea ice concentration observation value of each data point in the data subset with the roughness difference value of each data point to correct the roughness of the model.

[0008] Real-time measured data of polar oceans are input into the COARE model. During the operation of the COARE model, the correction function corresponding to the wind speed range to which the wind speed data in the real-time measured data belongs is called. Combined with the sea ice concentration observation value in the real-time measured data, the roughness of the model is corrected in real time to optimize the operation of the COARE model and obtain the output ocean-atmosphere flux that meets the polar sea ice coverage conditions.

[0009] The beneficial effects of the method for optimizing the COARE model based on sea ice concentration according to the present invention are as follows:

[0010] The method of this invention, by introducing sea ice concentration and establishing a zonal correction relationship, realizes the dynamic correction of roughness under polar sea ice conditions by the COARE model, breaks through the limitations of the original model in terms of wind speed applicability and polar adaptability, and significantly improves the accuracy and practicality of polar momentum flux calculation.

[0011] It should be noted that the core of this invention lies in addressing the systematic bias problem existing in the COARE model when applied to polar ocean regions. To accurately understand the application scenarios of this invention, the relevant ocean areas must first be defined:

[0012] Open ocean typically refers to large bodies of water far from the continental shelf and unaffected by coastlines, islands, or sea ice. Its air-sea exchange processes are primarily driven by the physical properties of the open ocean (such as waves and water temperature) and atmospheric conditions (such as wind fields). Existing COARE models are built and optimized based on data from such environments, but their parameterization schemes do not include sea ice, a crucial underlying surface factor.

[0013] The polar oceans refer to the ocean regions surrounding the Earth's North and South Poles, geographically located roughly within the polar circles. Their most prominent feature is the presence of seasonal or multi-year sea ice cover. The presence of sea ice drastically alters the physical properties (such as albedo and roughness) and dynamic processes of the ocean surface, transforming it from a relatively homogeneous open water surface into a complex mixed underlying surface composed of seawater, broken ice floes, snow, and other elements.

[0014] This invention significantly expands the applicability of the COARE model by introducing sea ice concentration and establishing zoning correction relationships. The optimized model overcomes the original model's limitation to open ocean applications, enabling it to be accurately applied to polar marine environments with sea ice. Regardless of sea ice concentration (from open water to completely frozen areas), the model dynamically adjusts sea surface roughness through a correction function, thus providing flux estimation results that better reflect actual physical processes. Simultaneously, through a wind speed zoning strategy, the optimized model overcomes the original model's poor applicability under low wind speeds (≤3 m / s) and high wind speeds (>10 m / s), achieving stable and accurate calculations across a wider wind speed distribution range in polar regions.

[0015] Furthermore, the optimized model of this invention possesses adaptive capabilities. Users only need to run the optimized COARE model in the conventional manner, ensuring that the input real-time data includes sea ice concentration observations. The model will automatically determine the relevant interval based on the input wind speed and call the corresponding correction function to complete real-time roughness correction, requiring no manual intervention throughout the process. This optimized model is also applicable in traditional open oceans without sea ice (such as tropical and temperate zones) and can automatically degenerate into a mode equivalent to the original COARE model. When the input sea ice concentration is zero or close to zero, the output value of the correction function will also approach zero, making the amount of roughness correction to the model negligible. Therefore, the optimized model maintains the same accuracy as the original model in open oceans, achieving backward compatibility.

[0016] Therefore, this invention provides a more universal ocean-atmosphere flux calculation model. It is not a specialized model limited to polar regions, but rather an enhanced model with adaptive capabilities to sea ice environments. In polar regions, it provides high-precision results through dynamic correction; in the open ocean, it maintains the excellent performance of the original model. This provides a more reliable tool for climate research and prediction in the global ocean, especially in polar oceans where data is scarce and model biases are significant.

[0017] Based on the above scheme, the method of optimizing the COARE model based on sea ice concentration of the present invention can be further improved as follows.

[0018] In one alternative approach, it also includes:

[0019] The initial historical measured data were obtained based on observations conducted in the polar ocean during the summer season using an eddy covariance system carried by a research vessel at fixed intervals of 30 minutes. All data in the initial historical measured data are average values ​​within 30 minutes.

[0020] Based on the acquired initial historical measured data, outlier criteria based on statistical standard deviation are used to filter the data in the initial historical measured data to remove outliers, and the filtered data is then time-aligned to obtain the historical measured data.

[0021] The outlier criterion based on statistical standard deviation is as follows: For time series data arranged in chronological order in the initial historical measured data, calculate the overall standard deviation of the difference between adjacent points of each data point in the time series data. When the difference of a data point is greater than or equal to 3 times the overall standard deviation, the data point is marked as an outlier. However, when the differences of five consecutive adjacent data points all meet this condition, the five adjacent data points are not considered outliers and are not removed.

[0022] Among the above-mentioned optional methods, by explicitly limiting the historical measured data to observations by marine research vessels in polar regions during summer, and by using a specific quality control method that includes the 3σ principle and the continuous point criterion to process the 30-minute average data, the dataset used for modeling is ensured to have high spatiotemporal specificity and physical reliability. This reduces the interference of observation noise and gross errors on the subsequent model optimization process from the source, and lays a solid data foundation for establishing stable and accurate correction relationships.

[0023] In one alternative approach, historical measured data includes: wind speed data, friction velocity data, momentum flux data, sea ice concentration observation data, radiation data, sea surface temperature data, atmospheric pressure data, relative humidity data, and air temperature data.

[0024] Among the above-mentioned optional methods, by specifically listing all key meteorological and oceanographic parameters such as wind speed, friction speed, and sea ice concentration contained in historical measured data, the data elements required for model optimization are fully defined, ensuring that the dataset on which the technical solution depends can comprehensively characterize the actual state of polar air-sea interaction, enhancing the feasibility and reproducibility of the solution, and avoiding implementation difficulties caused by missing data or ambiguous definitions.

[0025] In one alternative approach, the steps of obtaining the true roughness and the model roughness separately include:

[0026] Based on wind profile theory, the true roughness is calculated by using wind speed data and friction speed data, combined with the measured height.

[0027] Wind speed data, radiation data, sea surface temperature data, atmospheric pressure data, relative humidity data, and air temperature data are input into the COARE model, and the sea surface roughness parameters calculated internally by the COARE model are used as the model roughness.

[0028] Among the above-mentioned alternative methods, by respectively clarifying the methods of calculating the true roughness based on wind profile theory using wind speed and friction speed, and driving the COARE model to output its internal roughness by inputting specific parameters, the sources and calculation paths of the two roughnesses are clearly defined, ensuring that the true roughness as the benchmark has a solid physical theoretical basis. At the same time, it exposes the inherent bias of the original model when sea ice is not considered, providing a scientific basis for subsequent quantification of errors and establishment of correction relationships.

[0029] In one optional approach, the preset wind speed zoning strategy is as follows:

[0030] Based on the shift in the dominant physical processes in sea ice-atmosphere interactions, wind speed data are divided into low wind speed range, medium wind speed range, and high wind speed range.

[0031] Among them, the low wind speed range corresponds to the wind speed conditions dominated by the stable stratification of the sea ice surface and static effects; the medium wind speed range corresponds to the wind speed conditions dominated by the dynamic drag of sea ice and the influence of partially broken ice surface; and the high wind speed range corresponds to the wind speed conditions dominated by the complex underlying surface effects caused by fully developed turbulence and large-scale sea ice breakage.

[0032] Among the above-mentioned alternative methods, wind speed zones are divided based on the transformation of the dominant physical processes in sea ice-atmosphere interactions, and different zones are associated with specific physical mechanisms (such as stable stratification, dynamic drag, and complex underlying surface effects). This gives wind speed zoning clear physical meaning, overcomes the mechanical nature of simple numerical division, and ensures that the sea ice influence correction functions established under different dynamic conditions are more physically representative and have higher predictive accuracy.

[0033] In one alternative approach, the step of establishing a correction function for correcting model roughness, which correlates the sea ice concentration observations of each data point in each data subset within each wind speed interval with the roughness difference values ​​of each data point, includes:

[0034] For data points within the same wind speed range, the observed sea ice concentration is used as the independent variable and the roughness difference is used as the dependent variable. A multinomial fitting method is used to establish a correction function between the observed sea ice concentration and the roughness difference for each data point.

[0035] Among the above-mentioned alternative methods, by using a polynomial fitting method to establish a functional relationship between sea ice concentration observations and roughness differences, it is possible to effectively capture and characterize the nonlinear variation characteristics that may exist between the two, thereby obtaining better fitting accuracy than a simple linear relationship, so that the established correction function can more precisely reflect the complex influence of sea ice on sea surface roughness.

[0036] In one alternative approach, the step of establishing the correction function further includes:

[0037] Using a pre-given validation dataset associated with polar oceans to validate the functional relationship, the determination coefficient of the modified function relative to the validation data is calculated, and the significance level value of the modified function is tested.

[0038] The modified function is deemed to have passed the verification when the coefficient of determination of the modified function meets the preset standard or the significance level value of the modified function meets the preset standard.

[0039] In the above-mentioned alternative methods, by adding a step of statistical verification based on an independent validation dataset and using the coefficient of determination and significance level, an objective and quantitative reliability evaluation standard is provided for the established correction function. This ensures that only robust functional relationships that have passed rigorous statistical tests will be ultimately adopted for model optimization, greatly improving the reliability and credibility of the optimization results.

[0040] In one alternative approach, real-time measured data of the polar ocean is input into the COARE model. During the operation of the COARE model, based on the wind speed range to which the wind speed data in the real-time measured data belongs, a correction function corresponding to that wind speed range is called. Combined with the observed sea ice concentration values ​​in the real-time measured data, the model roughness is corrected in real time to optimize the operation of the COARE model and obtain the output ocean-atmosphere flux that conforms to the polar sea ice cover conditions. The steps include:

[0041] Input real-time measured data into the COARE model; determine the wind speed range based on real-time wind speed data; call the correction function corresponding to the wind speed range, and use the current roughness difference value obtained by substituting the real-time sea ice concentration observation value into the correction function as the roughness correction amount;

[0042] The roughness correction amount is algebraically added to the model roughness calculated internally by the COARE model to obtain the corrected roughness;

[0043] Based on the corrected roughness, the flux calculation process of the COARE model is performed, and the ocean-atmosphere flux that meets the polar sea ice cover conditions is finally output.

[0044] In the above-mentioned optional methods, by describing in detail the complete process of inputting real-time data, calling the correction function to calculate the correction amount, performing algebraic addition correction on the roughness inside the model, and finally outputting the flux, the specific working steps and internal correction mechanism of the optimized COARE model during operation are clarified, which demonstrates the operability of the technical solution of the present invention and ensures that the correction of sea ice influence can be automatically and in real time integrated into the model operation process.

[0045] Secondly, this invention provides a system for optimizing the COARE model based on sea ice concentration. The technical solution of this system is as follows:

[0046] The system based on the COARE model optimized by sea ice concentration includes:

[0047] The difference quantification module is used to obtain the true roughness representing the actual sea surface condition for each observation time based on historical measured data collected in the polar ocean within a specified time period, and to input the meteorological parameters required by the standard COARE model from the historical measured data of that observation time into the COARE model to obtain the uncorrected model roughness calculated internally by the model, and to determine the roughness difference value based on the true roughness and the model roughness.

[0048] The function construction module is used to divide the wind speed data in the historical measured data into multiple wind speed intervals according to the preset wind speed partitioning strategy. Based on the sea ice concentration observation value in the historical measured data and the data subset in each wind speed interval, a correction function is established to associate the sea ice concentration observation value of each data point in the data subset with the roughness difference value of each data point to correct the roughness of the model.

[0049] The dynamic optimization module is used to input real-time measured data of polar oceans into the COARE model. During the operation of the COARE model, based on the wind speed range to which the wind speed data in the real-time measured data belongs, it calls the correction function corresponding to that wind speed range and combines the observed sea ice concentration values ​​in the real-time measured data to correct the roughness of the model in real time, so as to optimize the operation of the COARE model and obtain the output ocean-atmosphere flux that conforms to the polar sea ice coverage conditions.

[0050] The beneficial effects of the COARE model optimization system based on sea ice concentration of the present invention are as follows:

[0051] The system of this invention, by introducing sea ice concentration and establishing a zonal correction relationship, realizes the dynamic correction of roughness under polar sea ice conditions by the COARE model, breaks through the limitations of the original model in terms of wind speed applicability and polar adaptability, and significantly improves the accuracy and practicality of polar momentum flux calculation.

[0052] Based on the above scheme, the COARE model based on sea ice concentration optimization of the present invention can be further improved as follows.

[0053] In one alternative approach, it also includes:

[0054] The data preprocessing module is used to acquire initial historical measured data based on observations conducted in the polar ocean during the summer season using an eddy covariance system carried by a research vessel at fixed time intervals of 30 minutes. All data in the initial historical measured data are average values ​​within 30 minutes.

[0055] Based on the acquired initial historical measured data, outlier criteria based on statistical standard deviation are used to filter the data in the initial historical measured data to remove outliers, and the filtered data is then time-aligned to obtain the historical measured data.

[0056] The outlier criterion based on statistical standard deviation is as follows: For time series data arranged in chronological order in the initial historical measured data, calculate the overall standard deviation of the difference between adjacent points of each data point in the time series data. When the difference of a data point is greater than or equal to 3 times the overall standard deviation, the data point is marked as an outlier. However, when the differences of five consecutive adjacent data points all meet this condition, the five adjacent data points are not considered outliers and are not removed.

[0057] Thirdly, the technical solution of an electronic device according to the present invention is as follows:

[0058] The invention includes a memory, a processor, and a program stored in the memory and running on the processor, wherein the processor, when executing the program, implements the steps of the method for optimizing the COARE model based on sea ice concentration as described in this invention.

[0059] Fourthly, the technical solution of a computer-readable storage medium provided by the present invention is as follows:

[0060] The computer-readable storage medium stores instructions that, when read, cause the computer-readable storage medium to perform the steps of the method for optimizing the COARE model based on sea ice concentration according to the present invention.

[0061] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0062] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0063] Figure 1 This is a flowchart illustrating a method for optimizing the COARE model based on sea ice concentration according to the present invention.

[0064] Figure 2 A flowchart illustrating the process of establishing the correction function for this invention;

[0065] Figure 3 This is a schematic diagram of the real-time optimization process of the model of the present invention;

[0066] Figure 4 This is a schematic diagram of the overall workflow of the present invention;

[0067] Figure 5 This is a schematic diagram of the structure of a system based on the COARE model for optimizing sea ice concentration according to the present invention;

[0068] Figure 6 This is a schematic diagram of the structure of an electronic device according to the present invention. Detailed Implementation

[0069] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0070] Figure 1 This diagram illustrates a flowchart of an embodiment of a method for optimizing a COARE model based on sea ice concentration, provided by the present invention. This method is executed by a system that optimizes a COARE model based on sea ice concentration. Figure 1 As shown, it includes the following steps:

[0071] S1. Based on historical measured data collected in the polar ocean within a specified time period, for each observation time, obtain the true roughness representing the actual sea surface condition, and input the meteorological parameters required by the standard COARE model from the historical measured data of that observation time into the COARE model to obtain the uncorrected model roughness calculated internally by the model. Then, determine the roughness difference value based on the true roughness and the model roughness. In S1:

[0072] 1) Historical measured data refers to the raw data set including wind speed, friction speed, sea ice concentration and other relevant meteorological parameters collected in the polar ocean through observation platforms such as research vessels. Its purpose is to provide real environmental basic data support for the embodiments of the present invention.

[0073] 2) True roughness refers to a physical quantity that characterizes the true degree of sea surface undulation based on field measured data and physical theory. Its function is to serve as a benchmark value to measure the estimation error of the model.

[0074] 3) Model roughness refers to the sea surface roughness calculated by the unoptimized original COARE model according to its built-in parameterization scheme. Its purpose is to expose the systematic biases of the model when the influence of sea ice is not taken into account.

[0075] 4) Roughness difference value refers to the difference obtained by comparing the actual roughness with the model roughness. Its function is to quantify the estimation error of the COARE model of sea surface roughness under the current sea ice conditions, and to provide a data basis for establishing correction relationships in the future.

[0076] Specifically, the process begins by acquiring field-measured data of the polar ocean through ship observations and performing quality control on the data, including outlier removal using statistical methods and time alignment. Subsequently, based on logarithmic wind profile theory, the actual sea surface roughness is calculated using measured wind speed and friction velocity data. At the same time, relevant meteorological parameters are input into the standard COARE model to obtain its calculated model roughness. Finally, the difference between the actual roughness and the model roughness is calculated to quantify the model error.

[0077] S2. According to the preset wind speed zoning strategy, the wind speed data in the historical measured data is divided into multiple wind speed intervals. Based on the sea ice concentration observation values ​​in the historical measured data and the data subsets within each wind speed interval, a correction function is established to correlate the sea ice concentration observation values ​​of each data point in the data subset with the roughness difference values ​​of each data point, used to correct the model roughness. In S2:

[0078] 1) Wind speed zoning strategy refers to dividing continuous wind speed data into several intervals with clear thresholds based on the characteristics of polar ocean wind speed distribution and the applicability limitations of the COARE model. Its purpose is to distinguish the differences in the influence of sea ice on roughness under different wind speed conditions.

[0079] 2) The correction function is a mathematical expression established by mathematical fitting method that can characterize the quantitative relationship between the difference between sea ice concentration and roughness within a specific wind speed range. Its function is to quantify the influence of sea ice into a computable correction term that depends on wind speed conditions.

[0080] Specifically, firstly, all historical wind speed data are divided into three preset wind speed ranges: low, medium, and high. Then, for the data in each wind speed range, the observed sea ice concentration from the historical data is used as the independent variable and the roughness difference is used as the dependent variable. A multinomial fitting method is used to establish a functional relationship between the two. Finally, the functional relationship is statistically verified using an independent validation dataset to ensure that it meets the reliability requirements before it can be used for model optimization.

[0081] S3. Input real-time measured data of the polar ocean into the COARE model. During the operation of the COARE model, based on the wind speed range to which the wind speed data in the real-time measured data belongs, call the correction function corresponding to that wind speed range, and combine the observed sea ice concentration values ​​in the real-time measured data to perform real-time correction of the model roughness, so as to optimize the operation of the COARE model and obtain the output ocean-atmosphere flux that conforms to the polar sea ice coverage conditions. In S3:

[0082] 1) Model roughness correction refers to the process of algebraically adding the output value of the correction function to the roughness originally calculated by the COARE model. Its function is to dynamically integrate the influence of sea ice on roughness into the model in order to compensate for the systematic bias of the original model.

[0083] 2) Ocean-atmosphere flux that meets the conditions of polar sea ice cover refers to the momentum flux calculated using the corrected roughness. Its role is to provide a physical quantity that accurately reflects the ocean-atmosphere interaction under the condition of polar sea ice presence, and to provide more reliable input data for climate simulation and other studies.

[0084] Specifically, firstly, when running the COARE model, in addition to regular meteorological parameters, sea ice concentration data must be input simultaneously. Next, the system automatically determines the preset wind speed range (e.g., low, medium, and high wind speed ranges) based on the input wind speed value. Then, it calls a pre-established function relationship corresponding to that wind speed range to calculate the roughness correction based on the current sea ice concentration. Subsequently, this correction is applied to the roughness value calculated internally by the COARE model to obtain the optimized roughness. Finally, the model uses this optimized roughness for subsequent calculations, outputting a more accurate ocean-atmosphere momentum flux. This process enables dynamic and adaptive correction of the core parameters of the COARE model, making it accurately applicable to polar sea ice cover conditions.

[0085] The technical solution of this embodiment introduces sea ice concentration and establishes a zonal correction relationship, thereby realizing the dynamic correction of roughness under polar sea ice conditions by the COARE model. This breaks through the limitations of the original model in terms of wind speed applicability and polar adaptability, and significantly improves the accuracy and practicality of polar momentum flux calculation.

[0086] In one alternative approach, it also includes:

[0087] The initial historical measured data were obtained based on observations conducted in the polar ocean during the summer season using an eddy covariance system carried by a research vessel at fixed intervals of 30 minutes. All data in the initial historical measured data are average values ​​within 30 minutes.

[0088] Based on the acquired initial historical measured data, outlier criteria based on statistical standard deviation are used to filter the data in the initial historical measured data to remove outliers, and the filtered data is then time-aligned to obtain the historical measured data.

[0089] The outlier criterion based on statistical standard deviation is as follows: For time series data arranged in chronological order in the initial historical measured data, calculate the overall standard deviation of the difference between adjacent points of each data point in the time series data. When the difference of a data point is greater than or equal to 3 times the overall standard deviation, the data point is marked as an outlier. However, when the differences of five consecutive adjacent data points all meet this condition, the five adjacent data points are not considered outliers and are not removed.

[0090] It should be noted that the polar summer period was chosen because it is the time when sea ice melts significantly and sea ice concentration varies the most (from completely frozen to open water). Data obtained during this period can more comprehensively capture the impact of sea ice concentration from high to low on roughness under various wind speed conditions, thus establishing a more representative correction function.

[0091] The initial measured dataset refers to the set of raw observation values ​​obtained directly from the observation platform.

[0092] Outlier criteria based on statistical standard deviation refer to the criteria for identifying outlier data by using the statistical characteristics of a data series. The specific implementation method is as follows: calculate the overall standard deviation (σΔx) of the difference Δx between adjacent points of the time series x, and check each point one by one. If Δx ≥ 3σΔx at a certain point, it is an outlier. However, if 5 consecutive adjacent data points meet the criteria, it is considered not an outlier.

[0093] Outliers are data points that deviate significantly from the normal range due to observation errors or interference; time alignment refers to the process of adjusting observation data with different timestamps to a unified time reference; historical measured data refers to the set of data that can be used for subsequent calculations after quality control.

[0094] In this embodiment, an initial measured dataset is first obtained through observations using an eddy covariance system mounted on a research vessel. This dataset contains various data points for parameters such as wind speed, friction velocity, momentum flux, sea ice concentration, radiation, sea surface temperature, atmospheric pressure, relative humidity, and air temperature, with a time resolution of 30 minutes. All data points are averages over a 30-minute period. Taking horizontal wind speed data as an example, the specific implementation method for outlier removal is explained:

[0095] Suppose there is a continuous sequence of wind speed observations x1, x2, x3, ..., x n First, calculate the difference sequence between adjacent data points: Δx1 = x2 - x1, Δx2 = x3 - x2, ..., Δx n-1 =x n -x n-1 Then calculate the overall standard deviation σΔx of the difference sequence; then examine each Δx point by point. i Value, when Δxi When Δx ≥ 3σΔx, the point is marked as an outlier; however, if five consecutive adjacent Δx values ​​all satisfy Δx ≥ 3σΔx, these points are not considered outliers.

[0096] After outlier removal, the physical rationality of friction velocity and momentum flux was checked, limiting friction velocity to greater than 0.05 m / s and less than 1 m / s, and momentum flux to less than 1 N / m. 2 Finally, all data that passed the test were time-aligned, unifying the observed values ​​of each parameter to the same timestamp, forming the final historical measured data for use in subsequent steps.

[0097] The formula used in outlier detection is based on statistical principles. Δx represents the difference between adjacent data points; σΔx represents the overall standard deviation of the difference sequence; and 3σΔx constitutes the outlier detection threshold. This quality control method effectively identifies and eliminates abnormal observations caused by instrument malfunctions, environmental interference, and other factors, while avoiding misclassifying real drastic changes as outliers, thus ensuring the physical rationality and statistical reliability of the data. Time alignment ensures the consistency of different parameters over time, providing the necessary conditions for subsequent synchronous calculations.

[0098] This quality control process ensures the reliability and consistency of the data, providing a high-quality data foundation for subsequent calculations.

[0099] In one alternative approach, historical measured data includes: wind speed data, friction velocity data, momentum flux data, sea ice concentration observation data, radiation data, sea surface temperature data, atmospheric pressure data, relative humidity data, and air temperature data.

[0100] In this embodiment, the historical measured data used to optimize the COARE model originates from polar oceanographic expeditions and is acquired through observations conducted by the eddy covariance system and other supporting sensors aboard the research vessel. This dataset contains a series of key meteorological and oceanographic parameters:

[0101] Among them, wind speed data and friction velocity data are obtained directly from observations based on the eddy covariance method, serving as the core inputs for calculating the true sea surface roughness; momentum flux data serve as the verification benchmark for the model optimization effect; sea ice concentration observation data are obtained by measuring the proportion of sea ice within the field observation field, and are the core input parameters of this invention, used to quantify the impact of sea ice on roughness; radiation data, sea surface temperature data, atmospheric pressure data, relative humidity data, and air temperature data are the meteorological driving parameters necessary for running the original COARE model.

[0102] All data were observed at an altitude of 15 meters above the sea surface, with a time resolution of 30 minutes, ensuring sufficient temporal accuracy and physical consistency in characterizing the air-sea exchange process. This complete dataset constitutes a full-chain data support from model-driven approaches and core innovations to effect verification.

[0103] In one alternative approach, the steps of obtaining the true roughness and the model roughness separately include:

[0104] Based on wind profile theory, the true roughness is calculated by using wind speed data and friction speed data, combined with the measured height.

[0105] Wind speed data, radiation data, sea surface temperature data, atmospheric pressure data, relative humidity data, and air temperature data are input into the COARE model, and the sea surface roughness parameters calculated internally by the COARE model are used as the model roughness.

[0106] It should be noted that wind profiler theory is a fluid dynamics theory describing the variation of wind speed with altitude. It is applicable under near-neutral conditions. Since the heat flux in sea ice areas is not significant, generally ranging from -10 to 10 W / m², it can be approximated as near-neutral. True roughness refers to a physical quantity that characterizes the true degree of sea surface undulation, calculated based on field measurement data and wind profiler theory. Its function is to serve as a benchmark value to measure the estimation error of the model. Model roughness refers to the sea surface roughness calculated by the unoptimized original COARE model according to its built-in parameterization scheme. Measurement height refers to the vertical height of the wind speed sensor above the sea level.

[0107] In this embodiment, the process of calculating the actual roughness and the model roughness is as follows:

[0108] First, the required parameters are extracted from historical measured data that has undergone quality control. The calculation of the true roughness z0 is based on the logarithmic wind profile theory and is performed using the formula z0=z / exp(kU / u*).

[0109] Where z represents the measurement height, which is 15 meters; k is the Karman constant, which is a fixed value of 0.4; U is the average wind speed observed at height z; and u* is the friction speed obtained from synchronous observation.

[0110] As a calculation example, assuming a set of measured data is: U = 8 m / s, u* = 0.2 m / s, substituting into the formula, we get: z0 = 15 / exp(0.4 × 8 / 0.2) = 15 / exp(16) ≈ 15 / 8.886e6 ≈ 1.687e-6 meters. This calculation result represents the true roughness of the actual sea surface conditions.

[0111] Simultaneously, the original COARE model is run in parallel to obtain model roughness. Meteorological parameters defined by the standard COARE model, such as wind speed, radiation, sea surface temperature, atmospheric pressure, and air temperature data from historical measured data, are input into the model. The COARE model performs calculations based on its internally integrated parameterization scheme built upon wind profile theory and Obuhoff similarity theory, ultimately directly outputting its calculated model roughness z. c Through the above process, roughness z0 based on physical reality and roughness z based on model algorithm were obtained respectively. c This is to calculate the difference between the two values ​​(difference Δz = z0 - z) later. c We are ready.

[0112] It should be noted that in the corresponding field, the COARE model, through input variables, can directly output air-sea fluxes, such as sensible heat flux, latent heat flux, momentum flux, friction velocity, drag coefficient, and roughness length. Furthermore, the COARE model requires input data including wind speed, radiation, sea surface temperature, atmospheric pressure, relative humidity, and air temperature; and output data including sensible heat flux, latent heat flux, momentum flux, friction velocity, drag coefficient, and roughness length. However, this invention focuses on momentum flux because roughness length is a process quantity closely related to momentum flux.

[0113] In one optional approach, the preset wind speed zoning strategy is as follows:

[0114] Based on the shift in the dominant physical processes in sea ice-atmosphere interactions, wind speed data are divided into low wind speed range, medium wind speed range, and high wind speed range.

[0115] Among them, the low wind speed range corresponds to the wind speed conditions dominated by the stable stratification of the sea ice surface and static effects; the medium wind speed range corresponds to the wind speed conditions dominated by the dynamic drag of sea ice and the influence of partially broken ice surface; and the high wind speed range corresponds to the wind speed conditions dominated by the complex underlying surface effects caused by fully developed turbulence and large-scale sea ice breakage.

[0116] In one alternative approach, the step of establishing a correction function for correcting model roughness, which correlates the sea ice concentration observations of each data point in each data subset within each wind speed interval with the roughness difference values ​​of each data point, includes:

[0117] For data points within the same wind speed range, the observed sea ice concentration is used as the independent variable and the roughness difference is used as the dependent variable. A multinomial fitting method is used to establish a correction function between the observed sea ice concentration and the roughness difference for each data point.

[0118] In an alternative approach, the steps of establishing the correction function further include:

[0119] Using a pre-given verification data set associated with the polar ocean for verifying the functional relationship, calculating the coefficient of determination of the correction function with respect to the verification data, and testing the significance level value of the correction function;

[0120] When the coefficient of determination of the correction function meets the preset standard or the significance level value of the correction function meets the preset standard, it is determined that the correction function passes the verification.

[0121] It should be noted that the wind speed zoning strategy refers to the criterion for dividing continuous wind speed data into intervals according to the wind speed distribution characteristics of the polar ocean and the original applicable limitations of the COARE model; the role of the low wind speed interval is to separately analyze the influence mechanism of sea ice on roughness under low wind speed conditions; the medium wind speed interval corresponds to the original applicable wind speed range of the COARE model; the role of the high wind speed interval is to characterize the influence characteristics of sea ice under the unique high wind speed conditions in the polar region.

[0122] The polynomial fitting method refers to the mathematical method of using a polynomial function to perform curve fitting on discrete data points; the correction function refers to the mathematical expression characterizing the quantitative relationship between the sea ice concentration and the roughness difference value within a specific wind speed interval, and its role is to quantify the influence of sea ice into a computable correction term.

[0123] The pre-given verification data set associated with the polar ocean for verifying the functional relationship refers to the measured data set from different time or space ranges that did not participate in the fitting process of the aforementioned relationship function, and its role is to objectively evaluate the established functional relationship; the coefficient of determination (R 2 ) refers to the degree of explanation of the variability of the verification data by the established relationship function, and its value range is from 0 to 1. The closer the value is to 1, the better the fitting effect of the function on the data; the significance level value (p-value) is a probability index used to test whether the relationship function has statistical significance, and its role is to judge whether there is a significant systematic relationship rather than a random association between the sea ice concentration and the roughness difference value; the preset standard refers to the judgment threshold set based on actual application requirements, and its role is to provide an objective criterion for the reliability of the functional relationship.

[0124] In this embodiment, the implementation process of wind speed zoning and establishing the functional relationship is as follows:

[0125] First, divide the wind speed data in the obtained historical measured data containing the roughness difference value Δz and the corresponding sea ice concentration I according to the preset wind speed zoning strategy: when the wind speed U ≤ 3 m / s, the data is classified into the low wind speed interval; when 3 < U ≤ 10 m / s, the data is classified into the medium wind speed interval; when U > 10 m / s, the data is classified into the high wind speed interval.

[0126] The pre-defined wind speed zoning strategy is scientifically based on the transformation of the physical mechanisms of sea ice-atmosphere interaction under different wind speeds in polar oceans. According to the general distribution of the influence weights of different physical processes, the key physical processes dominating sea surface roughness changes significantly near the critical points of approximately 3 m / s and 10 m / s. Specifically, below 3 m / s, stable stratification and static forces on the sea ice surface dominate; in the medium wind speed range of 3 to 10 m / s, the dynamic drag effect of sea ice and the influence of partially broken ice become the controlling factors; when the wind speed exceeds 10 m / s, fully developed turbulence, splash, and the complex underlying surface effects caused by large-scale sea ice breakage play a dominant role. Therefore, zoning based on the critical wind speeds of this physical mechanism transformation ensures that the functional relationship between sea ice concentration and roughness differences established in each zone has clear physical meaning and the highest fitting accuracy, thus providing a reliable theoretical basis for the optimization of the COARE model.

[0127] After the division is completed, such as Figure 2 As shown, for each wind speed interval's data subset, to reduce data noise and ensure the robustness of the established functional relationship, data preprocessing is first performed: sea ice concentration I is grouped according to a preset step size (e.g., 0.1), and the arithmetic mean of all roughness difference values ​​Δz within each group is calculated, thus forming a series of new data points with the average sea ice concentration as the independent variable and the average roughness difference as the dependent variable. To ensure statistical significance of the fit, the total number of new data points obtained in each wind speed interval must be no less than 5. For each preprocessed data subset within each wind speed interval, a polynomial fitting method is used, with sea ice concentration I as the independent variable and roughness difference value Δz as the dependent variable. The polynomial fitting method is implemented using the least squares method.

[0128] For the low and medium wind speed ranges, a quadratic polynomial function is used for fitting, with the function form being f(I) = c × I. 2 +d×I+e;

[0129] For the high wind speed range, a linear function is used for fitting, with the function form being f(I)=a×I+b.

[0130] Where I represents sea ice concentration, and a, b, c, d, and e are polynomial coefficients obtained by fitting using the least squares method. For example, after fitting a dataset of a certain medium wind speed range, the specific function f(I) = -0.23 × I is obtained. 2 +0.21×I+0.17, this function establishes a quantitative relationship between the difference in sea ice concentration and roughness within this wind speed range.

[0131] By using the above-mentioned partitioning fitting method, we can accurately characterize the differences in the influence of sea ice concentration on sea surface roughness under different wind speed conditions, laying the foundation for the accurate optimization of the COARE model.

[0132] After fitting, the coefficient of determination R of the obtained function is calculated based on the validation dataset. 2 and significance level value, when R satisfies 2 When the preset standard of p > 0.6 or p < 0.05 is met, the correction function is deemed to have passed verification and can be used in subsequent model optimization steps. Further:

[0133] First, an independent validation dataset was obtained. The sea ice concentration in this dataset can be obtained from the daily average data of NSIDC's passive microwave remote sensing products, while the roughness difference values ​​are obtained from field observations from other expeditions or at different times.

[0134] Substituting the sea ice concentration values ​​from the validation dataset into the established correction function, the predicted values ​​of roughness difference are calculated, and then the coefficient of determination R is calculated using statistical methods. 2 And the significance level p-value.

[0135] Coefficient of determination R 2 The calculation formula is R 2 =1-(SS res / SS tot ), where SS res Sum of squared residuals, representing the sum of squares of the differences between predicted and observed values; SS tot The total sum of squares represents the sum of squares of the differences between each actual observation and the mean of all actual observations.

[0136] The significance level (p-value) is typically obtained using an F-test or a t-test. The F-test primarily assesses whether the entire modified function is statistically significant, while the t-test assesses whether the coefficients of specific independent variables (such as sea ice concentration or its higher-order terms) within the function are significantly different from zero. Both tests aim to determine whether the established functional relationship significantly deviates from the null hypothesis (i.e., there is no linear relationship between all independent and dependent variables, or all coefficients are zero).

[0137] Through rigorous statistical verification, it was ensured that the established correction function could accurately reflect the impact of sea ice concentration on roughness differences, providing a reliable scientific basis for the optimization of the COARE model.

[0138] The preset standard set in this embodiment is: when the determination coefficient R 2 When the significance level p-value is greater than 0.6, the p-value does not need to be considered, and the correction function is directly deemed to have passed the validation; or, when the significance level p-value is less than 0.05, R² does not need to be considered. 2 The correction function was directly determined to pass the verification.

[0139] Among these, because when the significance level p-value < 0.05, even if the coefficient of determination R... 2 Even at a low level (e.g., as low as 0.1), the functional relationship is considered statistically significant, and therefore the verification is passed.

[0140] For example, the function f(I) = -0.02 × I is established for a certain medium wind speed range. 2 Validation is performed using +0.03×I+0.005, and the calculation is performed using an independent validation set containing 20 data sets. If R is obtained... 2 =0.65, then R is satisfied. 2 A standard of >0.6 can be used to directly determine that the function has passed the verification; if another function passes the verification, R... 2 =0.15 (this is a relatively low value), but p=0.03. Since p<0.05, the function is still considered to have passed the validation. This dual-standard validation mechanism ensures that the functional relationship of the modified function can be adopted when the interpretability is high, and can also be reasonably applied in complex physical scenarios where the interpretability is statistically significant but limited. This ensures that the functional relationship on which the subsequent model optimization steps depend has the necessary reliability and robustness.

[0141] In one alternative approach, real-time measured data of the polar ocean is input into the COARE model. During the operation of the COARE model, based on the wind speed range to which the wind speed data in the real-time measured data belongs, a correction function corresponding to that wind speed range is called. Combined with the observed sea ice concentration values ​​in the real-time measured data, the model roughness is corrected in real time to optimize the operation of the COARE model and obtain the output ocean-atmosphere flux that conforms to the polar sea ice cover conditions. The steps include:

[0142] Input real-time measured data into the COARE model; determine the wind speed range based on real-time wind speed data; call the correction function corresponding to the wind speed range, and use the current roughness difference value obtained by substituting the real-time sea ice concentration observation value into the correction function as the roughness correction amount;

[0143] The roughness correction amount is algebraically added to the model roughness calculated internally by the COARE model to obtain the corrected roughness;

[0144] Based on the corrected roughness, the flux calculation process of the COARE model is performed, and the ocean-atmosphere flux that meets the polar sea ice cover conditions is finally output.

[0145] It should be noted that algebraic addition refers to the mathematical operation of performing arithmetic addition on two values; the corrected roughness refers to the new roughness value after correction for the influence of sea ice concentration, and its purpose is to make the roughness calculated by the model closer to the actual situation under polar sea ice cover conditions.

[0146] Flux calculation process refers to the built-in algorithm flow in the COARE model that calculates ocean-atmosphere exchange flux based on input parameters; real-time correction refers to the roughness correction operation performed dynamically within each time step of the model operation; ocean-atmosphere flux that is more consistent with polar sea ice cover conditions refers to the physical quantity output that can accurately reflect the impact of sea ice presence on ocean-atmosphere exchange.

[0147] In this embodiment, the model correction and optimization process is as follows:

[0148] First, during the operation of the COARE model, such as Figure 3 As shown, when a set of data including wind speed U, sea ice concentration I, and other meteorological parameters defined by the standard COARE model are input, the model first automatically determines the wind speed range to which it belongs based on the input wind speed U, and then calls the validated correction function f(I) corresponding to that wind speed range. Substituting the current sea ice concentration I value into this function, the roughness correction amount f(I) is calculated.

[0149] Next, according to formula z c' =z c +f(I) corrects the roughness of the model calculated internally, where z c' Represents the corrected roughness; z c f(I) represents the roughness of the original output of the COARE model; f(I) represents the output value of the correction function, which is the correction amount calculated based on the current sea ice concentration I. By using an algebraic addition correction method, the sea ice influence is effectively integrated into the COARE model with minimal computational complexity, achieving a substantial improvement in model performance.

[0150] For example, suppose the current wind speed U = 12 m / s (belonging to the high wind speed range), and the sea ice concentration I = 0.7. Using the correction function for the high wind speed range f(I) = 0.1 × I + 0.02, we calculate f(0.7) = 0.1 × 0.7 + 0.02 = 0.09. If the roughness z calculated internally by the COARE model at this time... c =0.0012 meters, then the corrected roughness z c' =0.0012+0.09=0.0912 meters.

[0151] Then, the COARE model will use this modified roughness z c' Replace the original z c It continues its internal flux calculation process and finally outputs the optimized momentum flux.

[0152] This correction process is automatically executed at each time step of the COARE model's operation, achieving real-time dynamic correction of roughness. This allows the COARE model to fully consider the impact of sea ice concentration and output ocean-atmosphere fluxes that are more consistent with the actual conditions of sea ice cover in polar regions. This optimizes the COARE model, significantly improving its applicability and computational accuracy in polar oceans, and providing more accurate data support for polar ocean climate research.

[0153] like Figure 4 As shown, the method of the present invention mainly includes two interconnected stages:

[0154] The first phase is the modeling phase, which begins with acquiring historical measured data collected in the polar ocean within a specified time period. Subsequently, for each observation time, the true roughness representing the actual sea surface conditions is calculated in parallel. Meteorological parameters defined by the standard COARE model from the historical measured data are input into the COARE model to obtain its internally calculated, uncorrected model roughness, thereby determining the roughness difference value. Afterward, according to a pre-defined wind speed zoning strategy, the wind speed data from the historical measured data is divided into multiple wind speed intervals. For a subset of data within each wind speed interval, combined with sea ice concentration observations, a correction function is established to correlate sea ice concentration observations with the roughness difference value, used to correct the model roughness.

[0155] The second stage is the application stage: real-time measured data corresponding to historical measured data is input into the COARE model. First, the wind speed interval to which the data belongs is determined based on the real-time wind speed data, and a pre-established correction function corresponding to that interval is invoked. Next, based on real-time sea ice concentration observations, the current roughness difference value is determined as the roughness correction amount, and the model roughness calculated internally by the COARE model is corrected in real time. Finally, by optimizing the operation of the COARE model, the output ocean-atmosphere flux that conforms to the polar sea ice cover conditions is obtained.

[0156] Figure 5 A schematic diagram of an embodiment of a system 200 based on a sea ice concentration optimization COARE model provided by the present invention is shown. Figure 5 As shown, the system 200 includes:

[0157] The difference quantification module 210 is used to obtain the true roughness representing the actual sea surface condition for each observation time based on historical measured data collected in the polar ocean within a specified time period, and to input the meteorological parameters required by the standard COARE model from the historical measured data of the observation time into the COARE model to obtain the uncorrected model roughness calculated internally by the model, and to determine the roughness difference value based on the true roughness and the model roughness.

[0158] The function construction module 220 is used to divide the wind speed data in the historical measured data into multiple wind speed intervals according to the preset wind speed partitioning strategy. Based on the sea ice concentration observation value in the historical measured data and the data subset in each wind speed interval, a correction function is established to associate the sea ice concentration observation value of each data point in the data subset with the roughness difference value of each data point to correct the roughness of the model.

[0159] The dynamic optimization module 230 is used to input real-time measured data of polar oceans into the COARE model. During the operation of the COARE model, based on the wind speed range to which the wind speed data in the real-time measured data belongs, it calls the correction function corresponding to the wind speed range and combines the sea ice concentration observation value in the real-time measured data to correct the roughness of the model in real time, so as to optimize the operation of the COARE model and obtain the output ocean-atmosphere flux that meets the polar sea ice coverage conditions.

[0160] The technical solution of this embodiment introduces sea ice concentration and establishes a zonal correction relationship, thereby realizing the dynamic correction of roughness under polar sea ice conditions by the COARE model. This breaks through the limitations of the original model in terms of wind speed applicability and polar adaptability, and significantly improves the accuracy and practicality of polar momentum flux calculation.

[0161] In one alternative approach, the data preprocessing module is also included to acquire initial historical measured data based on observations conducted in the polar ocean during the summer season using an eddy covariance system carried by a research vessel at fixed time intervals of 30 minutes. All data in the initial historical measured data are average values ​​over 30 minutes.

[0162] Based on the acquired initial historical measured data, outlier criteria based on statistical standard deviation are used to filter the data in the initial historical measured data to remove outliers, and the filtered data is then time-aligned to obtain the historical measured data.

[0163] The outlier criterion based on statistical standard deviation is as follows: For time series data arranged in chronological order in the initial historical measured data, calculate the overall standard deviation of the difference between adjacent points of each data point in the time series data. When the difference of a data point is greater than or equal to 3 times the overall standard deviation, the data point is marked as an outlier. However, when the differences of five consecutive adjacent data points all meet this condition, the five adjacent data points are not considered outliers and are not removed.

[0164] The parameters and steps for implementing the corresponding functions of each module in the system 200 based on the COARE model for sea ice concentration optimization described above can be found in the embodiments of the method for optimizing the COARE model based on sea ice concentration described above, and will not be repeated here.

[0165] like Figure 6 As shown, an electronic device 300 according to an embodiment of the present invention includes a processor 320 coupled to a memory 310. The memory 310 stores at least one computer program 330, which is loaded and executed by the processor 320 to enable the electronic device 300 to implement any of the above-mentioned methods for optimizing the COARE model based on sea ice concentration. Specifically:

[0166] The electronic device 300 can vary considerably due to differences in configuration or performance. It may include one or more processors 320 (Central Processing Units, CPUs) and one or more memories 310, wherein the one or more memories 310 store at least one computer program 330, which is loaded and executed by the one or more processors 320 to enable the electronic device 300 to implement any of the methods for optimizing the COARE model based on sea ice concentration provided in the above embodiments. Of course, the electronic device 300 may also have wired or wireless network interfaces, a keyboard, and input / output interfaces for input and output. The electronic device 300 may also include other components for implementing device functions, which will not be elaborated here.

[0167] An embodiment of the present invention provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to enable a computer to implement any of the above-described methods for optimizing the COARE model based on sea ice concentration.

[0168] Alternatively, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, a floppy disk, and an optical data storage device, etc.

[0169] In an exemplary embodiment, a computer program product or computer program is also provided, comprising computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform any of the methods described above for optimizing the COARE model based on sea ice concentration.

[0170] It should be noted that the terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and represent a limitation on a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of this application described herein can be implemented in an order other than that shown or described.

[0171] Those skilled in the art will recognize that this invention can be implemented as a system, method, or computer program product. Therefore, this disclosure can be specifically implemented in the following forms: it can be entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, the invention can also be implemented as a computer program product contained in one or more computer-readable media, which includes computer-readable program code.

[0172] Any combination of one or more computer-readable media can be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0173] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for optimizing a COARE model based on sea ice concentration, characterized in that, The method comprises the following steps: Based on the historical measured data collected in the polar ocean within a specified time period, for each observation time, the true roughness representing the actual sea surface condition is obtained, and the meteorological parameter required by the standard COARE model is input into the COARE model in the observation time historical measured data, and the uncorrected model roughness generated by the internal calculation of the model is obtained, and the roughness difference value is determined according to the true roughness and the model roughness; According to the preset wind speed partition strategy, the wind speed data in the historical measured data is divided into multiple wind speed intervals, and based on the sea ice density observation value in the historical measured data and the data subset in each wind speed interval, a correction function for correcting the model roughness is established, which is related to the sea ice density observation value of each data point and the roughness difference value of each data point in the data subset; The real-time measured data of the polar ocean is input into the COARE model, and during the running of the COARE model, the correction function corresponding to the wind speed interval to which the wind speed data in the real-time measured data belongs is called based on the wind speed interval, and the model roughness is corrected in real time combined with the sea ice density observation value in the real-time measured data, so as to realize the optimization of the running of the COARE model and obtain the output of the ocean-atmosphere flux conforming to the polar sea ice coverage condition.

2. The method of optimizing COARE model based on sea ice concentration according to claim 1, characterized in that, The method further comprises the following steps: Obtain initial historical measured data collected by the eddy correlation system carried by the research ship in the polar ocean within a fixed time interval of 30 minutes during the summer season in the polar region, wherein all data in the initial historical measured data are average values within 30 minutes; Based on the obtained initial historical measured data, an outlier criterion based on statistical standard deviation is used to filter out outliers from the data in the initial historical measured data, and the filtered data is time-aligned to obtain historical measured data; The outlier criterion based on statistical standard deviation is that for the time series data arranged in time sequence in the initial historical measured data, the overall standard deviation of the difference value of each data point in the time series data is calculated, and when the difference value of a data point is greater than or equal to 3 times the overall standard deviation, the data point is marked as an outlier, but when the difference values corresponding to five consecutive adjacent data points all satisfy this condition, the five adjacent data points are not considered as outliers and are not removed.

3. The method of optimizing COARE model based on sea ice concentration according to claim 1, characterized in that, The historical measured data includes wind speed data, friction velocity data, momentum flux data, sea ice density observation value data, radiation data, sea surface temperature data, atmospheric pressure data, air relative humidity data, and air temperature data.

4. The method of optimizing COARE model based on sea ice concentration according to claim 3, characterized in that, The steps of obtaining the true roughness and the model roughness respectively comprise: Based on the wind profile theory, the true roughness is calculated by using the wind speed data and the friction velocity data combined with the measurement height; The wind speed data, radiation data, sea surface temperature data, atmospheric pressure data, air relative humidity data, and air temperature data are input into the COARE model, and the sea surface roughness parameter calculated internally by the COARE model is output as the model roughness.

5. The method of optimizing COARE model based on sea ice concentration according to claim 1, wherein, The preset wind speed partition strategy is: According to the transition of the dominant physical process in sea ice-atmosphere interaction, the wind speed data is divided into a low wind speed interval, a medium wind speed interval and a high wind speed interval; Among them, the low wind speed interval corresponds to the wind speed condition dominated by the stable stratification of the sea ice surface and the static force; the medium wind speed interval corresponds to the wind speed condition dominated by the sea ice dynamics drag and the influence of the partially broken ice surface; the high wind speed interval corresponds to the wind speed condition dominated by the fully developed turbulence and the complex underlying surface effect caused by the large area of sea ice breaking.

6. The method of optimizing COARE model based on sea ice concentration according to claim 3, characterized in that, For each data subset in each wind speed interval, the step of establishing a correction function for correcting the model roughness associated with the sea ice density observation value of each data point in the data subset and the roughness difference value of each data point includes: Taking the sea ice density observation value as the independent variable and the roughness difference value as the dependent variable, the data points in the same wind speed interval are established; a polynomial fitting method is used to establish the correction function between the sea ice density observation value and the roughness difference value of each data point.

7. The method of optimizing COARE model based on sea ice concentration according to claim 6, characterized in that, The step of establishing the correction function further includes: Using the pre-defined validation data set associated with the polar ocean for verifying the functional relationship, the determination coefficient of the correction function relative to the validation data is calculated, and the significance level value of the correction function is tested; When the determination coefficient of the correction function meets the preset standard or the significance level value of the correction function meets the preset standard, it is determined that the correction function passes the verification.

8. The method of optimizing COARE model based on sea ice concentration according to claim 1, wherein, The real-time measured data of the polar ocean is input into the COARE model, and in the running of the COARE model, based on the wind speed interval to which the wind speed data in the real-time measured data belongs, the correction function corresponding to the wind speed interval is called, and the model roughness is corrected in real time by combining the sea ice density observation value in the real-time measured data, so as to realize the optimization of the running of the COARE model and obtain the output of the ocean-atmosphere flux meeting the polar sea ice coverage condition, including: The real-time measured data is input into the COARE model; based on the real-time wind speed data, the wind speed interval is determined; the correction function corresponding to the wind speed interval is called, and the current roughness difference value obtained by substituting the real-time sea ice density observation value into the correction function is used as the roughness correction amount; The roughness correction amount and the model roughness generated by the internal calculation of the COARE model are algebraically added to obtain the corrected roughness; Based on the corrected roughness, the flux calculation process of the COARE model is executed, and finally the ocean-atmosphere flux meeting the polar sea ice coverage condition is output.

9. A system for optimizing COARE model based on sea ice concentration, characterized in that, It includes: The difference quantization module is used to obtain the real roughness representing the actual sea surface condition for each observation time based on the historical measured data collected in the polar ocean within a specified time period, and input the meteorological parameters required by the standard COARE model into the COARE model to obtain the uncorrected model roughness generated by the internal calculation of the model, and determine the roughness difference value according to the real roughness and the model roughness; The function construction module is configured to divide wind speed data in the historical measured data into multiple wind speed intervals according to a preset wind speed partition strategy, and to establish a correction function for correcting the model roughness based on the sea ice density observation values in the historical measured data and a data subset in each wind speed interval, the correction function relating the sea ice density observation values of each data point in the data subset to the roughness difference values of the data points. The dynamic optimization module is configured to input real-time measured data of the polar ocean into the COARE model, to call a correction function corresponding to a wind speed interval to which wind speed data in the real-time measured data belong during running of the COARE model, and to combine the sea ice density observation values in the real-time measured data to correct the model roughness in real time, so as to optimize the running of the COARE model and obtain output sea-ocean fluxes that meet the polar sea ice coverage condition.

10. The system for optimizing COARE model based on sea ice concentration according to claim 9, wherein, The data preprocessing module is configured to obtain initial historical measured data collected by an eddy correlation system carried by an expedition ship in the polar ocean at a fixed time interval of 30 minutes during a summer period in the polar region, wherein all data in the initial historical measured data are average values within 30 minutes. Based on the obtained initial historical measured data, an abnormal value criterion based on a statistical standard deviation is used to filter out abnormal values from the data in the initial historical measured data, and the filtered data is subjected to time alignment processing to obtain the historical measured data. The abnormal value criterion based on the statistical standard deviation is that, for time series data arranged in time sequence in the initial historical measured data, the overall standard deviation of difference values of adjacent points of each data point in the time series data is calculated, and when the difference value of a data point is greater than or equal to 3 times the overall standard deviation, the data point is marked as an abnormal value, but when the difference values corresponding to five consecutive adjacent data points all satisfy this condition, the five consecutive adjacent data points are not regarded as abnormal values to be removed. ​

Images