Shipborne meteorological and hydrological instrument data and numerical mode initial field data maneuvering fusion method

Abstract

The invention discloses a mobile fusion method for shipborne meteorological and hydrological instrument data and numerical model initial field data, belongs to the technical field of marine meteorological observation, numerical weather forecast and data fusion processing, and realizes mobile observation of meteorological basic elements such as wind, temperature, humidity, air pressure and the like based on mobile navigation of a ship. Point observation of the meteorological and hydrological instrument is converted into line observation; a space-time matching algorithm is adopted to realize time and space consistency matching of three kinds of data including a shore-based automatic meteorological station, an offshore meteorological hydrological instrument and numerical mode initial field data; dynamically adjusting the route according to the ship motion attitude; motorized fusion is carried out by adopting an LAPS numerical mode and an optimal interpolation method, so that line data can be converted into three-dimensional lattice point data; and carrying out real-time maneuvering correction on multi-source fusion data realized in the LAPS numerical mode by adopting a meteorological and hydrological instrument for mobile observation on the sea. By means of the mode, the cognitive precision of the temperature, humidity and wind under the marine meteorological conditions is improved.

Description

Technical Field

[0001] The present invention relates to the field of marine meteorological observation, numerical weather forecast and data fusion processing technology, and in particular to a method for maneuvering fusion of shipborne meteorological and hydrological instrument data and initial field data of numerical models. Background Art

[0002] Ship-borne meteorological and hydrological instruments are important tools for marine meteorological observations, which can obtain sea surface meteorological and hydrological parameters in real time, such as wind speed, wind direction, air temperature, sea surface temperature, salinity, etc. Automatic meteorological observation stations are often installed on the ground to measure various meteorological elements, such as temperature, humidity, air pressure, wind speed, wind direction, precipitation, etc. The initial field data of the numerical model is three-dimensional grid data, which is the basic data of numerical weather forecasting and provides an initial description of the state of the atmosphere and the ocean. The effective integration of these three can improve the accuracy of the understanding of marine meteorological conditions, optimize the weather forecast model, and thus improve the accuracy of the forecast.

[0003] Whether it is a ship-borne meteorological and hydrological instrument or a ground-based automatic meteorological observation station, all of them are single-station data, that is, "point" data. Interpolating "point" data into three-dimensional grid data is a technology widely used in meteorology, geographic information systems (GIS), environmental science and other fields. This technology allows researchers to estimate unknown data values ​​in the entire three-dimensional space based on known discrete point data (usually irregularly distributed), thereby generating a more detailed and continuous spatial data set.

[0004] At present, the commonly used interpolation methods include linear interpolation, spline interpolation, Kriging interpolation and cubic interpolation, etc. These methods have their own characteristics and are suitable for different data distributions and accuracy requirements.

[0005] (1) Linear interpolation: Estimate the value of unknown points based on the linear relationship between known points. This method is simple and intuitive, but may not work well when the data distribution is complex or the nonlinear relationship is strong.

[0006] (2) Spline interpolation: By constructing a series of polynomial curves to approximate known data points, the value of unknown points can be estimated. Spline interpolation can fit complex data distributions well, but the amount of calculation is relatively large.

[0007] (3) Kriging interpolation: A geostatistical method that takes into account the autocorrelation and heterogeneity of spatial data. Kriging interpolation is widely used in geology, environment and other fields, and can generate spatial data distribution that is more in line with reality.

[0008] (4) Cubic interpolation: Interpolation based on cubic polynomials can generate a smooth surface. This method works better when the data points are evenly distributed and smooth spatial data needs to be generated.

[0009] (5) Fusion based on numerical models such as LAPS, GSI, and MM5. This method integrates interpolation algorithms, energy conservation equations, momentum conservation equations, mass conservation equations, and water vapor conservation equations. The fusion results are often closer to the real atmospheric environment.

[0010] Existing fusion methods often lack flexibility and can only integrate data from meteorological and hydrological instruments on fixed platforms. They cannot be dynamically adjusted according to the weather forecast requirements of changes in the routes and speeds of ships on mobile platforms.

[0011] Based on this, the present invention designs a method for the mobile fusion of shipborne meteorological and hydrological instrument data and initial field data of numerical models to solve the above problems. Summary of the invention

[0012] In view of the above-mentioned shortcomings of the prior art, the present invention provides a method for maneuverably fusing shipborne meteorological and hydrological instrument data with initial field data of numerical models.

[0013] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0014] The method for mobile fusion of shipborne meteorological and hydrological instrument data and initial field data of numerical models includes the following steps:

[0015] Step 1: Make wind direction and wind speed corrections to the shipboard meteorological and hydrological instrument data according to the ship's route speed and attitude;

[0016] Step 2: Perform quality control on the data from the automatic weather station and the ship-borne meteorological and hydrological instruments, remove the data that exceed the climatological limit values ​​and the historical variation range, and process the missing and invalid data to make the data continuous; then perform a time consistency check and an internal consistency check on the data to complete the control of data quality;

[0017] Step 3: Match the observation data of the ship-borne meteorological and hydrological instrument with the data of multiple shore-based automatic weather stations in time and space to ensure the consistency of the two in time and space, and complete the matching and calibration of the observation data of the ship-borne meteorological and hydrological instrument with the data of multiple shore-based automatic weather stations;

[0018] Step 4: The data that have been time-space matched are integrated with the observation data of the ship-borne meteorological and hydrological instruments and the data of multiple shore-based automatic weather stations through the ground analysis module in LAPS (Local Data Prediction and Analysis System), and the data are output through two-dimensional variational analysis;

[0019] Step 5: The LAPS-fused data is corrected in real time using the linear quadratic estimation algorithm and then output in NETCDF format for use in the numerical weather forecast model.

[0020] Furthermore, the wind direction correction uses an inclination sensor to monitor the sensor inclination in real time and correct it through an algorithm; the wind speed correction uses a motion compensation algorithm or an intelligent correction algorithm.

[0021] Furthermore, missing and invalid data are processed by interpolation, mean substitution or model prediction.

[0022] Furthermore, the internal consistency check refers to the check that the relationship between the records of meteorological elements observed at the same time should conform to a certain physical connection, that is, the internal consistency check; it includes internal consistency checks for the same type of elements and internal consistency checks for elements of different types; among which the internal consistency check for the same type of elements is generally a logical check, such as the on-time value ≥ the minimum value within the hour, the on-time value ≤ the maximum value within the hour, and the hourly cumulative amount = the sum of the amounts per minute within the hour.

[0023] Furthermore, time-space matching includes time matching and space matching; time matching refers to the sampling time t of the shipborne meteorological and hydrological instrument. BA As a benchmark, synchronously take the current distance t BA The most recent time t A The automatic weather station data at the moment; the spatial matching refers to using the real-time updated longitude and latitude information (Lon_p, Lat_p) in the ship's navigation path as the longitude and latitude of the ship's meteorological and hydrographic instruments.

[0024] Furthermore, the ground analysis module uses data from automatic weather stations and meteorological and hydrological instruments to generate the spatial distribution field of ground meteorological elements through a series of interpolation, smoothing and analysis techniques, providing a spatial resolution of less than several kilometers and a temporal resolution of hours.

[0025] Furthermore, the algorithm of the ground analysis module is as follows:

[0026] (1) The data from the numerical model are interpolated onto the grid of LAPS output points using bilinear interpolation.

[0027] (2) Obtain the initial field based on the distance weight And the error value Then the initial field is corrected to obtain the analytical field:

[0028]

[0029] In the formula, B (xk,yk) is the kth automatic station data, is the analysis field, xk, yk are the longitude and latitude of the kth automatic station, w k is the weight parameter of the kth automatic station, w' k is the modified weight parameter of the kth automatic station, and c is the Barness filter parameter.

[0030] Furthermore, the linear quadratic estimation algorithm refers to the use of a series of measurements observed over time, including statistical noise and various other noises that cause observation errors, to generate estimates of unknown variables through a joint probability density distribution; for the prediction stage, an estimate of the current state variable and its uncertainty matrix are generated; these estimates are updated using a weighted average after observing the results of the next measurement, and more certain estimates are given more weight.

[0031] Furthermore, the basic formula of the linear quadratic estimation algorithm is:

[0032] β t =Φ t-1 β t-1 +ω t-1

[0033] Y t =X t β t +ν t

[0034] The above two formulas are the state equation and the measurement equation respectively, where: is the regression coefficient, is the state vector in the correction system; ω is the input noise; φ is the state transfer matrix; Y t represents the state measurement vector, X t is the prediction factor; ν is the state measurement noise.

[0035] Furthermore, the parameters in the basic formula of the linear quadratic estimation algorithm are corrected in a rolling manner, using the updated data to correct the state estimate, and continuously updating the measured information during the recursive process; the specific formula is as follows:

[0036] β t =β t-1 +ν t-1

[0037]

[0038] In the recursive system composed of the above six formulas, Y represents the observed value at time t, represents the forecast value at time t, represents the estimated value of the regression coefficient, R t express The error variance matrix of the extrapolated values, X t represents the predictor, is the predictor factor X t The transformed rank matrix, C t-1 express The error variance matrix of the filtered value, σt is the forecast error variance matrix, A t is the gain matrix, which is σ t The inverse matrix of .

[0039] Compared with the prior art, the present invention has the following beneficial effects: 1. The observation data of the meteorological and hydrological instrument is point data, that is, it is only the basic meteorological elements wind, temperature, humidity, air pressure, and salinity information at one observation point. By using a ship equipped with a meteorological and hydrological instrument for observation, the "point" data can be converted into "line" data; based on the mobile navigation of the ship, the mobile observation of the basic meteorological elements wind, temperature, humidity, air pressure, etc. is realized, thereby converting the "point" observation of the meteorological and hydrological instrument into "line" observation; the time-space matching algorithm is adopted to realize the time and space consistency matching of the three types of data, namely, the shore-based automatic weather station, the offshore meteorological and hydrological instrument, and the initial field data of the numerical model (global analysis field GFS); according to the ship's motion posture and the dynamic adjustment of the route, the problem of mobile fusion of the meteorological and hydrological instrument data on the moving platform ship is solved; and the "LAPS numerical model + optimal interpolation method" is used for mobile fusion to convert the "line" data into three-dimensional grid data, thereby solving the problem of small spatial coverage of the meteorological and hydrological instrument data;

[0040] 2. Use meteorological and hydrological instruments for mobile observation at sea to make real-time mobile corrections to the multi-source fusion data realized by the LAPS (Regional Analysis and Forecasting System) numerical model to improve the accuracy of the fusion data;

[0041] 3. By integrating real-time shipborne observation data and initial field data from numerical models, the accuracy of understanding of ocean meteorological conditions such as temperature, humidity and wind can be improved, thereby improving the accuracy of weather forecasts.

[0042] 4. Mobile fusion: According to the maneuverability of the unmanned boat, the time and space matching of meteorological and hydrological instruments is carried out in real time, and the meteorological and hydrological data after time and space matching are integrated into the initial field data of the numerical model to realize mobile fusion and mobile real-time correction.

[0043] 5. The present invention can be applied not only to the field of numerical weather forecasting, but also to the fields of marine environment monitoring, weather change research, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1It is a flow chart of the method for mobile fusion of shipborne meteorological and hydrological instrument data and initial field data of numerical model of the present invention;

[0046] Figure 2 Schematic diagram of the spatiotemporal matching between shipborne meteorological and hydrological instruments and other data.

[0047] In the figure, B represents a ship, and A represents an automatic weather station and a meteorological and hydrological instrument. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] Embodiment 1: In some embodiments, please refer to the drawings of the specification Figure 1-Figure 2 The method for maneuvering fusion of shipborne meteorological and hydrological instrument data and initial field data of numerical models includes the following steps:

[0050] Step 1: Make wind direction and wind speed corrections to the shipboard meteorological and hydrological instrument data according to the ship's route speed and attitude;

[0051] Step 2: Perform quality control on the data from the automatic weather station and the ship-borne meteorological and hydrological instruments, remove the data that exceed the climatological limit values ​​and the historical variation range, and process the missing and invalid data to make the data continuous; then perform a time consistency check and an internal consistency check on the data to complete the control of data quality;

[0052] Step 3: Match the observation data of the ship-borne meteorological and hydrological instrument with the data of multiple shore-based automatic weather stations in time and space to ensure the consistency of the two in time and space, and complete the matching and calibration of the observation data of the ship-borne meteorological and hydrological instrument with the data of multiple shore-based automatic weather stations;

[0053] Step 4: The data that have been time-space matched are integrated with the observation data of the ship-borne meteorological and hydrological instruments and the data of multiple shore-based automatic weather stations through the ground analysis module in LAPS (Local Data Prediction and Analysis System), and the data are output through two-dimensional variational analysis;

[0054] Step 5: The LAPS-fused data is corrected in real time using the linear quadratic estimation algorithm and then output in NETCDF format for use in the numerical weather forecast model.

[0055] Embodiment 2: In some embodiments, Figure 1-Figure 2As shown, as a preferred embodiment of the present invention, the wind direction correction mentioned in step 1 uses an inclination sensor to monitor the sensor inclination in real time and correct it through an algorithm; the wind speed correction mentioned in step 1 uses a motion compensation algorithm or an intelligent correction algorithm;

[0056] The wind data of meteorological and hydrographic instruments, including wind speed and direction, are all generated by the wind sensors of shipborne meteorological and hydrographic instruments. The wind sensor data are easily affected by the ship's course speed and attitude, resulting in measurement errors of wind direction and wind speed, requiring wind direction correction and wind speed correction.

[0057] Wind direction correction: As ships are affected by wind, waves, currents and other factors during navigation, the ship's attitude will change during navigation, causing the sensor inclination to change, which will affect the accuracy of the shipboard anemometer data. Therefore, it is necessary to correct the sensor inclination in real time to ensure the accuracy of the data. Common correction methods include the use of

[0058] Wind speed correction: When measuring wind speed, the shipborne anemometer is often affected by the ship speed, which leads to deviations in the measurement results. In order to eliminate the influence of the ship speed on the shipborne anemometer data, the following algorithm is used:

[0059] Motion compensation algorithm: Using the information provided by the ship's navigation system, the wind speed data is motion compensated. This algorithm can eliminate the interference of ship motion on wind speed measurement and improve the accuracy of the data.

[0060] Intelligent correction algorithms: Combining artificial intelligence and machine learning technologies, we develop intelligent correction algorithms that can automatically learn and identify ship motion patterns and make real-time corrections to wind speed data accordingly.

[0061] The missing and invalid data mentioned in step 2 are processed by interpolation, mean substitution or model prediction;

[0062] Interpolation method: To maintain the continuity of data, the values ​​of adjacent data points are used to estimate the values ​​of missing data. Common interpolation methods include linear interpolation, polynomial interpolation, Lagrange interpolation, etc.

[0063] Mean substitution method: When the proportion of missing and invalid data is small and the data distribution in the data set is relatively uniform, the missing data is replaced by the mean of other similar data points in the data set.

[0064] Model prediction method: When there is an obvious correlation or predictable trend between data, a prediction model is established using the existing data, and then the model is used to predict the value of the missing data.

[0065] The super-climatological limit value processing mentioned in step 2 refers to checking whether the data of each element exceeds the critical value of the meteorological element that cannot be exceeded from the climatological point of view;

[0066] The super-historical variation range mentioned in step 2 is achieved by checking whether the value of each element exceeds the maximum and minimum values ​​that have occurred in history; within the specified geographical and temporal range, the data is checked for the main variation range of the station based on the climate statistical values; the main variation range of meteorological elements varies with different geographical regions and seasons, and its value range should not exceed the climatological limit value range.

[0067] The time consistency check mentioned in step 2 refers to the check whether the changes in meteorological records within a certain time range have a specific regularity. The time consistency check includes: maximum allowable change rate check and minimum required change rate check.

[0068] The internal consistency check mentioned in step 2 refers to the check that the relationship between the records of meteorological elements observed at the same time should conform to a certain physical connection, that is, the internal consistency check. It includes: internal consistency check of the same type of elements, internal consistency check of different types of elements; the internal consistency check of the same type of elements is generally a logical check, such as the on-time value ≥ the minimum value within the hour, the on-time value ≤ the maximum value within the hour, the hourly cumulative amount = the sum of the amounts per minute within the hour, etc.

[0069] The time-space matching mentioned in step 3 includes time matching and space matching; the time matching refers to the sampling time t of the shipborne meteorological and hydrological instrument. BA As a benchmark, synchronously take the current distance t BA The most recent time t A The automatic weather station data at the moment; the spatial matching refers to using the real-time updated longitude and latitude information (Lon_p, Lat_p) in the ship's navigation path as the longitude and latitude of the ship's meteorological and hydrographic instruments.

[0070] Time matching: shore-based automatic weather stations and ship-borne meteorological and hydrographic instruments often do not use the same time synchronization, so the data generated by these two devices are not synchronized. Since the data time resolution of both devices is 1 minute, the fusion mode usually takes observation data every 10 minutes. For this reason, the sampling time t of the ship-borne meteorological and hydrographic instrument is used. BA As a benchmark, synchronously take the current distance t BA The most recent time t A The automatic weather station data at time is fused. Assume that the fusion mode takes the time of observation data as 0:

[0071] Numerical mode acquisition time t 0 →The most recent t BA →The most recent t A

[0072] Spatial matching: Since the shore-based automatic weather station data and the initial field data of the numerical model can be directly input into the LAPS numerical model to read and analyze the spatial coordinates and longitude and latitude information, these two types of data and information do not need to be spatially matched. It is only necessary to control the location of the shore-based automatic station within the grid range of the LAPS model; the ship-borne meteorological and hydrographic instrument, because it is installed on the ship, its longitude and latitude information will change as the ship sails, so the spatial information of the ship-borne meteorological and hydrographic instrument must be confirmed. In view of the fact that the length of the hull is usually 5 meters to 50 meters, this application uses the longitude and latitude information (Lon_p, Lat_p) updated in real time in the ship's navigation path as the longitude and latitude of the ship-borne meteorological and hydrographic instrument.

[0073] The LAPS (Local Data Prediction and Analysis System) mentioned in step 4 is an emerging mesoscale data fusion analysis system; this analysis system uses the information of the ground observation network, vertical detectors, aviation satellites, etc., which constitute the sea, land and air trinity meteorological detection network, to analyze and process, and finally obtain a high-resolution three-dimensional space meteorological grid field. In information processing, LAPS will evaluate the credibility, mutual constraints, spatial representativeness and constraint relationships between various information sources, and form a data fusion module, a data analysis module, and an access prediction model module; the fusion of automatic stations and meteorological and hydrological instruments mainly calls the ground analysis module.

[0074] The ground analysis module mentioned in step 4 uses data from automatic weather stations and meteorological and hydrological instruments to generate the spatial distribution field of ground meteorological elements through a series of interpolation, smoothing and analysis techniques. It can provide a spatial resolution of less than several kilometers and a temporal resolution of hours, capture subtle changes in ground meteorological scenes, and improve the accuracy and reliability of model numerical simulations.

[0075] The algorithm of the ground analysis module is as follows:

[0076] (1) The data from the numerical model are interpolated onto the grid of LAPS output points using bilinear interpolation.

[0077] (2) Obtain the initial field based on the distance weight And the error value Then the initial field is corrected to obtain the analytical field:

[0078]

[0079] In the formula, B (xk,yk) is the kth automatic station data, is the analysis field, xk, yk are the longitude and latitude of the kth automatic station, w k is the weight parameter of the kth automatic station, w' kis the modified weight parameter of the kth automatic station, and c is the Barness filter parameter. Through two-dimensional variational analysis, the output product is an lsx file.

[0080] The linear quadratic estimation algorithm mentioned in step 5 uses a series of measurements observed over time, including statistical noise and various other noises that cause observation errors, to generate estimates of unknown variables through joint probability density distributions. These variables are often more accurate than variables based on a single measurement. The correction is completed through a two-stage process. For the prediction stage, estimates of the current state variables and their uncertainty matrix are generated; these estimates will be updated using weighted averages after observing the results of the next measurement, and more certain estimates will be given more weight.

[0081] The basic formula of the linear quadratic estimation algorithm is:

[0082] β t =Φ t-1 β t-1 +ω t-1

[0083] Y t =X t β t +ν t

[0084] The above two formulas are the state equation and the measurement equation respectively, where: is the regression coefficient, is the state vector in the correction system; ω is the input noise; φ is the state transfer matrix; Y t represents the state measurement vector, X t is the prediction factor; ν is the state measurement noise.

[0085] The parameters in the basic formula of the linear quadratic estimation algorithm are corrected in a rolling manner, and the state estimation is corrected using the updated data, and the measured information is continuously updated during the recursive process; the specific formula is as follows:

[0086] β t =β t-1 +ν t-1

[0087]

[0088] In the recursive system composed of the above six formulas, Y represents the observed value at time t, represents the forecast value at time t, represents the estimated value of the regression coefficient, R t express The error variance matrix of the extrapolated values, X t represents the predictor, is the predictor factor X t The transformed rank matrix, C t-1 express The error variance matrix of the filtered value, σ t is the forecast error variance matrix, A t is the gain matrix, which is σ t The inverse matrix of .

[0089] The rolling correction method uses the updated data to correct the state estimate, and continuously updates the measured information during the recursive process to correct the fusion result of the LAPS mode.

[0090] The fused data is output in NETCDF format for use in numerical weather forecast models. At the same time, various fusion products of meteorological and hydrological parameters, such as meteorological maps and hydrological maps, can be generated as needed.

[0091] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for the mobile fusion of shipborne meteorological and hydrological instrument data and initial field data of numerical models, characterized in that: The following steps are involved: Step 1: Make wind direction and wind speed corrections to the shipboard meteorological and hydrological instrument data according to the ship's route speed and attitude; Step 2: Perform quality control on the data from the automatic weather station and the ship-borne meteorological and hydrological instruments, remove the data that exceed the climatological limit values ​​and the historical variation range, and process the missing and invalid data to make the data continuous; then perform a time consistency check and an internal consistency check on the data to complete the control of data quality; Step 3: Match the observation data of the ship-borne meteorological and hydrological instrument with the data of multiple shore-based automatic weather stations in time and space to ensure the consistency of the two in time and space, and complete the matching and calibration of the observation data of the ship-borne meteorological and hydrological instrument with the data of multiple shore-based automatic weather stations; Step 4: The data that have been time-space matched are integrated with the observation data of the ship-borne meteorological and hydrological instruments and the data of multiple shore-based automatic weather stations through the ground analysis module in LAPS (Local Data Prediction and Analysis System), and the data are output through two-dimensional variational analysis; Step 5: The LAPS-fused data is corrected in real time using the linear quadratic estimation algorithm and then output in NETCDF format for use in the numerical weather forecast model.

2. The wind direction correction and wind speed correction mentioned in step 1 of claim 1 are characterized in that: Wind direction correction uses an inclination sensor to monitor the sensor inclination in real time and corrects it through an algorithm; wind speed correction uses a motion compensation algorithm or an intelligent correction algorithm.

3. The missing and invalid data processing mentioned in step 2 of claim 1 is characterized in that: Missing and invalid data are processed by interpolation, mean substitution or model prediction.

4. The internal consistency check mentioned in step 2 of claim 1 is characterized in that: Internal consistency check refers to the check that the relationship between the records of meteorological elements observed at the same time should conform to a certain physical connection, that is, internal consistency check; it includes internal consistency check of the same type of elements and internal consistency check of different types of elements; among which the internal consistency check of the same type of elements is generally a logical check, such as the on-time value ≥ the minimum value within the hour, the on-time value ≤ the maximum value within the hour, and the hourly cumulative amount = the sum of the amounts per minute within the hour.

5. The spatiotemporal matching mentioned in step 3 of claim 1 is characterized in that: Time-space matching includes time matching and space matching; time matching refers to the sampling time t of the shipborne meteorological and hydrological instrument. BA As a benchmark, synchronously take the current distance t BA The most recent time t A The automatic weather station data at the moment; the spatial matching refers to using the real-time updated longitude and latitude information (Lon_p, Lat_p) in the ship's navigation path as the longitude and latitude of the ship's meteorological and hydrographic instruments.

6. The ground analysis module mentioned in step 4 of claim 1, characterized in that: The ground analysis module uses data from automatic weather stations and meteorological and hydrological instruments to generate the spatial distribution field of ground meteorological elements through a series of interpolation, smoothing and analysis techniques, providing a spatial resolution of less than several kilometers and a temporal resolution of hours.

7. The ground analysis module mentioned in claim 1, characterized in that: The algorithm of the ground analysis module is as follows: (1) The data from the numerical model are interpolated onto the grid of LAPS output points using bilinear interpolation. (2) Obtain the initial field based on the distance weight And the error value Then the initial field is corrected to obtain the analytical field: In the formula, B (xk,yk) is the kth automatic station data, is the analysis field, xk, yk are the longitude and latitude of the kth automatic station, w k is the weight parameter of the kth automatic station, w' k is the modified weight parameter of the kth automatic station, and c is the Barness filter parameter.

8. The linear quadratic estimation algorithm mentioned in step 5 of claim 1, characterized in that: The linear quadratic estimation algorithm refers to the use of a series of measurements observed over time, including statistical noise and various other noises that cause observation errors, to generate estimates of unknown variables through a joint probability density distribution; for the prediction stage, an estimate of the current state variable and its uncertainty matrix are generated; these estimates will be updated using a weighted average after observing the results of the next measurement, and more certain estimates will be given more weight.

9. The linear quadratic estimation algorithm mentioned in claim 8, characterized in that The basic formula of the linear quadratic estimation algorithm is: b t =Φ t-1 b t-1 +oh t-1 Y t =X t b t +n t The above two formulas are the state equation and the measurement equation respectively, where: is the regression coefficient, is the state vector in the correction system; ω is the input noise; φ is the state transfer matrix; Y t represents the state measurement vector, X t is the prediction factor; ν is the state measurement noise.

10. The parameters in the basic formula of the linear quadratic estimation algorithm of claim 9, characterized in that: The parameters in the basic formula of the linear quadratic estimation algorithm are corrected in a rolling manner, using the updated data to correct the state estimate, and continuously updating the measured information during the recursive process; the specific formula is as follows: b t =b t-1 +n t-1 In the recursive system composed of the above six formulas, Y represents the observed value at time t, represents the forecast value at time t, represents the estimated value of the regression coefficient, R t express The error variance matrix of the extrapolated values, X t represents the predictor, is the predictor factor X t The transformed rank matrix, C t-1 express The error variance matrix of the filtered value, σ t is the forecast error variance matrix, A t is the gain matrix, which is σ t The inverse matrix of .

Images