Typhoon process sea wave mode driving wind field construction method based on SAR data assimilation
By using SAR data assimilation method, CyclObs wind farm data is assimilated to the WRF model to generate a high-precision typhoon process wave mode drive wind farm, which solves the problem of underestimation of wind speed in traditional tropical cyclone wave simulation and significantly improves the simulation accuracy.
Patent Information
- Application Number
- CN202510132580.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-06
- Publication Date
- 2025-05-30
AI Technical Summary
Traditional tropical cyclone (TC) wave simulation relies on reanalysis data, and there is a problem of underestimation of wind speed, resulting in significant deviations in effective wave height prediction of wave simulation. In the prior art, SAR data is not fully integrated with the 3DVAR module of the WRF model, resulting in low TC simulation accuracy.
Using a SAR data assimilation method, CyclObs wind farm data inversion by Sentinel-1 and Radarsat-2 images is obtained, and these data are assimilated into the WRF model using WRF 3DVAR assimilation module to generate a high-precision typhoon process wave mode driven wind farm.
Through SAR data assimilation, the accuracy of the wave mode driving wind farm under typhoon sea conditions is improved, the problem of underestimation of wind speed in traditional reanalysis data is solved, and the accuracy of numerical simulation of waves in TC process is significantly improved.
Smart Images
Figure CN120068712A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of meteorological and ocean numerical simulation, and specifically, to a method for constructing a typhoon process wave model-driven wind field based on SAR data assimilation. Background Art
[0002] Typhoons play an important role in the water mass transport in the mid-latitudes and the heat exchange at the air-sea interface. In addition, the extreme sea conditions caused by strong winds cause several damages during typhoons, especially in coastal waters. Against the background of global climate change, the intensity and trajectory of typhoons are responsible for the inter-annual changes in wave distribution. Since the 1980s, the third-generation wave model (WAM) has been developed based on solving the energy conservation wave propagation equation. Subsequently, based on the principles and framework of WAM, the National Oceanic and Atmospheric Administration (NOAA) of the United States developed WAVEWATCH-III (WW3), and Delft University of Technology developed the SWAN model. The differences between the latest versions of WW3 and SWAN are small, and both provide optional ice-wave interaction grids and parameterizations. In previous studies, the applicability of WW3 and SWAN coupled with circulation models for wave simulation considering background dynamics was confirmed.
[0003] The advantage of mesoscale meteorological numerical models is that they can perform post-forecasting and prediction of atmospheric dynamics with relatively high spatial and temporal resolutions. There are three popular numerical models in the atmospheric community, namely Weather Research and Forecasting (WRF), PSU / NCAR MM5 model, Rapid Refresh (RAP), and Global and Regional Assimilation and Prediction System (GRAPS). Some studies have been devoted to simulating the tropical cyclone (TC) wind field using various physical parameterizations. It has been confirmed that the trajectories of the simulated TC wind fields are relatively accurate, but there are often underestimations in the TC wind speeds simulated by WRF, especially for super hurricanes. This limitation may be caused by the distortion of the initialization and boundary conditions. To improve WRF simulation, observations and remote sensing products can be incorporated during the modeling process.
[0004] Traditional tropical cyclone (TC) wave simulations mostly rely on reanalysis data such as ERA-5, and there is a problem of wind speed underestimation (for example, the maximum wind speed of ERA-5 is about 30% lower than the measured value), resulting in significant prediction deviations in the significant wave height (SWH) of wave simulation. In the prior art, although SAR can invert high-resolution wind fields, it has not been fully combined with the 3DVAR module of the WRF model to optimize the TC simulation accuracy. Conventional assimilation methods (such as direct interpolation) are prone to unstable assimilation results because they do not consider the error covariance between the background field and the observed data. Summary of the Invention
[0005] Aiming at the defects in the prior art, the purpose of the present invention is to provide a method for constructing a wind field driven by a typhoon process wave model based on SAR data assimilation, which can improve the accuracy of the wind field driven by the wave model under typhoon sea conditions.
[0006] To solve the above problems, the technical solution of the present invention is as follows:
[0007] A method for constructing a wind field driven by a typhoon process wave model based on SAR data assimilation, comprising the following steps:
[0008] Obtain CyclObs wind field data from SAR images, corresponding ERA-5 reanalysis wind field data, SFMR wind field data, HY-2 altimeter data, and NDBC buoy data;
[0009] Use the 3DVAR assimilation method based on the WRF assimilation module to assimilate the CyclObs wind field data to generate a wind field driven by the typhoon process wave model;
[0010] Verify the accuracy of the driven wind field;
[0011] Verify the accuracy of the WW3 wave model driven by the generated wind field.
[0012] Preferably, the step of obtaining CyclObs wind field data from SAR images, corresponding ERA-5 reanalysis wind field data, SFMR wind field data, HY-2 altimeter data, and NDBC buoy data specifically includes: collecting CyclObs wind field data of more than 60 Sentinel-1 and Radarsat-2 images during 15 typhoon periods, ERA-5 reanalysis wind field data with a time resolution of 1 hour and a spatial resolution of 0.1°, SFMR wind field data during typhoon periods, and HY-2 altimeter data and NDBC buoy data covering the globe.
[0013] Preferably, the step of using the 3DVAR assimilation method based on the WRF assimilation module to assimilate the CyclObs wind field data to generate a wind field driven by the typhoon process wave model specifically includes: using the 3DVAR assimilation module in WRF to cyclically assimilate the CyclObs wind field data of no less than 4 SAR images during each typhoon period into the WRF model to generate the corresponding model-driven wind field.
[0014] Preferably, the step of using the 3DVAR assimilation method based on the WRF assimilation module to assimilate the CyclObs wind field data to generate a wind field driven by the typhoon process wave model specifically includes:
[0015] Prepare background field data: use the previous forecast results of WRF as the initial guess and the SAR wind field data observation data;
[0016] Process the data, convert the original observed data into a format readable by WRFDA, and perform temporal and spatial interpolation to make it match the model grid and time window. Eliminate abnormal data and check whether the deviation between the observation and the background field is within the allowable range.
[0017] Calculate the difference between the background field and the observation, combine the error information of both, and find the optimal solution by constructing a cost function.
[0018] Generate a new initial field: Use the optimized result as the new starting point of WRF and re-run the forecast.
[0019] Preferably, the step of verifying the accuracy of the wind field driven by the verification specifically includes: verifying the WRF assimilated wind speed simulation results through SFMR wind field data, performing spatio-temporal matching on the two types of data, and screening using the conditions that the time is less than half an hour and the spatial distance is less than 0.5 km to evaluate the accuracy of the WRF assimilated wind field simulation.
[0020] Preferably, the step of verifying the accuracy of the WW3 wave model driven by the generated wind field specifically includes: verifying the WW3 simulation results through HY-2 altimeter data and NDBC buoy data, performing spatio-temporal matching on the two types of data, and screening using the conditions that the time is less than half an hour and the spatial distance is less than 0.5 km to evaluate the accuracy of the WRF assimilated wind field driving the WW3 model simulation.
[0021] Compared with the prior art, the present invention uses the CyclObs wind field data during 15 typhoons retrieved from the obtained Sentinel-1 and Radarsat-2 images, assimilates a large amount of CyclObs wind field data into the WRF model through the 3DVAR module of WRF, performs spatio-temporal matching with SFMR data, inputs the WRF assimilated wind field into the WW3 model, and based on the high-precision wind field assimilated by SAR and WRF 3DVAR, solves the problem of underestimation of wind speed in traditional reanalysis data and improves the accuracy of numerical simulation of TC process waves. Description of the Drawings
[0022] By reading the detailed description of the non-limiting embodiments with reference to the following drawings, other features, objects, and advantages of the present invention will become more apparent:
[0023] Figure 1 It is a flow chart of the method for constructing the typhoon process wave model driving wind field based on SAR data assimilation of the present invention;
[0024] Figure 2 It is an example diagram of the CyclObs typhoon wind field;
[0025] Figure 3It is an example diagram of the reanalysis wind field of the European Centre for Medium-Range Weather Forecasts;
[0026] Figure 4 It is an example diagram of the HY-2 altimeter and the selected buoy positions;
[0027] Figure 5 It is a wind speed result diagram generated after the WRF assimilates SAR;
[0028] Figure 6 It is a diagram of the variation of wind speed with distance for WRF assimilation and ERA-5;
[0029] Figure 7 It is a spatio-temporal matching diagram of the wind speed between WRF assimilation and SFMR;
[0030] Figure 8a 、 8b It is a comparison diagram of the significant wave height simulated by WW3 and the HY-2 altimeter product;
[0031] Figure 9a 、 9b It is a comparison diagram of the significant wave height simulated by WW3 and the NDBC buoy data. Detailed implementation manners
[0032] The present invention will be described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but do not limit the present invention in any form. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several changes and improvements can still be made. These all belong to the protection scope of the present invention.
[0033] Specifically, the present invention provides a method for constructing a wind field driven by a typhoon process wave model based on SAR data assimilation, as Figure 1 shown, the method includes the following steps:
[0034] S1: Obtain CyclObs wind field data from SAR images, the corresponding ERA-5 reanalysis wind field data, SFMR wind field data, HY-2 altimeter data, and NDBC buoy data;
[0035] Specifically, a total of CyclObs wind field data from more than 60 Sentinel-1 and Radarsat-2 images during 15 typhoon periods, ERA-5 reanalysis wind field data with a time resolution of 1 hour and a spatial resolution of 0.1°, SFMR wind field data during typhoon periods, and HY-2 altimeter data covering the globe and National Data Buoy Center (NDBC) buoy data were collected.
[0036] Figure 2This is an example of the wind field of Typhoon CyclObs obtained at TC Fiona at 22:34 on September 23, 2022, where the red line represents the track of the aircraft carrying the SFMR. Figure 3 This is an example of the TCFiona wind field from the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis (ERA-5). Figure 4 This is an example diagram of the HY-2 altimeter and the selected buoy position, where the red triangle is the buoy position.
[0037] S2: Use the 3DVAR assimilation method based on the WRF assimilation module to assimilate the CyclObs wind field data to generate the typhoon process wave mode driven wind field;
[0038] Specifically, using the 3DVAR assimilation module in WRF, the CyclObs wind field data of no less than 4 SAR images during each typhoon are cyclically assimilated into the WRF model to generate the corresponding model-driven wind field. The specific steps are as follows: 1. Prepare background field data: use the previous WRF forecast results as the initial guess and SAR wind field data observation data; 2. Process the data, convert the original observation data into WRFDA readable format (ASCII), and perform time and space interpolation to match the model grid and time window; remove abnormal data (such as values beyond a reasonable time or space range), and check whether the deviation between the observation and the background field is within the allowable range; 3. Calculate the difference between the background field and the observation, combine the error information of the two, and use the constructed cost function to find the "optimal solution". The specific expression is as follows:
[0039]
[0040] Among them, x b is the background field, x is the analysis field, y 0 is the observed data, B is the background error covariance matrix, R is the observation error covariance matrix, H is the observation operator, and the minimum cost function is used to find the analysis field with the smallest J(x); 4. Generate a new initial field: use the optimized result as the new starting point of WRF and re-run the forecast. Figure 5 The wind speed results generated by WRF assimilation of SAR are shown, where the red line indicates the SFMR data observation point at that moment.
[0041] S3: Verify the accuracy of the driven wind field;
[0042] Specifically, the WRF assimilated wind speed simulation results were verified by SFMR wind field data. The two types of data were matched in time and space, and the time less than half an hour and the spatial distance less than 0.5 kilometers were used as conditions for screening to evaluate the accuracy of the WRF assimilated wind field simulation.
[0043] Generate model-driven wind fields through assimilation Figure 5The variation of wind speed with distance in the WRF assimilation of the SFMR observation point and ERA-5 is as follows Figure 6 shown. The wind speed generated by WRF assimilation is closer to the SFMR observation results and significantly higher than the ERA-5 data. Figure 7 The comparison between WRF assimilation and SFMR measurement results is shown, with an RMSE of 3.48 m / s for wind speed and a correlation coefficient (Cor) of 0.89. The results indicate that the driving wind field generated by WRF assimilation is reliable.
[0044] S4: Verify the accuracy of the WW3 wave model driven by the generated wind field.
[0045] Specifically, the WW3 simulation results are verified through HY-2 altimeter data and NDBC buoy data. The two types of data are matched in space and time, and screened by the conditions of time less than half an hour and spatial distance less than 0.5 km, which are used to evaluate the accuracy of the WW3 model simulation driven by the WRF assimilation wind field.
[0046] Figure 8a , 8b is a comparison chart of the significant wave height of the WW3 simulation and the HY-2 altimeter product, where Figure 8a is the result chart of the wave model driven by the WRF assimilation wind field, Figure 8b is the result chart of the wave model driven by the ERA-5 wind field. The statistical results of using the WRF assimilation wind field have an RMSE of 0.78 m, a correlation coefficient (Cor) of 0.91, a standard deviation (SI) of 0.44, and a Bias of -0.36, which are better than the statistical results of using the ERA-5 wind field with an RMSE of 0.82 m, a correlation coefficient (Cor) of 0.82, a standard deviation (SI) of 0.46, and a Bias of -0.67. Similarly, the comparison with the measurement results of the NDBC buoy is as Figure 9a , 9b shown, Figure 9a , 9b is a comparison chart of the significant wave height of the WW3 simulation and the NDBC buoy data, where Figure 9a is the result chart of the wave model driven by the WRF assimilation wind field, Figure 9bFor the result diagram driven by the ERA-5 wind field for use in the wave model, the statistical results using WRF assimilation show an RMSE of 0.66 m, a correlation coefficient (Cor) of 0.89, a standard deviation (SI) of 0.41, and a Bias of -0.20, which are better than the statistical results using the ERA-5 wind field with an RMSE of 0.73 m, a correlation coefficient (Cor) of 0.85, a standard deviation (SI) of 0.45, and a Bias of -0.56. The SWH using WRF assimilation has a significant improvement in the comparison results, especially in terms of bias. Substantially, the wave model using the driving wind field generated by WRF assimilation of SAR can effectively improve the underestimation problem of the ERA5 driving results in high sea states.
[0047] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.
Claims
1. A method for constructing a typhoon process wave pattern driven wind field based on SAR data assimilation, characterized in that: The method comprises the following steps: Obtain CyclObs wind field data from SAR images, the corresponding ERA-5 reanalysis wind field data, SFMR wind field data, HY-2 altimeter data, and NDBC buoy data; The 3DVAR assimilation method based on the WRF assimilation module is used to assimilate the CyclObs wind field data to generate the typhoon process wave mode driven wind field; Verify the accuracy of the driven wind field; Verify the accuracy of generating wind-site driven WW3 wave models.
2. The method for constructing a typhoon process wave pattern driven wind field based on SAR data assimilation according to claim 1, characterized in that: The steps of obtaining CyclObs wind field data from SAR images, corresponding ERA-5 reanalysis wind field data, SFMR wind field data, HY-2 altimeter data, and NDBC buoy data specifically include: collecting CyclObs wind field data of more than 60 Sentinel-1 and Radarsat-2 images during 15 typhoons, ERA-5 reanalysis wind field data with a time resolution of 1 hour and a spatial resolution of 0.1°, SFMR wind field data during typhoons, and HY-2 altimeter data and NDBC buoy data covering the world.
3. The method for constructing a typhoon process wave pattern driven wind field based on SAR data assimilation according to claim 1, characterized in that: The step of assimilating the CyclObs wind field data using the 3DVAR assimilation method based on the WRF assimilation module to generate the typhoon process wave mode driven wind field specifically includes: using the 3DVAR assimilation module in WRF, cyclically assimilating the CyclObs wind field data of no less than 4 SAR images during each typhoon into the WRF model to generate the corresponding mode driven wind field.
4. The method for constructing a typhoon process wave pattern driven wind field based on SAR data assimilation according to claim 3 is characterized in that: The step of assimilating the CyclObs wind field data using the 3DVAR assimilation method based on the WRF assimilation module to generate the typhoon process wave mode driven wind field specifically includes: Prepare background field data: use the previous WRF forecast results as initial guesses and SAR wind field data observation data; Process the data, convert the raw observation data into a WRFDA readable format, and perform temporal and spatial interpolation to match the pattern grid and time window, remove abnormal data, and check whether the deviation between the observation and the background field is within the allowable range; Calculate the difference between the background field and the observation, combine the error information of the two, and use the constructed cost function to find the optimal solution; Generate a new initial field: Use the optimized results as a new starting point for WRF and rerun the forecast.
5. The method for constructing a typhoon process wave pattern driven wind field based on SAR data assimilation according to claim 1, characterized in that: The step of verifying the accuracy of the driven wind field specifically includes: verifying the WRF assimilated wind speed simulation results through SFMR wind field data, matching the two data in time and space, and screening them by using time less than half an hour and spatial distance less than 0.5 kilometers as conditions to evaluate the accuracy of the WRF assimilated wind field simulation.
6. The method for constructing a typhoon process wave pattern driven wind field based on SAR data assimilation according to claim 1, characterized in that: The step of verifying the accuracy of the generated wind field driven WW3 wave model specifically includes: verifying the WW3 simulation results through HY-2 altimeter data and NDBC buoy data, matching the two types of data in time and space, and screening by using time less than half an hour and spatial distance less than 0.5 kilometers as conditions to evaluate the accuracy of the WRF assimilated wind field driven WW3 model simulation.