Direct assimilation method for ground-based microwave radiometer

Abstract

The invention discloses a direct assimilation method for a ground-based microwave radiometer, and relates to the technical field of meteorological observation, and the method comprises the steps: calling an RTTOF-gb rapid radiation transmission mode, enabling a derivation type in a WRFDA system to correspond to an RTTOF-gb derivation type, enabling RTTOV, RTTOF-gb and the WRFDA system to be coupled at the same time, and converting a background field of a numerical weather forecast mode into simulated brightness temperature data; the source of the deviation and the nonlinear relation between the source and the deviation are considered, a deviation correction forecast factor is calculated according to the original observation brightness temperature and the background field, deviation correction is conducted on the original observation brightness temperature, and the corrected observation brightness temperature is obtained; and based on the corrected observation brightness temperature data, performing three-dimensional variation assimilation processing on the background field to obtain an analysis field, and outputting a diagnosis file. According to the method, a good deviation correction effect can be achieved, adjustment can be conveniently carried out according to actual application requirements, RTTOF-gb and RTTOV can be used at the same time, and combined direct assimilation of various foundation microwave radiometers and satellite-borne microwave radiometers can be achieved.

Description

Technical Field

[0001] The present invention relates to the field of meteorological observation technology, and in particular to a direct assimilation method for ground-based microwave radiometers. Background Art

[0002] Numerical weather forecasting relies on the efficient fusion of multi-source observation data by the data assimilation system. Among them, ground-based microwave radiometers have unique value in short-term forecasting due to their minute-level high-frequency observation capabilities and their advantages in fine capture of the atmospheric boundary layer. This type of data assimilation technology is currently dominated by indirect assimilation, that is, data fusion is achieved by inverting temperature and humidity parameters, but there are three limitations: uncertainty and large errors are introduced during the inversion process, thereby limiting the effectiveness of assimilation; the inversion process makes error estimation complex, and it is difficult to quantify and process these errors through the error covariance matrix; secondary assimilation leads to the reuse of background information and reduces the independence of observations. In contrast, direct assimilation avoids inversion errors by processing raw radiation brightness temperature data, and has now become the preferred solution for operational applications.

[0003] The observation operator is the core component for realizing the direct assimilation of ground-based microwave radiometers. Its function is to convert variables such as temperature and water vapor in the numerical model into radiation data equivalent to radiation observation. The direct assimilation technology of spaceborne microwave radiometers is relatively mature, and usually uses a fast radiation transfer model as the observation operator, such as RTTOV (Radiative Transfer for the Television Infrared Observation Satellite Operational Vertical Sounder). However, the ground-based microwave radiometer has an essential difference from the spaceborne platform in the bottom-up observation path, which makes the spaceborne radiation transfer model unable to be directly applied. In recent years, the fast radiation transfer model RTTOV-gb (ground-based version of RTTOV) suitable for ground-based microwave radiometers has been developed, providing strong support for the direct assimilation of ground-based microwave radiometers.

[0004] However, the direct assimilation technology of ground-based microwave radiometer data in the existing technology does not perform nonlinear bias correction. The assimilation method assumes that the observation error and background error follow an unbiased Gaussian distribution. But in practice, observation instrument errors, limitations of radiation transfer models, and background errors in numerical models may lead to systematic deviations between observations and simulated radiation, thus affecting the assimilation quality. Bias correction is one of the key technologies for direct assimilation of ground-based microwave radiometers, which aims to identify and correct the deviations between observations and backgrounds to improve assimilation accuracy. When using variational assimilation, the existing technology does not perform bias correction and ignores the systematic deviations between observations and backgrounds; when using ensemble Kalman filtering for assimilation, the existing technology uses linear bias correction, but does not consider the source of the bias and the nonlinear bias component.

[0005] The topography and geomorphology of the southwest region are complex. Affected by large terrains such as the plateau and the Hengduan Mountains, the model background error is often biased and very complex. In addition, the applicability of RTTOV-gb is limited in these areas because its training data mainly comes from global profile data, rather than specifically for the plateau. Therefore, when performing direct assimilation of ground-based microwave radiometers in these areas, there may be complex nonlinear biases between observations and simulations. If appropriate bias correction is not performed during direct assimilation, the assimilation effect will be greatly limited and the benefits of observations will be reduced. Summary of the invention

[0006] Aiming at complex terrains such as the plateau and the Hengduan Mountains, the present invention provides a direct assimilation method for ground-based microwave radiometers, aiming to effectively correct the deviation of OMB (observation minus background) through deviation correction. According to the observation characteristics of ground-based microwave radiometers, the source of the deviation (prediction factor of deviation correction) and the nonlinear relationship between the source and the deviation are considered, and the contribution of each prediction factor in the correction process is evaluated.

[0007] The technical solution adopted by the present invention is as follows: A direct assimilation method for ground-based microwave radiometers, including: Read the original radiation observation data of the ground-based microwave radiometer and pre-process it to obtain the original observation brightness temperature data; Call the RTTOV-gb fast radiative transfer model, match the derived types in the WRFDA (weather research and forecasting model data assimilation system) system with the RTTOV-gb derived types, so that the RTTOV, RTTOV-gb and WRFDA systems are coupled simultaneously, and then the background field of the numerical weather forecast model is converted into simulated brightness temperature data; Based on the machine learning algorithm, the source of the deviation and the nonlinear relationship between the source and the deviation are considered. The deviation correction prediction factor is calculated according to the original observed brightness temperature data and the background field. The deviation of the original observed brightness temperature data is corrected based on the deviation correction prediction factor to obtain the corrected observed brightness temperature data. Calculate observational residual data based on simulated brightness temperature data and corrected observed brightness temperature data, and select corrected observed brightness temperature data required for assimilation based on observed residual data; Using the assimilation module of the WRFDA system, the background field is subjected to three-dimensional variational assimilation according to the screened and corrected observed brightness temperature data. The analysis field is obtained by iterating the minimization of the cost function and the diagnostic file is output.

[0008] As a further description of the above technical solution, the preprocessing method includes: (a) Time zone conversion, unifying observation data into Coordinated Universal Time; (b) Check the temporal and spatial consistency, compare with the background field, and select the observation data required for assimilation time and region; (c) Cloud and rain detection: cloud area detection is performed on the selected observation data using the AGRI (Advanced Geostationary Radiation Image) cloud classification product of the Fengyun-4B satellite, and rain area detection is performed on the selected observation data using the rain gauge data of the microwave radiometer; (d) Observation data output: The observation data after cloud and rain detection are classified and summarized according to the instrument type, and the required files for assimilation are output one by one to obtain the original observation brightness temperature data.

[0009] As a further description of the above technical solution, the method for converting the background field of the numerical weather forecast model into simulated brightness temperature data specifically includes: reading the coefficient file of the ground-based microwave radiometer, reading the meteorological element data from the background field, performing horizontal and vertical spatial interpolation, and assigning them to the derived arrays corresponding to the RTTOV-gb model respectively, and calculating the simulated brightness temperature data of the ground-based microwave radiometer through RTTOV-gb.

[0010] As a further description of the above technical solution, when introducing RTTOV-gb in the WRFDA system and building the ground-based microwave radiometer direct assimilation module, it is necessary to replace keywords in the RTTOV-gb source code, replace the file name involving the code itself, and also replace the definition and call name of the relevant interface, module, library in the code to avoid conflict with RTTOV and realize the coexistence of the two transmission modes.

[0011] As a further description of the above technical solution, the method for obtaining the corrected observed brightness temperature data based on the machine learning algorithm specifically includes: S31, WRFDA system calculates the bias correction prediction factor based on the original observed brightness temperature data and background field; S32, based on the bias correction prediction factor outputted from step S31, the bias correction model obtained by pre-training is used to predict the bias of OMB (observed brightness temperature minus simulated brightness temperature). The bias correction model is constructed based on the random forest algorithm, which can be used in the variational assimilation framework, and takes into account the source of the bias and the nonlinear relationship between the source of the bias and the bias according to the observation characteristics and terrain position of the ground-based microwave radiometer, and can evaluate the importance of the source of the bias. During training, it is trained separately for different instruments and different channels; S33, performing deviation correction on the original observed brightness temperature data according to the deviation of the OMB predicted in step S32 to obtain the corrected observed brightness temperature data.

[0012] As a further description of the above technical solution, the deviation correction prediction factors selected in step S32 include 1000-700hPa atmospheric thickness, 700-500hPa atmospheric thickness, 500-300hPa atmospheric thickness, 1000-300hPa atmospheric thickness, 200-50hPa atmospheric thickness, 2 meter temperature, 2 meter humidity, 10 meter meridional wind, 10 meter zonal wind, ground pressure, ground temperature, atmospheric precipitable water, observed brightness temperature, longitude and latitude.

[0013] As a further description of the above technical solution, for the corrected observed brightness temperature data, when it meets any of the exclusion conditions, it will be excluded, otherwise it is the corrected observed brightness temperature data required for assimilation; the exclusion conditions include: the corresponding observed residual data is greater than 20K; the corresponding observed residual data is greater than 3 times the error; it is the observation data under cloud and rain environment; the cloud water integral of the entire layer at the observation point is greater than 0.2g / kg. As a further description of the above technical solution, three-dimensional variational assimilation processing is performed based on minimization iteration, and the objective cost function of the minimization iteration is defined as: ; in, and denote the analysis field and the background field respectively; is the corrected observed brightness temperature data; and denote the background error covariance matrix and the observation error covariance matrix respectively; Represents the observation operator.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1) In previous studies on direct assimilation of ground-based microwave radiometers, variational assimilation did not adopt bias correction, and the bias correction method in ensemble assimilation did not consider the bias source. The method of the present invention fills the gap in the bias correction technology of ground-based microwave radiometers under variational assimilation and can be extended to ensemble assimilation.

[0015] 2) The present invention considers the source of deviation in more detail, and considers the nonlinear relationship between the source of deviation and deviation, which is conducive to achieving better deviation correction effect and convenient for adjustment according to the needs of actual application.

[0016] 3) Based on the modified RTTOV-gb, the present invention establishes a ground-based microwave radiometer direct assimilation module in WRFDA, which can realize the direct assimilation of various ground-based microwave radiometers, and sets a universal interface, which is conducive to the subsequent realization of direct assimilation of more types of ground-based microwave radiometers.

[0017] 4) The present invention solves the coupling problem between the radiation transmission modes RTTOV-gb and RTTOV required by ground-based microwave radiometers and satellite-borne microwave radiometers in the data assimilation system WRFDA, so that the two can be used at the same time. This setting is conducive to the subsequent joint direct assimilation of satellite-borne and ground-based microwave radiometers.

[0018] 5) The present invention can evaluate the contribution of prediction factors in deviation correction, which is helpful to discover the source of deviation and fundamentally solve these deviations. It also provides a reference for other deviation correction methods to consider the source of deviation.

[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the embodiments of the present invention are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 is a flow chart of the direct assimilation method described in the Examples; Figure 2 RTTOV-gb code processing flow chart in the embodiment; Figure 3 is the training and evaluation process of the deviation correction model in the embodiment; Figure 4is the OMB deviation and standard deviation of each channel of HATPRO (the Humidity And Temperature PROfiler) in the embodiment; Figure 5 is a predictor contribution graph in the embodiment, i.e., the importance of predictors used in the bias correction model. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions 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 only part of the embodiments of the present invention, not all of the embodiments.

[0023] Please refer to Figure 1 The embodiment of the present invention provides a direct assimilation method for a ground-based microwave radiometer, including: (1) Read the original radiation observation data of the ground-based microwave radiometer and preprocess it to obtain the original observation brightness temperature data.

[0024] This step uses Python to read the original radiation observation data of the ground-based microwave radiometer and preprocess the original data: (a) Time zone conversion, unifying observation data into Coordinated Universal Time; (b) Check the temporal and spatial consistency, compare with the background field, and select the observation data required for assimilation time and region; (c) Cloud and rain detection: cloud area detection is carried out using the AGRI cloud classification product of the Fengyun-4B satellite, and rain area detection is carried out using the rain gauge data of the microwave radiometer; (d) Observation data output: The multi-station data are classified and summarized according to the instrument type, and the required files for assimilation are output one by one to obtain the original observation brightness temperature data.

[0025] (2) Call the RTTOV-gb rapid radiative transfer model to convert the background field of the numerical weather forecast model into simulated brightness temperature data.

[0026] In the embodiment of the present invention, the observation operator is constructed using the fast radiation transfer model to directly assimilate the radiation amount without introducing the inversion process, so that the assimilated observation accuracy is higher and the error is more controllable, which is more effective for improving the initial value in the numerical weather model. Based on the RTTOV-gb fast radiation transfer model, the direct assimilation of the ground-based microwave radiometer can be realized. For single frequency radiation, the atmospheric transmission model is calculated as follows: ; in is the radiation received by the ground; is the atmospheric temperature; is the amount of Planck radiation at a specific temperature; is the transmittance from the top of the atmosphere to the ground; is the microwave cosmic background temperature, set to 2.728 K. RTTOV-gb uses a linear relationship to calculate the optical thickness and transmittance of the atmosphere based on the given instrument elevation angle, atmospheric temperature and humidity profile, and air pressure, thereby obtaining the simulated radiation (brightness temperature).

[0027] The existing WRFDA system does not implement the access of RTTOV-gb and the construction of the ground-based microwave radiometer assimilation module. The embodiment of the present invention compares the coupling status of WRFDA and RTTOV, establishes a coupling framework of WRFDA and RTTOV-gb in parallel, and avoids the conflict of simultaneous coupling of RTTOV and RTTOV-gb.

[0028] RTTOV, RTTOV-gb and WRFDA are coupled simultaneously: The WRFDA system uses RTTOV to construct observation operators to achieve satellite data assimilation. The present invention uses RTTOV-gb based on the WRFDA system to form observation operators to achieve direct assimilation of ground-based microwave radiometers. RTTOV-gb is developed based on RTTOV. The two do not consider the scenario of simultaneous use. The function module names are exactly the same, and the keyword for calling the interface is rttov, but the actual calculations are different. If the direct assimilation of ground-based microwave radiometers is to be achieved without affecting the assimilation of satellite data, it is necessary to avoid conflicts with the existing RTTOV when introducing RTTOV-gb in WRFDA. The present invention adopts the method of keyword replacement of all codes involving interfaces, modules, and libraries in the RTTOV-gb source code. It is necessary to replace the file names involving the code itself, as well as the definitions and calling names of related interfaces, modules, and libraries in the code. The specific process is as follows: Figure 2 As shown: 1) Compile RTTOV-gb normally; 2) Get key files: traverse the .h suffix files in the src directory, the .F90 suffix files in the src directory, the .o suffix files in the obj directory, the .mod suffix files in the mod directory, and the .interface suffix files in the include directory; 3) Get keywords: remove the path and suffix of the key file name; 4) Prepare uncompiled RTTOV-gb code; 5) Get specific files: traverse the .F90, .h, .f suffix files and Makefile files in the src / main, src / test, src / coef_io, src / other, and src / parallel directories; 6) Replace the file name of a specific file: add _gb to the main name, for example, Y.f90 is replaced with Y_gb.f90; 7) Replace the file content of a specific file: Replace the keywords in the file. If the keyword X is not preceded or followed by letters or underscores, add _gb after the keyword, for example, replace X with X_gb.

[0029] After the replacement is completed, start compiling RTTOV-gb. At this time, the keywords of RTTOV-gb interfaces, libraries, and modules are all rttov_gb, which are independent of the keywords of RTTOV-related interfaces, libraries, and modules, and the two do not affect each other. Comparing the current coupling status of WRFDA and RTTOV, a coupling framework of WRFDA and RTTOV-gb is established in parallel based on the modified and compiled RTTOV-gb to achieve the coordinated coupling of the assimilation system with the two sets of radiation transmission modes, namely, satellite-borne and ground-based. Combined with RTTOV-gb, the observation operators and their tangent linear adjoints required for direct variational assimilation of microwave radiometer brightness temperature observations are established, and the conversion between brightness temperature observations and model variables such as temperature, humidity, and wind field is completed. A direct assimilation module of ground-based microwave radiometer is established, including observation reading, assimilation, and diagnostic file output.

[0030] The embodiment of the present invention establishes an independent control structure of the ground-based microwave radiometer in WRFDA, and realizes the direct assimilation of the ground-based microwave radiometer without affecting the assimilation of other types of data. First, the derived type in the WRFDA system is matched with the derived type of RTTOV-gb, and the coefficient file of the ground-based microwave radiometer is read; the meteorological elements such as temperature, humidity, and air pressure are read from the background field, and the horizontal and vertical spatial interpolation is performed, and the values ​​are respectively assigned to the derived arrays corresponding to the RTTOV-gb mode; the simulated brightness temperature of the ground-based microwave radiometer is calculated by RTTOV-gb.

[0031] (3) Based on the machine learning algorithm, the source of the deviation and the nonlinear relationship between the source and the deviation are considered. The deviation correction prediction factor is calculated according to the original observed brightness temperature data and the background field. The deviation of the original observed brightness temperature data is corrected based on the deviation correction prediction factor to obtain the corrected observed brightness temperature data, as follows: S31, WRFDA system calculates the bias correction prediction factor based on the original observed brightness temperature and background field.

[0032] In this embodiment, by constructing relevant calculation codes in the WRFDA system, the deviation correction prediction factor is calculated based on the original observed brightness temperature and background field.

[0033] S32, based on the deviation correction prediction factor outputted in step S31, the deviation of the current OMB is predicted using the pre-trained deviation correction model.

[0034] In this embodiment, based on the output diagnostic information, the following elements are selected as deviation correction predictors: 1000-700hPa atmosphere thickness, 700-500hPa atmosphere thickness, 500-300hPa atmosphere thickness, 1000-300hPa atmosphere thickness, 200-50hPa atmosphere thickness, 2m temperature, 2m humidity, 10m meridional wind, 10m zonal wind, surface pressure, surface temperature, atmospheric precipitable water, observed brightness temperature, longitude, latitude.

[0035] Variable assimilation assumes that both observation and background errors follow an unbiased Gaussian distribution. However, due to instrument errors, limitations of radiation transfer models, and background errors in numerical weather prediction models, observed brightness temperatures and simulated brightness temperatures contain errors, and their errors may be biased distributions. Bias correction is a key process in radiation data assimilation, the purpose of which is to identify and eliminate these biases. In the real atmosphere, the observed brightness temperature and simulated brightness temperature are regarded as the true value T plus their respective differences, as shown in the following formula: ; in, To observe brightness temperature, To simulate brightness temperature, The statistical expectation of can represent systematic bias, so it is critical that OMB assesses the characteristics of systematic bias and corrects for it.

[0036] For some areas, affected by the large plateau terrain and complex terrain, the deviation of OMB will be very obvious. No deviation correction or simple deviation correction cannot meet the assimilation requirements. Therefore, based on machine learning, this paper proposes a deviation correction method for direct assimilation of ground-based microwave radiometers. This method is based on the random forest algorithm in machine learning and aims to correct the OMB deviation. It has the following characteristics: 1) It can be used in the variational assimilation framework; 2) Consider the source of bias, i.e., the prediction factor of bias correction, based on the observation characteristics and terrain location of ground-based microwave radiometers; 3) Consider the nonlinear relationship between predictors and bias; 4) the importance of predictors can be assessed; 5) Conduct separate training for different instruments and channels.

[0037] In order to evaluate the reliability of the method, the bias correction model needs to be trained and evaluated. This embodiment conducts a three-month experiment to obtain a large number of samples. The experiment uses WRF (weather research and forecasting model) as the numerical weather forecast model, and uses NCEP-GFS (the National Centers for Environmental Prediction Global Forecast System) data to initialize once every 6 hours. WRFDA runs in monitoring mode (no minimization iteration is performed in this mode) every hour to obtain a large number of samples.

[0038] The specific process is as follows Figure 3 As shown, including: 1) Preprocessing of raw radiation observations from multi-site and multi-model ground-based microwave radiometers. The preprocessing process is the same as the previous preprocessing method.

[0039] 2) Simulated brightness temperature calculation Based on RTTOV-gb, the observation operator is constructed to complete the conversion of model space variables to observation space variables. Based on the model background pressure, temperature, humidity profile and the given observation instrument elevation angle, the observation operator calculates the optical thickness and transmittance of the simulated atmosphere to obtain the simulated radiation, that is, the simulated brightness temperature.

[0040] 3) Sample statistics and segmentation a. Sample information acquisition In order to obtain training and evaluation samples, this step is based on the ground-based microwave radiometer direct assimilation module constructed in WRFDA to calculate the output prediction variables and prediction factors to form multi-site and multi-time samples.

[0041] b. Acquisition of training set To train the bias correction model, 70% of the data in the sample information is randomly selected as training samples, that is, a training set is constructed.

[0042] c. Test set acquisition In order to evaluate the effect of bias correction, the remaining 30% of the data in the sample information is selected as the test sample, that is, the test set is constructed.

[0043] 4) Deviation Correction Model Training a. Hyperparameter Tuning The essence of using random forest to build a bias correction model is to use the forecast factor as the characteristic variable of the model and OMB as the predictor variable, fit the relationship between the forecast factor and OMB, and predict the OMB deviation through the forecast factor. The forecast factors include 1000-700hPa atmospheric thickness, 700-500hPa atmospheric thickness, 500-300hPa atmospheric thickness, 1000-300hPa atmospheric thickness, 200-50hPa atmospheric thickness, 2m temperature, 2m humidity, 10m meridional wind, 10m zonal wind, surface pressure, surface temperature, atmospheric precipitable water, observed brightness temperature, longitude, and latitude.

[0044] There are two types of parameters in the bias correction model: model parameters and hyperparameters. Model parameters are initialized and updated during the learning process. Hyperparameters define the architecture of the model and cannot be updated during the model training process. They must be set before training. The setting of hyperparameters greatly affects the training time and prediction performance of the model. The process of determining the hyperparameter configuration and the ideal model architecture is called hyperparameter tuning. Hyperparameter tuning is a key element of effective machine learning models. The random forest algorithm has four key hyperparameters: the number of trees in the forest, the maximum depth of the tree, the minimum number of samples required to split the internal nodes, and the minimum number of samples required for the leaf nodes. These hyperparameters are set to a certain range and combined to form a hyperparameter grid, which contains all the possible hyperparameters within the preset range. Based on the training set data, the hyperparameter grid is traversed and the model is trained and scored to obtain the score of each hyperparameter combination. Among them, for a certain hyperparameter setting, the model training and scoring adopts a 5-fold cross-validation method: after dividing the data into 5 equal parts. Select the first set of data as test data, and the remaining four sets as training data. Use the training data to train the random forest model, and use the test data to evaluate the effect of the trained model, and calculate the model score S1. Based on this, select the second, third, fourth, and fifth sets of data as test data, and calculate the model scores S2, S3, S4, and S5. Finally, average the five scores to obtain the hyperparameter score S.

[0045] b. Optimal hyperparameter settings According to the scoring results, the hyperparameter combination with the highest score S is selected as the optimal hyperparameter.

[0046] c. Bias Correction Model Based on the optimal hyperparameter settings, the model is retrained using the training set. Each type of instrument and each unit is trained separately and saved locally.

[0047] 5) Bias Correction Model Evaluation The randomly selected test set data accounts for 30% of the total samples and does not participate in model training. This data can be used to test the effectiveness of the bias correction model.

[0048] a. Original OMB Based on the test set data, the original OMB is calculated and statistically obtained.

[0049] b. Forecasted OMB Based on the test set data, for each type of instrument and channel, the corresponding deviation correction model is loaded, the selected prediction factors are input, and the OMB is predicted to obtain the predicted OMB. express.

[0050] c. Revised OMB Based on the original OMB and the predicted OMB, the revised OMB is obtained. Use to express: ; d. Comparative evaluation Based on the original OMB and the revised OMB, the effect of model bias correction and the importance of predictor factors were evaluated.

[0051] Figure 4 The deviation correction effect of each channel of HATPRO is shown, where (a) shows the OMB deviation with and without deviation correction, and (b) shows the OMB standard deviation with and without deviation correction. Before deviation correction, the OMB deviation of HATPRO is between 0 and 2K, and the deviation of the K band (especially channels 4 to 7) is smaller than that of the V band. After deviation correction, the deviation of each channel is about 0K. As for the STD (Standard Deviation) of OMB, its value is between 2 and 4K without deviation correction, and the values ​​of channels 4 to 7 are smaller than those of other channels. After deviation correction, the STD of OMB fluctuates between 0.5 and 1.5K. After applying the deviation correction model, both the OMB deviation and the OMB standard deviation are greatly reduced, by 0.77K (i.e., a reduction of 99.8%) and 1.6K (i.e., a reduction of 65.0%), respectively. The corrected OMB distribution exhibits a Gaussian characteristic centered at zero, indicating that the systematic OMB bias has been effectively alleviated.

[0052] It is crucial to diagnose the contribution of each predictor. Figure 5The importance of predictors in HATPRO bias correction is demonstrated. The higher the score, the stronger the correlation between the predictor and the OMB bias. Observed brightness temperature, total atmospheric precipitable water, and surface pressure are important factors in K-band bias correction. For the V-band, observed brightness temperature, latitude, and surface pressure are the most influential predictors. Compared with other predictors, the contribution of the atmospheric thickness predictor is relatively small; but among them, the contribution of the 1,000-700hPa thickness predictor is relatively large, which is related to the fact that ground-based microwave radiometers mainly observe the radiation of the lower atmosphere. It is worth noting that surface pressure plays a key role in the bias correction of the temperature channel, which is the reason for the positive OMB bias observed at the plateau station.

[0053] S33, performing deviation correction on the original observed brightness temperature data according to the deviation of the OMB predicted in step S32 to obtain the corrected observed brightness temperature data.

[0054] Since the bias correction is implemented based on Python code, it cannot be mixed with the WRFDA code. The coupling between the two is not at the memory level, but through file reading and writing. WRFDA outputs the variables required for bias correction by calculating diagnostic information such as prediction factors, and the bias correction model outputs the correction results. Since WRFDA can only read the observation information y, which is not the bias that needs to be corrected, the present invention implements the bias that needs to be corrected on the original observed brightness temperature y. According to the formula ; Calculate the corrected observed brightness temperature , and output the corrected observation data, classify and summarize the multi-station data according to the instrument type, and output the required files for assimilation one by one, so as to obtain the corrected observation brightness temperature data.

[0055] (4) Three-dimensional variational assimilation a. Calculate the observation residual data based on the simulated brightness temperature data and the corrected observed brightness temperature data.

[0056] ; In the formula represents the observed residual, is the observed brightness temperature after bias correction, is the simulated brightness temperature.

[0057] b. Filter out the corrected observational brightness temperature data required for assimilation based on the observational residual data.

[0058] Due to the influence of instrument detection itself and the atmospheric environment, the observation data will inevitably have errors. Before entering the three-dimensional variational assimilation, certain technical means must be used to eliminate the data with large errors. The remaining data is the corrected observation brightness temperature data required for assimilation, which can ensure the quality of the observation data entering the assimilation module.

[0059] The quality control of the ground-based microwave radiometer in the embodiment of the present invention mainly includes the following: Observation quality control code: included in the original data, provided by the instrument manufacturer and observers, to eliminate data with problems in observation quality.

[0060] Absolute Deviation Test: Elimination Observations greater than 20K.

[0061] Relative deviation test: Elimination Observations with errors greater than 3 times.

[0062] Observation cloud and rain area detection: Based on the cloud and rain area markings, observations under cloud and rain environments are eliminated.

[0063] Simulated cloud and rain area detection: Eliminate observations where the cloud water integral of the entire layer at the observation point is greater than 0.2g / kg in the elimination mode.

[0064] c. Using the assimilation module of the WRFDA system, based on the background error covariance matrix, the observation error covariance matrix, the screened corrected observation brightness temperature, the observation operator and its tangent adjoint, the background field is iterated by three-dimensional variational minimization to finally obtain the analysis field.

[0065] In this embodiment, the target cost function of the minimization iteration is defined as: ; in, For the target price, and denote the analysis field and the background field respectively; is the corrected observed brightness temperature data; and denote the background error covariance matrix and the observation error covariance matrix respectively; Represents the observation operator.

[0066] d. Diagnostic file output.

[0067] Based on the background field and analysis field , and output diagnostic files such as the mean, root mean square error, etc. of OMB and OMA (observation minus analysis) in the observation space can be used to further evaluate the assimilation effect.

[0068] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A direct assimilation method for ground-based microwave radiometers, characterized in that: include: Read the original radiation observation data of the ground-based microwave radiometer and pre-process it to obtain the original observation brightness temperature data; Call the RTTOV-gb fast radiative transfer model, correspond the derived types in the WRFDA system to the RTTOV-gb derived types, so that the RTTOV, RTTOV-gb fast radiative transfer models and the WRFDA system are coupled simultaneously, and then the background field of the numerical weather forecast model is converted into simulated brightness temperature data; Based on the machine learning algorithm, the source of the deviation and the nonlinear relationship between the source and the deviation are considered. The deviation correction prediction factor is calculated according to the original observed brightness temperature data and the background field. The deviation of the original observed brightness temperature data is corrected based on the deviation correction prediction factor to obtain the corrected observed brightness temperature data. Calculate observational residual data based on simulated brightness temperature data and corrected observed brightness temperature data, and select corrected observed brightness temperature data required for assimilation based on the observed residual data; Using the assimilation module of the WRFDA system, the background field is subjected to three-dimensional variational assimilation processing based on the screened and corrected observational brightness temperature data to obtain the analysis field and output the diagnostic file.

2. The direct assimilation method for ground-based microwave radiometer according to claim 1, characterized in that: Preprocessing methods include: (a) Time zone conversion, unifying observation data into Coordinated Universal Time; (b) Check the temporal and spatial consistency, compare with the background field, and select the observation data required for assimilation time and region; (c) Cloud and rain detection: cloud area detection is performed on the selected observation data using the AGRI cloud classification product of the Fengyun-4B satellite, and rain area detection is performed on the selected observation data using the rain gauge data of the microwave radiometer; (d) Observation data output: The observation data after cloud and rain detection are classified and summarized according to the instrument type, and the required files for assimilation are output one by one to obtain the original observation brightness temperature data.

3. The direct assimilation method for ground-based microwave radiometer according to claim 1, characterized in that: The method for converting the background field of the numerical weather forecast model into simulated brightness temperature data specifically includes: reading the coefficient file of the ground-based microwave radiometer, reading the meteorological element data from the background field, performing horizontal and vertical spatial interpolation, and assigning them to the derived arrays corresponding to the RTTOV-gb model respectively, and calculating the simulated brightness temperature data of the ground-based microwave radiometer through RTTOV-gb.

4. The direct assimilation method for ground-based microwave radiometer according to claim 3, characterized in that: When introducing RTTOV-gb into the WRFDA system and building the ground-based microwave radiometer direct assimilation module, it is necessary to replace keywords in the RTTOV-gb source code, including replacing the file name of the code itself, as well as the definition and call names of the relevant interfaces, modules, and libraries in the code.

5. The direct assimilation method for ground-based microwave radiometer according to claim 1, characterized in that: The method for obtaining the corrected observation brightness temperature data based on the machine learning algorithm specifically includes: S31, WRFDA system calculates the bias correction prediction factor based on the original observed brightness temperature data and background field; S32, based on the bias correction prediction factor outputted from step S31, using the bias correction model obtained by pre-training to predict the bias of the observed brightness temperature minus the simulated brightness temperature, the bias correction model is constructed based on the random forest algorithm, which can be used in the variational assimilation framework, and takes into account the source of the bias and the nonlinear relationship between the bias source and the bias for the observation characteristics and terrain position of the ground-based microwave radiometer, and can evaluate the importance of the bias source, and is trained separately for different instruments and different channels during training; S33, performing deviation correction on the original observed brightness temperature data according to the deviation of the observed brightness temperature predicted in step S32 minus the simulated brightness temperature, to obtain the corrected observed brightness temperature data.

6. The direct assimilation method for ground-based microwave radiometer according to claim 5, characterized in that: The deviation correction prediction factors selected in step S32 include 1000-700hPa atmospheric thickness, 700-500hPa atmospheric thickness, 500-300hPa atmospheric thickness, 1000-300hPa atmospheric thickness, 200-50hPa atmospheric thickness, 2-meter temperature, 2-meter humidity, 10-meter meridional wind, 10-meter zonal wind, surface air pressure, surface temperature, atmospheric precipitable water, observed brightness temperature, longitude and latitude.

7. The direct assimilation method for ground-based microwave radiometer according to claim 1, characterized in that: For the corrected observational brightness temperature data, if it meets any of the exclusion conditions, it will be excluded, otherwise it is the corrected observational brightness temperature data required for assimilation; the exclusion conditions include: the corresponding observational residual data is greater than 20K; the corresponding observational residual data is greater than 3 times the error; it is the observation data under cloud and rain environment; the cloud water integral of the entire layer at the observation point is greater than 0.2g / kg.

8. The direct assimilation method for ground-based microwave radiometer according to claim 1, characterized in that: The three-dimensional variational assimilation processing is performed based on minimization iteration, and the objective cost function of the minimization iteration is defined as: ； in, For the target price, and denote the analysis field and the background field respectively; is the corrected observed brightness temperature data; and denote the background error covariance matrix and the observation error covariance matrix respectively; Represents the observation operator.

Images