Land assimilation system and method based on traffic weather application

Abstract

The invention discloses a land assimilation system based on traffic weather application. Comprising a data acquisition unit used for regularly acquiring meteorological data of an automatic observation station, a quality control unit used for implementing a quality control process on the data acquired by the automatic observation station, and a data fusion unit used for fusing the acquired meteorological data, the product generation unit is used for outputting the product data processed by the data fusion unit; through a high-quality data processing and fusion technology, an accurate land surface actual condition analysis field is used as driving data application, the performance of a regional numerical forecasting service system can be improved, automatic site observation is interpolated into grid data so as to obtain the actual condition analysis field, and on the basis that a land surface assimilation system parameterization scheme is determined, the real-time performance of the land surface assimilation system is improved. And key technologies such as an automatic station selection strategy, automatic station observation quality control and an interpolation method are updated, so that the quality of a national live analysis field with the resolution of 1-3 kilometers is further improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of meteorological analysis, in particular to a land surface assimilation system and method based on traffic meteorological application. BACKGROUND

[0002] Land surface meteorological live refers to real-time or recent meteorological observation data on the earth's surface, which reflects the atmospheric conditions near the ground. Technical personnel often combine observation data with model predictions to perform land surface data assimilation to generate the best estimate of the ground state. The core of land surface data assimilation is to effectively fuse different resolution direct and indirect multi-source observation data based on the dynamic framework of the land surface process model, and automatically adjust the land surface model based on observation data to reduce errors.

[0003] Since the 1990s, the research on land surface data assimilation systems has increased, and the use of "offline" land surface models to assimilate remote sensing and ground observation data has become the main feature of land surface assimilation systems. Based on the land surface assimilation system, better accuracy of land surface meteorological live analysis field data can be generated. This live analysis field data can not only be coupled into a meteorological numerical prediction model to improve prediction accuracy, but also be used as driving data for traffic meteorological models, thereby providing initial driving fields for traffic road prediction.

[0004] Land surface assimilation systems often have the following problems:

[0005] 1. The quality and quantity of ground observation data used by the land surface assimilation system, as well as the selection and application of interpolation methods, have a significant impact on the accuracy of land surface live analysis field grid products.

[0006] 2. The live analysis field data in existing land surface assimilation systems, whether coupled into a model or used to drive a model, has a high requirement for the calculation efficiency of the land surface assimilation system, which needs to shorten the lag time of product generation as much as possible to ensure that users can quickly obtain the land surface live analysis field results at the current time.

[0007] Therefore, one of the problems to be solved by those skilled in the art is to use more efficient calculation methods and strategies to improve the timeliness of live analysis fields. SUMMARY

[0008] To solve the above problems, the present application provides the following technical solutions:

[0009] The land surface assimilation system based on traffic meteorological application comprises

[0010] A data acquisition unit for regularly acquiring meteorological data from automatic observation stations;

[0011] a quality control unit for implementing a quality control process on data collected by the automatic observation station;

[0012] a data fusion unit for fusing the collected meteorological data;

[0013] a product generation unit for outputting product data processed by the data fusion unit;

[0014] an inspection and evaluation unit for objectively inspecting and evaluating the grid data product of the live analysis field.

[0015] a land surface assimilation method based on traffic meteorological application, which is implemented by using the land surface assimilation system based on traffic meteorological application, and includes the following steps:

[0016] S1, the data collection unit collects real-time data from the ground meteorological observation station (AWS) every day, and receives the forecast field data of the CMA-BJ (CMA-BJ) through an interface or / and a file transmission mode;

[0017] S2, the data fusion unit optimizes the data collected by the ground meteorological observation station:

[0018] S2-1, the observation data is integrated, and the missing data and invalid marks in the standard of the World Meteorological Organization are marked and removed;

[0019] S2-2, the observation data of the same time and station are combined, and the repeated data are deleted;

[0020] S2-3, the extreme value check is performed on the results, and the observations outside the numerical interval are removed;

[0021] S2-4, the spatial consistency check is performed, the Barnes spatial consistency check method is used, the spherical distance between the checked point and other adjacent points is calculated first, so that the distance weight of each adjacent point is calculated, and finally the difference between the weighted average value of the physical quantity of the checked point and the physical quantity of all adjacent points around the checked point is calculated, and the checked point is removed when the difference is greater than 4 times the standard deviation;

[0022] S3, the data fusion unit fuses the optimized data to generate a land surface meteorological live analysis field grid data product;

[0023] S4, the inspection and evaluation unit selects the CMA-BJ prediction data as backup to make up for the missing observation data.

[0024] On the basis of the above technical solutions, the application can also be improved as follows.

[0025] Further, the.

[0026] Further, the.

[0027] Further, in the step S1, the real-time data collected by the data collection unit includes air temperature, air pressure, relative humidity, wind speed, precipitation of the ground meteorological observation station (AWS), and shortwave radiation data of CMA-BJ.

[0028] Further, in the step S1, the data collection unit starts the main program every hour at 21 minutes to obtain the CMA-BJ and AWS data.

[0029] Further, in the step S2-3, the relative humidity interval value is 0-100%, the wind speed interval value is 0-70 m / s, and the wind direction interval value is 0-359°.

[0030] Further, in the step S3, the data fusion scheme of the data fusion unit is a thin plate smoothing spline model and an improved Cressman weight revision.

[0031] Further, the land surface weather real-time analysis field grid data product includes near-surface air temperature, specific humidity, wind speed, air pressure, precipitation, shortwave radiation, and longwave radiation, wherein the downward shortwave radiation is directly interpolated by CMA-BJ data, and the downward longwave radiation data is calculated by air temperature, air pressure, and relative humidity through an empirical formula.

[0032] Further, in the step S3, the altitude observation is an auxiliary variable of the spline model, and the relative humidity, wind speed, and precipitation are CMA-BJ data as auxiliary variables of the spline model.

[0033] Further, in the step S4, the evaluation method of the objective test and evaluation of the analysis field grid data product by the test and evaluation unit is as follows:

[0034] S4-1, the data collection unit is started every hour at 52 minutes to collect the near-surface atmospheric data predicted by CMA-BJ;

[0035] S4-2, the ground meteorological automatic station observation data not entering the data fusion unit in the step S3 are used as independent observation to compare with the real-time analysis field grid data output, and the bias (ME) and root mean square error (RMSE) of the values of various meteorological elements in the real-time analysis field grid data product are calculated, and the calculation formula is as follows:

[0036]

[0037] Wherein, n is the number of observation stations, O i is observation, A i is the corresponding element value of the real-time analysis field.

[0038] Beneficial effects

[0039] Compared with the prior art, the application can improve the performance of the regional numerical prediction business system by using high-quality data processing and fusion technology and accurate land surface real-time analysis field as driving data, and can further improve the quality of the real-time analysis field with a resolution of 1-3 kilometers nationwide on the basis of the parameterization scheme of the land surface assimilation system by updating the key technologies such as automatic station selection strategy, automatic station observation quality control and interpolation method.

[0040] The real-time analysis field data product takes into account the timeliness and accuracy, and can not only provide a high-precision land driving field for the (business) prediction system, but also provide an atmospheric driving field or real-time analysis grid product for industry meteorological services in real time. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed for the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0042] Figure 1 A flowchart for generating a real-time analysis field grid data product is shown in Figure 4.

[0043] Figure 2 A system framework diagram of the application is shown in Figure 5.

[0044] Figure 3 A quality control process diagram of automatic station observation data is shown in Figure 6. DETAILED DESCRIPTION

[0045] In order to facilitate the understanding of the application, the application will be described more fully below with reference to the related drawings. The drawings show several embodiments of the application. However, the application can be realized in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the application more thorough and comprehensive.

[0046] The land surface assimilation system based on traffic meteorological application includes a data acquisition unit for acquiring meteorological data of automatic observation stations at regular intervals, a quality control unit for implementing a quality control process on the data collected by the automatic observation stations, a data fusion unit for fusing the collected meteorological data, a product generation unit for outputting the product data processed by the data fusion unit, and a verification and evaluation unit for objective verification and evaluation of the real-time analysis field grid data product.

[0047] The system updates the data hourly, which is meteorological driving data, mainly relying on real-time meteorological data from ground meteorological observation stations (AWS) and the forecast field data from the proprietary unit's CMA-BJ short-term numerical model. The main program is started every day at 21 minutes per hour on a high-performance computer cluster, CMA-BJ and AWS data are obtained, fused by a thin plate spline model, and locally corrected. Downward shortwave radiation is directly interpolated using CMA-BJ data, and downward longwave radiation is calculated using an empirical formula.

[0048] However, due to the large amount of calculation of the 1-3 km resolution grid data nationwide, in order to ensure the timeliness of the product, data blocking and multi-variable multi-task parallel computing methods are used, combined with the use of dependent job scripts to control the entire process to improve computing efficiency, and a matrix method is used instead of the conventional loop correction method for nationwide unified correction to improve the reporting efficiency of the product.

[0049] After the system operation scheme is updated, the generation time of real-time data products is not more than 35 minutes after the whole point.

[0050] In addition, the multi-time CMA-BJ forecast is used as backup data for automatic station observation to make up for the lack of observation data and further improve the stability of system operation. Then the prepared driving data is written into the format required by the land surface assimilation system. After the operation of the land surface assimilation system, grid data products including ground temperature, humidity, wind speed, wind direction, pressure, radiation and precipitation can be output.

[0051] At the same time, due to the real-time operation of the land surface assimilation system, the number of automatic station observation sites nationwide is more than 60,000 (encryption stations), and the average spatial resolution of data is about 3 km. However, the quality of most automatic station observations needs to be improved, so the quality of automatic station observation data needs to be controlled.

[0052] At present, the time resolution of mature high-resolution regional numerical prediction business systems in China is 1-3 hours, and the spatial resolution can reach 3-9 kilometers. The accurate land real-time analysis field as driving data can improve the performance of regional numerical prediction business systems. In addition, the road, traffic and other industry meteorological services have a strong demand for high-resolution land real-time analysis field. Therefore, focusing on traffic meteorological applications, based on the CMA-BJ land assimilation system framework and dense automatic station observation information, a 1-3 km resolution ground real-time analysis field is developed, which can not only provide high-precision land driving field for (business) prediction systems, but also provide real-time atmospheric driving field or real-time analysis grid products for industry meteorological services.

[0053] The observation of automatic station is interpolated into grid data based on the RegCM-Land assimilation system to obtain the real-time analysis field. The factors affecting the effect of the real-time analysis field of the land assimilation system mainly include the model parameters of the land assimilation system, the interpolation scheme and the quality of the observation data. Therefore, on the basis of determining the parameterization scheme of the land assimilation system, the quality of the real-time analysis field with a resolution of 1-3 km nationwide is further improved through the update of key technologies such as automatic station selection strategy, automatic station observation quality control and interpolation method. The real-time analysis field data product takes into account the timeliness and accuracy.

[0054] In order to explain the above technical solutions in detail, a land assimilation method based on traffic meteorological application is disclosed, comprising the following steps:

[0055] S1, the data acquisition unit collects real-time data from the ground meteorological observation station (AWS) every day, and receives the forecast field data of the CMA-BJ through the interface or / and file transmission mode;

[0056] The real-time data collected by the data acquisition unit includes the air temperature, air pressure, relative humidity, wind speed, precipitation of the ground meteorological observation station (AWS) and the shortwave radiation data of CMA-BJ. At the same time, the data acquisition unit starts the main program every hour at 21 minutes every day to obtain the CMA-BJ and AWS data;

[0057] S2, the data fusion unit optimizes the data collected by the ground meteorological observation station:

[0058] S2-1, the observation data is integrated, and the missing data and invalid symbols in the standard of the World Meteorological Organization are marked out;

[0059] Among them, the data integration inspection includes time window inspection, observation area inspection and total number of observation data inspection of observation data, the relative humidity interval value is 0-100%, the wind speed interval value is 0-70 m / s, and the wind direction interval value is 0-359°;

[0060] S2-2, the observation data of the same time and station are combined, and the repeated data are deleted;

[0061] S2-3, the extreme value inspection is carried out for the result, and the observation outside the numerical interval is removed;

[0062] S2-4, the spatial consistency inspection is carried out, the Barnes spatial consistency inspection method is adopted, the spherical distance between the inspected point and other adjacent points is calculated first, so as to calculate the distance weight of each adjacent point, and finally the difference between the weighted average value of the physical quantity of the inspected point and all adjacent points is calculated, and the inspected point is removed when the difference is greater than 4 times the standard deviation;

[0063] S3, the data fusion unit fuses the optimized data to generate land surface weather real-time analysis field grid data products;

[0064] The data fusion scheme of the data fusion unit is a thin plate smoothing spline model and an improved Cressman weight revision. Since the temperature, pressure and altitude of the five variables observed by the meteorological ground station are closely related, the altitude observation is taken as an auxiliary variable of the spline model, the relative humidity, wind speed and precipitation are taken as auxiliary variables of the spline model by using the CMA-BJ data, and the trend surface is established by using the thin plate smoothing spline model. Since the spline surface has a certain smoothness, the estimated value of the trend surface at the station and the observation at the point are not consistent. In order to improve the product accuracy of the grid points around the observation station, the trend surface needs to be revised by using the improved Cressman weight.

[0065] S4, the test and evaluation unit selects the CMA-BJ forecast data as backup to make up for the missing observation data;

[0066] S4-1, the data acquisition unit is started every hour at 52 minutes of each day to collect the near-surface atmospheric data of the CMA-BJ forecast;

[0067] S4-2, the observation data of the ground meteorological automatic station which does not enter the data fusion unit in step S3 is used as independent observation to compare with the output real-time analysis field grid data, and the bias (ME) and root mean square error (RMSE) of each meteorological element value in the real-time analysis field grid data product are calculated, and the calculation formula is as follows:

[0068]

[0069] Wherein, n is the number of observation stations, O i is observation, A i is the corresponding element value of the real-time analysis field.

[0070] The above describes only the preferred specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can make equivalent replacement or change according to the technical scheme and the inventive concept of the present application within the technical range disclosed by the present application, which should be covered in the protection scope of the present application.

Claims

1. A land surface assimilation system based on traffic meteorology applications, characterized in that, include The data acquisition unit is used to collect meteorological data from automatic observation stations at regular intervals; The quality control unit is used to implement quality control procedures on the data collected by the automatic observation station. The data fusion unit is used to fuse the collected meteorological data; The product generation unit is used to output product information processed by the data fusion unit. The inspection and evaluation unit is used for the objective inspection and evaluation of real-world analysis field grid data products.

2. A land surface assimilation method based on traffic meteorology applications, implemented using the land surface assimilation system based on traffic meteorology applications as described in claim 1, comprising the following steps: S1. The data acquisition unit collects real-time data from the ground meteorological observation station (AWS) on a daily schedule and receives forecast field data from the Beijing Urban Meteorological Research Institute's Ruitu-Short-Term Rapid Update Cyclic Numerical Prediction System (CMA-BJ) via interface and / or file transfer. S2, the data fusion unit optimizes the data collected by the ground meteorological observation station: S2-1. Overall observational data, marking missing data and removing invalid labels from the standards set by the World Meteorological Organization; S2-2. Merge observation data from the same time and station, and remove duplicate data; S2-3. Perform extreme value checks on the results and eliminate observations outside the numerical range; S2-4. Spatial consistency check: The Barnes spatial consistency check method is adopted. First, the spherical distance between the checked point and other adjacent points is calculated, thereby calculating the distance weight of each adjacent point. Finally, the difference between the physical quantity of the checked point and the weighted average of the physical quantities of all surrounding adjacent points is calculated. When the difference is greater than 4 times the standard deviation, the checked point is removed. S3, the data fusion unit fuses the optimized data to generate a grid data product of the land surface meteorological condition analysis field; S4. The verification and evaluation unit selects CMA-BJ forecast data as a backup to make up for missing observation data.

3. The land surface assimilation method based on traffic meteorology applications according to claim 2, characterized in that: In step S1, the real-time data collected by the data acquisition unit includes air temperature, air pressure, relative humidity, wind speed, wind direction, precipitation from the ground meteorological observation station (AWS), as well as shortwave radiation data from CMA-BJ.

4. The land surface assimilation method based on traffic meteorology application according to claim 3, characterized in that: In step S1, the data acquisition unit starts the main program at 21 minutes past every hour to acquire CMA-BJ and AWS data.

5. The land surface assimilation method based on traffic meteorology application according to claim 4, characterized in that: In step S2-1, the overall data check includes checking the time window of the observation data, checking the observation area, and checking the total number of observation data.

6. The land surface assimilation method based on traffic meteorology application according to claim 5, characterized in that: In steps S2-3, the relative humidity range is 0-100%, the wind speed range is 0-70m / s, and the wind direction range is 0-359°.

7. A land surface assimilation method based on traffic meteorology applications according to claim 6, characterized in that: In step S3, the data fusion scheme of the data fusion unit is a thin plate smooth spline model and an improved Cressman weight correction.

8. A land surface assimilation method based on traffic meteorology applications according to claim 7, characterized in that: The land surface meteorological data analysis field grid products include near-surface air temperature, specific humidity, wind speed, air pressure, precipitation, shortwave radiation, and longwave radiation. Among them, downward shortwave radiation is directly interpolated using CMA-BJ data, while downward longwave radiation data is calculated using air temperature, air pressure, and relative humidity through empirical formulas.

9. A land surface assimilation method based on traffic meteorology applications according to claim 8, characterized in that: In step S3, altitude observation is used as an auxiliary variable for the spline model, while relative humidity, wind speed, and precipitation are used as auxiliary variables for the spline model using CMA-BJ data.

10. A land surface assimilation method based on traffic meteorology applications according to claim 9, characterized in that: In step S4, the evaluation method for the objective inspection and evaluation of the analysis field grid data product by the inspection and evaluation unit is as follows: S4-1: The data acquisition unit is activated at 52 minutes past the hour every day to collect near-surface atmospheric data from the CMA-BJ forecast. S4-2. Using the observation data from the automatic weather stations that did not enter the data fusion unit in step S3 as independent observations, compare them with the output real-time analysis field grid data to calculate the deviation (ME) and root mean square error (RMSE) of various meteorological element values ​​in the real-time analysis field grid data product. The calculation formula is as follows: Where n is the number of observation stations, O i It is observation, A i These are the element values ​​of the corresponding real-world analysis field.

Images