Methods, apparatus, electronic equipment and program products for testing forecast models

By constructing a multi-dimensional verification algorithm based on meteorological elements, and calculating verification index values ​​and degradation factors, the problem of low verification accuracy of forecast models in existing technologies is solved, enabling comprehensive evaluation and improvement guidance of forecast models.

CN120780941BActive Publication Date: 2026-11-17ZHONGKE TIANJI METEOROLOGICAL TECH CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202510887771.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-06-27
Publication Date
2026-11-17
Estimated Expiration
2045-06-27

AI Technical Summary

Technical Problem

In existing technologies, the verification methods for forecast models mainly rely on static thresholds and single indicators, resulting in low verification accuracy and an inability to comprehensively assess the consistency of spatial structure and the coupling relationship of multiple elements.

Method used

By acquiring observational and forecast data for the target area and time period, different verification algorithms are constructed based on meteorological elements to calculate verification index values ​​and degradation factors, generate visualized verification results, and improve the adaptability and accuracy of the verification.

Benefits of technology

It enables multi-dimensional evaluation of forecast models, allowing for the monitoring and assessment of degradation processes, improving the accuracy and reliability of verification, and providing directional guidance for improving forecast models.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120780941B_ABST
    Figure CN120780941B_ABST
Patent Text Reader

Abstract

This application provides a method, apparatus, electronic device, and program product for verifying forecasts from forecast models. The method includes: acquiring target parameters input by a user, including a target area, meteorological elements, and a target time period; determining multiple observational data points for the target area within the target time period and multiple forecast data points corresponding to multiple forecast models; performing verification processing on the multiple observational data points and multiple forecast data points based on the meteorological elements to obtain a verification index value corresponding to each forecast model; determining a degradation factor corresponding to each forecast model based on the verification index value, whereby the degradation factor represents the driving contribution of meteorological elements to the degradation of forecast skill under the forecast model; and generating visualized verification results corresponding to each of the multiple forecast models based on the multiple verification index values ​​and multiple degradation factors, thereby improving the accuracy of verifying forecast models.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of weather forecasting, and more particularly to a method, apparatus, electronic equipment, and program product for verifying forecasts from forecast models. Background Technology

[0002] With the surge in multi-source heterogeneous forecast data (such as global models, regional models, and Grib / NetCDF format data released by different institutions), the need to test numerical weather prediction models has shifted from a single numerical error assessment to a comprehensive consideration of spatial structure consistency, multi-factor coupling relationships, and long-term stability.

[0003] Currently, the evaluation of forecast models can be based on traditional statistical indicators, comparing the numerical differences between forecast and observed values. For example, for a model of a specific region, using data in a certain format, significance testing can be conducted by calculating the regional mean.

[0004] In the aforementioned technologies, current testing methods typically employ static thresholds and single indicators, resulting in low accuracy of the testing and forecasting models. Summary of the Invention

[0005] This application provides a method, apparatus, electronic device, and program product for verifying forecasts from forecast models, in order to solve the technical problem of low accuracy in verifying forecast models in the prior art.

[0006] Firstly, this application provides a method for verifying forecasts from forecast models, the method comprising:

[0007] Obtain the target parameters input by the user, including the target area, meteorological elements, and target time period;

[0008] Determine multiple forecast data corresponding to multiple observation data and multiple forecast models for the target area within the target time period;

[0009] Based on the meteorological elements, the multiple observation data and the multiple forecast data are verified and processed to obtain the verification index value corresponding to each forecast model;

[0010] Based on the test index value, the degradation factor corresponding to each forecast model is determined. The degradation factor is used to represent the driving contribution of the meteorological elements to the degradation of forecast skills under the forecast model.

[0011] Based on multiple test index values ​​and multiple degradation factors, visualized test results corresponding to multiple forecast models are generated.

[0012] In this way, different verification methods are constructed based on different meteorological elements, which improves the adaptability of verification and determines the degradation factors corresponding to different regions of each forecast model, so as to realize the monitoring and evaluation of the degradation process and improve the accuracy of the verification forecast model.

[0013] Optionally, according to the method described above, for any forecast model, based on the meteorological elements, the multiple observation data and the multiple forecast data are tested to obtain the test index value corresponding to the forecast model, including:

[0014] Based on the meteorological elements, the target algorithm is determined from multiple verification algorithms;

[0015] According to the target algorithm, the multiple observation data and the multiple forecast data are verified to obtain the first index value corresponding to the forecast model;

[0016] The first index value is resampled to obtain the test index value corresponding to the forecast model.

[0017] In this way, different testing methods are constructed based on different meteorological elements, which improves the adaptability of the testing and the accuracy of the testing forecast model.

[0018] Optionally, in the method described above, the meteorological elements include visibility, precipitation, temperature, zonal wind, meridional wind, and humidity; the multiple testing algorithms include a dynamic threshold testing algorithm and a continuous variable testing algorithm; and a target algorithm is determined from among the multiple testing algorithms based on the meteorological elements, including:

[0019] If the meteorological element is visibility or precipitation, then the target algorithm is determined to be a dynamic threshold verification algorithm;

[0020] If the meteorological elements are temperature, zonal wind, meridional wind, and humidity, then the target algorithm is determined to be a continuous variable testing algorithm.

[0021] In this way, different testing methods are constructed based on different meteorological elements, which improves the adaptability of the testing and the accuracy of the testing forecast model.

[0022] Optionally, in the method described above, the target algorithm is a dynamic threshold verification algorithm. Based on the target algorithm, the plurality of observation data and the plurality of forecast data are verified to obtain a first index value corresponding to the forecast model, including:

[0023] Get the preset half width of the sliding window;

[0024] The first time period is determined based on the target time period and the preset sliding window half width;

[0025] Based on the preset quantile level and multiple observation data corresponding to the first time period, determine the dynamic threshold corresponding to the time point;

[0026] Based on the dynamic threshold, multiple observation data and multiple forecast data corresponding to the first time period are classified and processed to obtain the observation level of multiple observation data and the forecast level of multiple forecast data.

[0027] Based on the observation levels of multiple observation data and the forecast levels of multiple forecast data, the verification data corresponding to the time point is determined. The verification data includes multiple verification levels, as well as the number of hits, missed reports, false alarms, and correct rejections corresponding to each verification level.

[0028] Based on the test data corresponding to multiple time points, the first index value corresponding to the forecast model is determined.

[0029] In this way, by adopting a dynamic threshold test algorithm to determine the dynamic threshold, the adaptability to nonlinear variables and regional heterogeneity is improved, thereby enhancing the accuracy of the test prediction model.

[0030] Optionally, the method described above, determining the first index value corresponding to the forecast model based on test data corresponding to multiple time points, includes:

[0031] According to the order of multiple inspection levels from smallest to largest, the level inspection data corresponding to multiple inspection levels are determined sequentially from the inspection data corresponding to multiple time points. The level inspection data includes the number of hits, the number of missed reports, the number of false alarms, and the number of correct rejections corresponding to the same inspection level at multiple time points.

[0032] For any given inspection level, the inspection score for that inspection level is determined based on the inspection data corresponding to that inspection level.

[0033] The first indicator value corresponding to the forecasting model is determined based on the inspection scores corresponding to multiple inspection levels and the weight values ​​corresponding to each of the multiple inspection levels.

[0034] In this way, by comprehensively considering and weighting multi-dimensional data at different verification levels, the performance of the forecast model can be fully evaluated, providing a clear direction for the improvement of the forecast model, thereby improving the accuracy and reliability of the forecast model.

[0035] Optionally, in the method described above, the target algorithm is a continuous variable testing algorithm. Based on the target algorithm, the multiple observation data and the multiple forecast data are tested to obtain the first index value corresponding to the forecast model, including:

[0036] Obtain the latitude-weighted value corresponding to the target region;

[0037] Based on the multiple observation data, the multiple forecast data, and the latitude weighted value, the target algorithm determines multiple first index values ​​corresponding to the forecast model. The first index values ​​include anomaly correlation coefficient, deviation, and root mean square error.

[0038] In this way, by using the continuous variable testing algorithm, the system can evaluate the spatial consistency of the forecast model and the performance of large-scale weather systems, thereby improving the accuracy of the forecast model testing.

[0039] Optionally, the method described above, for any forecast model, determines the degradation factor corresponding to the forecast model based on the test index value, including:

[0040] Determine the type of indicator corresponding to the test indicator value;

[0041] Determine the target formula and reference value based on the indicator type;

[0042] Based on the test index value and the reference value, the degradation index corresponding to the test index value is determined by the target formula; based on the degradation index and the weight values ​​of multiple stations in the target area, the degradation factor corresponding to the forecast model is determined.

[0043] In this way, the degradation factor can be used to dynamically assess the degradation trend of forecasting skills as forecast lead time decreases, thereby providing directional guidance for model improvement.

[0044] Secondly, this application provides a verification device for forecasts from forecast models, the device comprising:

[0045] The acquisition module is used to acquire target parameters input by the user, including target area, meteorological elements, and target time period;

[0046] The determination module is used to determine multiple observation data and multiple forecast data corresponding to multiple forecast models for the target area within the target time period;

[0047] The first verification module is used to verify the multiple observation data and the multiple forecast data according to the meteorological elements, and obtain the verification index value corresponding to each forecast model.

[0048] The second verification module is used to determine the degradation factor corresponding to each forecast model based on the verification index value. The degradation factor is used to represent the driving contribution of the meteorological elements to the degradation of forecast skills under the forecast model.

[0049] The display module is used to generate visual test results corresponding to multiple forecast models based on multiple test index values ​​and multiple degradation factors.

[0050] Optionally, in the apparatus described above, the first inspection module is specifically used for:

[0051] Based on the meteorological elements, the target algorithm is determined from multiple verification algorithms;

[0052] According to the target algorithm, the multiple observation data and the multiple forecast data are verified to obtain the first index value corresponding to the forecast model;

[0053] The first index value is resampled to obtain the test index value corresponding to the forecast model.

[0054] Optionally, in the apparatus described above, the meteorological elements include visibility, precipitation, temperature, zonal wind, meridional wind, and humidity; the multiple testing algorithms include a dynamic threshold testing algorithm and a continuous variable testing algorithm; and the first testing module is specifically used for:

[0055] If the meteorological element is visibility or precipitation, then the target algorithm is determined to be a dynamic threshold verification algorithm;

[0056] If the meteorological elements are temperature, zonal wind, meridional wind, and humidity, then the target algorithm is determined to be a continuous variable testing algorithm.

[0057] Optionally, in the apparatus described above, the target algorithm is a dynamic threshold testing algorithm, and the first testing module is specifically used for:

[0058] Get the preset half width of the sliding window;

[0059] The first time period is determined based on the target time period and the preset sliding window half width;

[0060] Based on the preset quantile level and multiple observation data corresponding to the first time period, determine the dynamic threshold corresponding to the time point;

[0061] Based on the dynamic threshold, multiple observation data and multiple forecast data corresponding to the first time period are classified and processed to obtain the observation level of multiple observation data and the forecast level of multiple forecast data.

[0062] Based on the observation levels of multiple observation data and the forecast levels of multiple forecast data, the verification data corresponding to the time point is determined. The verification data includes multiple verification levels, as well as the number of hits, missed reports, false alarms, and correct rejections corresponding to each verification level.

[0063] Based on the test data corresponding to multiple time points, the first index value corresponding to the forecast model is determined.

[0064] Optionally, in the apparatus described above, the first inspection module is specifically used for:

[0065] According to the order of multiple inspection levels from smallest to largest, the level inspection data corresponding to multiple inspection levels are determined sequentially from the inspection data corresponding to multiple time points. The level inspection data includes the number of hits, the number of missed reports, the number of false alarms, and the number of correct rejections corresponding to the same inspection level at multiple time points.

[0066] For any given inspection level, the inspection score for that inspection level is determined based on the inspection data corresponding to that inspection level.

[0067] The first indicator value corresponding to the forecasting model is determined based on the inspection scores corresponding to multiple inspection levels and the weight values ​​corresponding to each of the multiple inspection levels.

[0068] Optionally, in the apparatus described above, the target algorithm is a continuous variable testing algorithm, and the first testing module is specifically used for:

[0069] Obtain the latitude-weighted value corresponding to the target region;

[0070] Based on the multiple observation data, the multiple forecast data, and the latitude weighted value, the target algorithm determines multiple first index values ​​corresponding to the forecast model. The first index values ​​include anomaly correlation coefficient, deviation, and root mean square error.

[0071] Optionally, in the apparatus described above, for any given forecast, the second verification module is specifically used for:

[0072] Determine the type of indicator corresponding to the test indicator value;

[0073] Determine the target formula and reference value based on the indicator type;

[0074] Based on the test index value and the reference value, the degradation index corresponding to the test index value is determined by the target formula.

[0075] The degradation factor corresponding to the forecast model is determined based on the degradation index and the weight values ​​of multiple stations in the target area.

[0076] Thirdly, this application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0077] The memory stores computer-executed instructions;

[0078] The processor executes computer execution instructions stored in the memory to implement the method described in any of the first aspects.

[0079] Fourthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any of the first aspects.

[0080] Fifthly, this application provides a computer program product, including a computer program that, when executed by a computer, implements the method as described in any one of the first aspects.

[0081] The forecast verification method, apparatus, electronic equipment, and program product provided in this application acquire user-input target parameters, including target area, meteorological elements, and target time period. It then determines multiple observational data points for the target area within the target time period and multiple forecast data points corresponding to each forecast model. Based on the meteorological elements, it performs verification processing on the multiple observational and forecast data points to obtain verification index values ​​for each forecast model. Based on the verification index values, it determines the degradation factor corresponding to each forecast model. The degradation factor represents the driving contribution of meteorological elements to the degradation of forecast skill under the forecast model. Based on the multiple verification index values ​​and multiple degradation factors, it generates visualized verification results corresponding to each forecast model. In this way, different verification methods are constructed based on different meteorological elements, improving the adaptability of the verification process. Furthermore, by determining the degradation factor corresponding to each forecast model, it enables the monitoring and evaluation of the degradation process, thereby improving the accuracy of the verified forecast models. Attached Figure Description

[0082] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0083] Figure 1 This is a schematic diagram of the structure of an inspection system provided in an embodiment of this application;

[0084] Figure 2 A flowchart illustrating a method for verifying a forecast pattern provided in an embodiment of this application;

[0085] Figure 3 A flowchart illustrating another method for verifying a forecast pattern provided in this application embodiment;

[0086] Figure 4 A flowchart illustrating another method for verifying a forecast pattern provided in an embodiment of this application;

[0087] Figure 5 A schematic diagram of the structure of a forecast pattern verification device provided in an embodiment of this application;

[0088] Figure 6This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0089] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0090] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0091] It should be noted that although the terms "first," "second," etc., are used to describe various types of information in the embodiments of this application, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. Optionally, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information.

[0092] It should be understood that the terms "comprising" or "including" indicate the presence of the previously mentioned features, steps, or operations, but do not preclude the presence, occurrence, or addition of one or more other features, steps, or operations. The terms "and / or," etc., used in this application can be interpreted as inclusive, or mean any one or any combination thereof. Optionally, "A and / or B" means "any one of the following: A; B; A and B." Additionally, the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0093] With the surge in multi-source heterogeneous forecast data (such as global models, regional models, and Grib / NetCDF format data released by different institutions), the need to test numerical weather prediction models has shifted from a single numerical error assessment to a comprehensive consideration of spatial structure consistency, multi-factor coupling relationships, and long-term stability.

[0094] Currently, the evaluation of forecast models can be based on traditional statistical indicators, comparing the numerical differences between forecast and observed values. For example, for a model of a specific region, using data in a certain format, significance testing can be conducted by calculating the regional mean.

[0095] In the aforementioned technologies, current testing methods typically employ static thresholds and single indicators, resulting in low accuracy of the testing and forecasting models.

[0096] To address the aforementioned technical problems, this application provides a method for verifying forecast models. This method involves determining multiple observational data points for a target area within a target time period and multiple forecast data points corresponding to different forecast models. Based on meteorological elements, the method performs verification processing on these data points to obtain verification index values ​​for each forecast model. Based on these index values, a degradation factor for each forecast model is determined. Finally, based on these multiple index values ​​and degradation factors, visualized verification results for each forecast model are generated. This approach constructs different verification methods based on different meteorological elements, improving the adaptability of the verification process. Furthermore, by determining the degradation factor for each forecast model, the method enables the monitoring and evaluation of the degradation process, thereby improving the accuracy of forecast model verification.

[0097] Below, in conjunction with Figure 1 An example is given to illustrate the verification system for the forecast model.

[0098] Figure 1 This is a schematic diagram of an inspection system provided in an embodiment of this application. Please refer to... Figure 1 , Figure 1 It can include a data layer, an indicator calculation layer, an analysis layer, and a visualization layer.

[0099] The data layer can be used to automatically download multiple forecast data and observation data from multiple forecast models, and supports various data formats, different pressure levels, and multidimensional meteorological elements. For example, multiple data formats can include Grib, NetCDF, and CSV formats. Different pressure levels can include 200 hPa, 500 hPa, 700 hPa, 850 hPa, and 925 hPa. Multidimensional meteorological elements can include precipitation, temperature, zonal wind, meridional wind, and humidity.

[0100] The data layer can also be used to download and reanalyze data.

[0101] The data downloaded by the data layer can have different spatiotemporal resolutions.

[0102] The data layer can automate data downloads through a timed task scheduling mechanism, ensuring the acquisition of real-time and accurate forecast and observation data. This significantly reduces the complexity of manual operations and greatly improves the real-time performance of data acquisition and system processing efficiency.

[0103] After the data download is complete, the system can use a preset program script to perform data quality control.

[0104] Data quality control can include deleting abnormal files, checking for data anomalies and deleting files containing abnormal data, and repairing or correcting abnormal data that can be repaired.

[0105] For example, abnormal files are files that cannot be opened normally, or files with less data than normal.

[0106] After deleting a file, the data layer can resubmit the data download task for download. If the resubmission fails three times, the download task will not be submitted again.

[0107] Data quality control can ensure the accuracy and reliability of input data and avoid deviations in test results due to data defects.

[0108] The data layer can also be used to convert downloaded data into a uniform file format, data format, and spatiotemporal resolution, and to align the data with a spatial grid structure.

[0109] For example, a single meteorological element of a pressure layer in each forecast model can be saved as a data file with NetCDF data format, hourly temporal resolution, and 0.1°x0.1° spatial resolution.

[0110] The data layer overcomes the compatibility issues caused by inconsistent data formats in existing technologies by standardizing data from various forecast models from different sources, significantly improving the efficiency and accuracy of data processing. The data layer successfully solves the technical challenges of diverse data sources and inconsistent formats, enabling unified management and analysis of multiple data sources.

[0111] The index calculation layer can be used to quantify the forecast performance of multiple forecast models.

[0112] The indicator calculation layer can include multiple test indicators.

[0113] For example, multiple test indicators can include root mean square error, standard deviation, correlation coefficient, threat score, and reliability.

[0114] The analysis layer can be used to diagnose and attribute the errors of multiple forecast models based on multiple test index values.

[0115] For example, the analysis layer can quantify the driving contribution of various meteorological factors to the degradation of forecasting skills at different spatiotemporal scales, construct a traceable and interpretable degradation factor map, and provide accurate reference for subsequent meteorological model improvement.

[0116] The visualization layer can be designed based on a lightweight B / S architecture, supporting high-performance interactive operation on the web page, and has flexible layer switching, legend zooming and parameter setting capabilities, adapting to multi-terminal browsing environments.

[0117] The visualization layer can have a "one-click skill report generation" function, which automatically summarizes the core assessment results and visualization graphics to form standardized or customized output reports, greatly improving the automation and intelligence level of the business assessment process.

[0118] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0119] The technical solutions shown in this application will now be described in detail through specific embodiments. It should be noted that the following embodiments may exist independently or in combination with each other; for the same or similar content, the description will not be repeated in different embodiments.

[0120] Figure 2 This is a flowchart illustrating a method for verifying a forecast pattern provided in an embodiment of this application. The execution entity in this embodiment can be a processor. The processor can be implemented in software or a combination of software and hardware. Please refer to... Figure 2 The method includes:

[0121] S201. Obtain the target parameters input by the user.

[0122] The target parameters include the target area, meteorological elements, and target time period.

[0123] The target area can be a location that needs to be tested with multiple forecast models.

[0124] Meteorological elements can include precipitation, temperature, zonal wind, meridional wind, and humidity.

[0125] The target period can be a historical period that needs to be tested with multiple forecast models.

[0126] Target parameters can be used to describe the parameters that users want to perform data analysis on.

[0127] An interactive interface can be displayed, which may include controls such as text boxes, selection boxes, and sliders. In response to user interaction with the target area selection control, meteorological element selection control, sliding operation of the slider corresponding to the target time period, and clicking operation of the submit button, the control values ​​corresponding to multiple controls are obtained, and the target parameters input by the user are determined based on the multiple control values.

[0128] The target parameters input by the user can be obtained through the command-line interface.

[0129] It should be noted that the target parameters input by the user can be obtained according to any feasible implementation method, and the embodiments of this application do not limit this.

[0130] S202. Determine multiple observation data and multiple forecast data corresponding to multiple forecast models for the target area within the target time period.

[0131] The observation data can be physical quantities such as temperature, humidity, and wind speed collected by equipment such as ground weather stations and automatic weather stations, covering the surface and near-surface atmosphere.

[0132] Forecast data can be generated by predicting atmospheric conditions, weather phenomena, and related environmental parameters using meteorological models, statistical algorithms, or target algorithms corresponding to forecast models.

[0133] Based on the target parameters, data can be downloaded through the program script in the data layer to obtain multiple downloaded data corresponding to meteorological elements in the target area within the target time period. The multiple downloaded data are then subjected to quality control and standardization processing to obtain multiple observation data and multiple forecast data corresponding to multiple forecast models for the target area within the target time period.

[0134] Quality control processes may include deleting abnormal files, checking for data anomalies and deleting files containing abnormal data, and repairing or correcting abnormal data that can be repaired.

[0135] Standardization can be used to convert downloaded data into a uniform file format, data format, and spatiotemporal resolution, and to align the data with a spatial grid structure.

[0136] It should be noted that, according to any feasible implementation method, multiple observation data and multiple forecast data corresponding to multiple forecast models for the target area within the target time period can be determined, and the embodiments of this application do not limit this.

[0137] S203. Based on meteorological elements, multiple observation data and multiple forecast data are verified and processed to obtain the verification index value corresponding to each forecast model.

[0138] Different meteorological elements require different testing methods and different testing indicators.

[0139] The test index value can be the index value corresponding to the test index.

[0140] The test index values ​​can be used to compare the quantitative results calculated from forecast data and observation data, and to objectively evaluate the accuracy, reliability and effectiveness of the forecast.

[0141] The first mapping relationship can be obtained. Based on the meteorological elements and the first mapping relationship, the target algorithm can be determined. Based on the target algorithm, multiple observation data and multiple forecast data are tested and processed to obtain the test index value corresponding to each forecast model.

[0142] The first mapping relationship may include multiple meteorological elements and the algorithm corresponding to each meteorological element.

[0143] Meteorological elements can be input into the algorithm selection model to obtain the target algorithm. Based on the target algorithm, multiple observation data and multiple forecast data are tested and processed to obtain the test index value corresponding to each forecast model.

[0144] It should be noted that, according to any feasible implementation method, multiple observation data and multiple forecast data corresponding to multiple forecast models for the target area within the target time period can be determined, and the embodiments of this application do not limit this.

[0145] S204. Based on the test index values, determine the degradation factor corresponding to each forecast model.

[0146] The degradation factor is used to represent the driving contribution of meteorological elements to the degradation of forecasting skills under forecasting models.

[0147] The test index values ​​may include threat score, latitude-weighted anomaly correlation coefficient, root mean square error, and bias, etc., and are not limited here.

[0148] The test index values ​​can include site index values ​​corresponding to multiple sites.

[0149] Optionally, for any forecast model, the degradation factor corresponding to the forecast model can be determined based on the test index value in the following way: obtain the reference value corresponding to the test index value, determine the degradation index corresponding to multiple sites based on the index value and reference value of multiple sites, and determine the degradation factor corresponding to the forecast model based on the degradation index corresponding to multiple sites and the weight value corresponding to multiple sites.

[0150] For example, for indicators such as root mean square error and deviation, the larger the degradation index, the worse the forecast effect.

[0151] The degradation index corresponding to this indicator type can be determined using the following formula:

[0152]

[0153] Among them, D s,f,t M can represent the degradation index corresponding to station s, meteorological element f, and target time period t. s,f,t This can represent the test index value corresponding to station s, meteorological element f, and target time period t, where T s,f It can represent the reference value corresponding to the test index value.

[0154] For the anomaly correlation coefficient with threat score and latitude weighting as the indicator type, the larger the degradation index, the better the forecast effect.

[0155] The degradation index corresponding to this indicator type can be determined using the following formula:

[0156]

[0157] Among them, D s,f,t M can represent the degradation index corresponding to station s, meteorological element f, and target time period t. s,f,t This can represent the test index value corresponding to station s, meteorological element f, and target time period t, where T s,f It can represent the reference value corresponding to the test index value.

[0158] The reference values ​​can be set based on historical data, sliding window quantiles, etc., and are not limited here.

[0159] Optionally, the degradation factor corresponding to the forecast model can be determined based on the degradation index corresponding to multiple stations in the target area and the weight values ​​corresponding to multiple stations in the following manner:

[0160]

[0161] Among them, w i D can represent the weight value of the i-th site. i,f,t D can represent the degradation index corresponding to the i-th station for meteorological element f and target time period t. f It can represent the degradation factor corresponding to the forecast model.

[0162] The weight value of the i-th station can be calculated using the following formula:

[0163] Since the distribution of site observations is not uniform, to avoid weight stacking in dense areas, following the principle of "giving less weight to densely populated sites and more weight to sparsely populated sites," a normalization coefficient can be introduced for regional balancing.

[0164]

[0165] ρ i Number of nearby stations (e.g., within 50km) around the station.

[0166] Optionally, for any forecast model, the degradation factor corresponding to the forecast model can be determined based on the test index value in the following ways: determine the index type corresponding to the test index value; determine the target formula and reference value based on the index type; determine the degradation index corresponding to the test index value through the target formula based on the test index value and reference value; and determine the degradation factor corresponding to the forecast model based on the degradation index and the weight values ​​of multiple stations in the target area.

[0167] In this way, based on the degradation index, the main degradation factors in different regions of the forecast model can be identified, providing direction for model improvement.

[0168] It should be noted that the degradation factor corresponding to the forecast model can be determined according to any feasible implementation method, and the embodiments of this application do not limit this.

[0169] Optionally, after determining the degradation factor corresponding to the forecast model, the method further includes: establishing a skill degradation contribution matrix and a contribution matrix between the degradation factor and the spatial region based on the degradation factor corresponding to the forecast model.

[0170] The skill degradation contribution matrix can be used to represent the degradation intensity distribution of degradation factors across multiple sites in the target region.

[0171] Each element in the contribution matrix of degradation factor and spatial region can represent the normalized intensity of skill degradation caused by the degradation factor in the corresponding region of the site.

[0172] The contribution matrix of degradation factors and spatial regions can be used to identify "which degradation factors lead to skill degradation in which regions".

[0173] S205. Based on multiple test index values ​​and multiple degradation factors, generate visual test results corresponding to multiple forecast models.

[0174] Visualized verification results can be used for comparative verification of spatial, temporal, and multiple forecast models.

[0175] Based on multiple test index values ​​and multiple degradation factors, multiple preset scripts can be used to generate visualized test results corresponding to multiple forecast modes.

[0176] The visual test results can include spatial skill maps, jigsaw puzzle skill radar charts, time-series skill trend charts, and uncertainty display charts.

[0177] Spatial skill maps can automatically match grading standards and generate color gradient maps based on the spatial distribution characteristics of test index values ​​fused with map background, achieving a "one-map-to-the-world" assessment pattern perception.

[0178] Spatial skill maps can also be used to integrate the main degradation factors of each site / region (different degradation factors can be represented by different colors), which can intuitively determine where and why degradation occurs, and assist in regional correction and pattern optimization.

[0179] Radar charts, through multi-mode, multi-element, and multi-level dimension comprehensive comparisons, can showcase the advantages and disadvantages of different forecast models in key meteorological indicators in the form of radar chart splicing matrices, thereby enhancing the interpretability analysis between models.

[0180] Time series technique trend charts can precisely show the changing trends of test index values ​​corresponding to different forecast models over time, helping to identify stability and abrupt change points.

[0181] Uncertainty visualization can innovatively output skill indicator distribution histograms and confidence interval box plots side by side, improving the robustness and reliability of decision-making references.

[0182] The forecast model verification method provided in this embodiment obtains user-input target parameters, including target area, meteorological elements, and target time period. It then determines multiple observational data points for the target area within the target time period and multiple forecast data points corresponding to each forecast model. Based on the meteorological elements, it performs verification processing on the multiple observational and forecast data points to obtain verification index values ​​for each forecast model. Based on the verification index values, it determines the degradation factor corresponding to each forecast model. The degradation factor represents the driving contribution of meteorological elements to the degradation of forecast skill under the forecast model. Based on the multiple verification index values ​​and multiple degradation factors, it generates visualized verification results corresponding to each forecast model. In this way, different verification methods are constructed based on different meteorological elements, improving the adaptability of the verification process. Furthermore, by determining the degradation factor corresponding to each forecast model, it enables the monitoring and evaluation of the degradation process, thereby improving the accuracy of forecast model verification.

[0183] Below, in conjunction with Figure 3 The process (S203) of verifying and processing multiple observation data and multiple forecast data based on meteorological elements to obtain the verification index value corresponding to the forecast model for any forecast model is explained.

[0184] Figure 3 A flowchart illustrating another method for verifying a forecast pattern provided in this application embodiment. Based on the above embodiments, see [link to relevant documentation]. Figure 3 The method includes:

[0185] S301. Based on meteorological elements, determine the target algorithm from multiple verification algorithms.

[0186] Meteorological elements include visibility, precipitation, temperature, zonal wind, meridional wind, and humidity.

[0187] Zonal winds are winds that travel along the latitude (east-west direction).

[0188] Zonal winds can include 10-meter winds and 100-meter winds, depending on the altitude at which they are measured.

[0189] Meridional winds refer to winds that travel along the direction of meridians (north-south direction).

[0190] Meridional wind can also include 10-meter wind and 100-meter wind, depending on the height being measured.

[0191] Several testing algorithms are available, including dynamic threshold testing algorithms and continuous variable testing algorithms.

[0192] The first mapping relationship can be obtained, and the target algorithm can be determined from multiple verification algorithms based on meteorological elements.

[0193] Optionally, the target algorithm can be determined from multiple testing algorithms based on meteorological elements in the following way: if the meteorological element is visibility or precipitation, the target algorithm is determined to be a dynamic threshold testing algorithm; if the meteorological element is temperature, zonal wind, meridional wind and humidity, the target algorithm is determined to be a continuous variable testing algorithm.

[0194] S302. Based on the target algorithm, multiple observation data and multiple forecast data are verified and processed to obtain the first index value corresponding to the forecast model.

[0195] The first indicator value can be the indicator value calculated according to the target algorithm.

[0196] Optionally, the target algorithm is a dynamic threshold verification algorithm, which can be used to verify multiple observation data and multiple forecast data to obtain the first index value corresponding to the forecast model in the following way: determine multiple time points corresponding to the target time period; obtain a preset sliding window half-width; determine the first time period based on the target time period and the preset sliding window half-width; determine the dynamic threshold corresponding to the time point based on the preset quantile level and the multiple observation data corresponding to the first time period; classify the multiple observation data and multiple forecast data corresponding to the first time period based on the dynamic threshold to obtain the observation level and forecast level of the multiple observation data; determine the verification data corresponding to the time point based on the observation level and forecast level of the multiple observation data; and determine the first index value corresponding to the forecast model based on the verification data corresponding to the multiple time points.

[0197] The inspection data includes multiple inspection levels, as well as the number of hits, missed reports, false alarms, and correct rejections for each inspection level.

[0198] The dynamic threshold can be determined using the following formula:

[0199] θ t =Quantile q (X t-N:t+N )

[0200] t can be a point in time, N can be a preset half-width of the sliding window, and X can be a time point. t-N:t+N For the first time period, θt is the dynamic threshold at time point t, and q can be the quantile level. q It can represent a quantile calculation function.

[0201] Optionally, the first indicator value corresponding to the forecast model can be determined based on the test data corresponding to multiple time points in the following manner: In the test data corresponding to multiple test levels, determine the level test data corresponding to each of the multiple test levels in ascending order; for any test level, determine the test score of the test level based on the level test data corresponding to the test level; and determine the first indicator value corresponding to the forecast model based on the test scores corresponding to the multiple test levels and the weight values ​​corresponding to the multiple test levels.

[0202] The graded inspection data includes the number of hits, missed reports, false alarms, and correct rejections for the same inspection grade at multiple time points.

[0203] Optionally, the inspection score for the inspection level can be determined based on the inspection data corresponding to the inspection level, and the inspection score may include a threat score (TS) and a quitable threat score (ETS):

[0204]

[0205] in,

[0206] Where k can represent the inspection level, TS k This can represent the threat score corresponding to the inspection level k, ETS k H can represent the threat score corresponding to the inspection level k. k It can represent the number of hits, M k It can represent the number of missed reports, F k It can represent the number of false alarms, C k It can indicate the number of correctly rejected reports.

[0207] Optionally, the first indicator value corresponding to the forecast model can be determined by the following method based on the test scores corresponding to multiple test levels and the weight values ​​corresponding to each of the multiple test levels: determine the product factor of the test score and the corresponding weight value for each test level, add the product factors corresponding to multiple test levels, and determine the first indicator value corresponding to the forecast model.

[0208] Optionally, the target algorithm is a continuous variable verification algorithm, which can be used to verify multiple observation data and multiple forecast data to obtain the first index value corresponding to the forecast model in the following way: obtain the latitude weighted value corresponding to the target area; and determine multiple first index values ​​corresponding to the forecast model through the target algorithm based on multiple observation data, multiple forecast data, and latitude weighted value.

[0209] The first indicator value includes the anomaly correlation coefficient, the deviation value, and the root mean square error value.

[0210] The latitude-weighted values ​​corresponding to the target area are preset.

[0211] The target algorithm may include correlation algorithms for calculating the anomaly correlation coefficient, correlation algorithms for calculating the deviation, and correlation algorithms for calculating the root mean square error.

[0212] It should be noted that the first index value corresponding to the forecast model can be determined according to any feasible implementation method, and the embodiments of this application do not limit this.

[0213] S303. Resample the first index value to obtain the test index value corresponding to the forecast model.

[0214] The quantity corresponding to the first indicator value can be determined, and N repeated resampling operations are performed on the first indicator value to obtain N sample sequences. The test indicator value is N sample sequences.

[0215] The resampling operation involves selecting one index value from the first index values.

[0216] The implementation details of each step in this application embodiment can be found in the description of the corresponding steps or operations in the above method embodiments; repeated content will not be repeated.

[0217] The forecast model verification method provided in this embodiment determines a target algorithm from multiple verification algorithms based on meteorological elements; performs verification processing on multiple observation data and multiple forecast data according to the target algorithm to obtain a first index value corresponding to the forecast model; and resampling the first index value to obtain the verification index value corresponding to the forecast model. In this way, different verification methods are constructed based on different meteorological elements, improving the adaptability of the verification and the accuracy of the verified forecast model.

[0218] Below, in conjunction with Figure 4 The technical solutions shown in this application will be illustrated through specific examples.

[0219] Figure 4 A flowchart illustrating another method for verifying a forecast pattern provided in this application embodiment. See also... Figure 4 The method includes:

[0220] S401. Obtain the target parameters input by the user.

[0221] S402. Determine multiple observation data and multiple forecast data corresponding to multiple forecast models for the target area within the target time period.

[0222] S403. Based on meteorological factors, determine the target algorithm from multiple verification algorithms.

[0223] S404. Based on the target algorithm, multiple observation data and multiple forecast data are tested and processed to obtain the first index value corresponding to each forecast model.

[0224] S405. Resample multiple first index values ​​to obtain the test index value corresponding to each forecast model.

[0225] S406. For any forecast model, determine the index type corresponding to the test index value.

[0226] S407. Determine the target formula and reference value according to the indicator type.

[0227] S408. Based on the test index value and the reference value, determine the degradation index corresponding to the test index value through the target formula.

[0228] S409. Determine the degradation factor corresponding to the forecast model based on the degradation index.

[0229] S410. Based on multiple test index values ​​and multiple degradation factors, generate visual test results corresponding to multiple forecast models.

[0230] The implementation details of each step in this application embodiment can be found in the description of the corresponding steps or operations in the above method embodiments; repeated content will not be repeated.

[0231] In this way, different verification methods are constructed based on different meteorological elements, which improves the adaptability of verification and determines the degradation factor corresponding to each forecast model, enabling the degradation process to be monitored and evaluated, thereby improving the accuracy of the verified forecast models.

[0232] Figure 5 This is a schematic diagram of a forecast pattern verification device provided in an embodiment of this application. Please refer to... Figure 5 The forecast pattern verification device 500 includes an acquisition module 501, a determination module 502, a first verification module 503, a second verification module 504, and a display module 505, wherein...

[0233] The acquisition module 501 is used to acquire target parameters input by the user, the target parameters including target area, meteorological elements and target time period;

[0234] The determining module 502 is used to determine multiple observation data and multiple forecast data corresponding to multiple forecast models for the target area within the target time period;

[0235] The first verification module 503 is used to verify the multiple observation data and the multiple forecast data according to the meteorological elements, and obtain the verification index value corresponding to each forecast model.

[0236] The second verification module 504 is used to determine the degradation factor corresponding to each forecast model based on the verification index value. The degradation factor is used to represent the driving contribution of the meteorological elements to the degradation of forecast skills under the forecast model.

[0237] Display module 505 is used to generate visualized test results corresponding to multiple forecast modes based on multiple test index values ​​and multiple degradation factors.

[0238] Optionally, in the apparatus described above, the first inspection module 503 is specifically used for:

[0239] Based on the meteorological elements, the target algorithm is determined from multiple verification algorithms;

[0240] According to the target algorithm, the multiple observation data and the multiple forecast data are verified to obtain the first index value corresponding to the forecast model;

[0241] The first index value is resampled to obtain the test index value corresponding to the forecast model.

[0242] Optionally, in the apparatus described above, the meteorological elements include visibility, precipitation, temperature, zonal wind, meridional wind, and humidity; the multiple testing algorithms include a dynamic threshold testing algorithm and a continuous variable testing algorithm; and the first testing module 503 is specifically used for:

[0243] If the meteorological element is visibility or precipitation, then the target algorithm is determined to be a dynamic threshold verification algorithm;

[0244] If the meteorological elements are temperature, zonal wind, meridional wind, and humidity, then the target algorithm is determined to be a continuous variable testing algorithm.

[0245] Optionally, in the apparatus described above, the target algorithm is a dynamic threshold testing algorithm, and the first testing module 503 is specifically used for:

[0246] Determine multiple time points corresponding to the target time period;

[0247] Get the preset half width of the sliding window;

[0248] Based on the target time period and the preset sliding window half width, a first time period is determined; based on the preset quantile level and multiple observation data corresponding to the first time period, a dynamic threshold corresponding to the time point is determined;

[0249] Based on the dynamic threshold, multiple observation data and multiple forecast data corresponding to the first time period are classified and processed to obtain the observation level of multiple observation data and the forecast level of multiple forecast data.

[0250] Based on the observation levels of multiple observation data and the forecast levels of multiple forecast data, the verification data corresponding to the time point is determined. The verification data includes multiple verification levels, as well as the number of hits, missed reports, false alarms, and correct rejections corresponding to each verification level.

[0251] Based on the test data corresponding to multiple time points, the first index value corresponding to the forecast model is determined.

[0252] Optionally, in the apparatus described above, the first inspection module 503 is specifically used for:

[0253] According to the order of multiple inspection levels from smallest to largest, the level inspection data corresponding to multiple inspection levels are determined sequentially from the inspection data corresponding to multiple time points. The level inspection data includes the number of hits, the number of missed reports, the number of false alarms, and the number of correct rejections corresponding to the same inspection level at multiple time points.

[0254] For any given inspection level, the inspection score for that inspection level is determined based on the inspection data corresponding to that inspection level.

[0255] The first indicator value corresponding to the forecasting model is determined based on the inspection scores corresponding to multiple inspection levels and the weight values ​​corresponding to each of the multiple inspection levels.

[0256] Optionally, in the apparatus described above, the target algorithm is a continuous variable testing algorithm, and the first testing module 503 is specifically used for:

[0257] Obtain the latitude-weighted value corresponding to the target region;

[0258] Based on the multiple observation data, the multiple forecast data, and the latitude weighted value, the target algorithm determines multiple first index values ​​corresponding to the forecast model. The first index values ​​include anomaly correlation coefficient, deviation, and root mean square error.

[0259] Optionally, in the apparatus described above, the second inspection module 504 is specifically used for:

[0260] Determine the type of indicator corresponding to the test indicator value;

[0261] Determine the target formula and reference value based on the indicator type;

[0262] Based on the test index value and the reference value, the degradation index corresponding to the test index value is determined by the target formula.

[0263] Based on the degradation index, the degradation factor corresponding to the forecast model is determined.

[0264] Figure 6This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Please refer to... Figure 6 The electronic device 600 may include: a memory 601, a processor 602, and a transceiver 603.

[0265] Memory 601 is used to store program instructions;

[0266] The processor 602 is used to execute the program instructions stored in the memory so that the electronic device 600 performs the above-described method.

[0267] Transceiver 603 may include a transmitter and / or a receiver. The transmitter may also be referred to as a transmitter, transmitter port, or transmitter interface, and the receiver may also be referred to as a receiver port, receiver interface, or similar descriptions. Exemplarily, memory 601, processor 602, and transceiver 603 are interconnected via bus 604.

[0268] This application also provides a computer program product that can be executed by a processor, and when the computer program product is executed, the above-described method can be implemented.

[0269] The forecast pattern verification device, electronic device, computer-readable storage medium, and computer program product of this application embodiment can execute the technical solution shown in the above forecast pattern verification method embodiment. Their implementation principle and beneficial effects are similar, and will not be repeated here.

[0270] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0271] It should be further noted that although the steps in the flowchart are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0272] It should be understood that the above-described device embodiments are merely illustrative, and the device of this application can also be implemented in other ways. For example, the division of units / modules in the above embodiments is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units, modules, or components may be combined, or integrated into another system, or some features may be ignored or not executed.

[0273] Furthermore, unless otherwise specified, the functional units / modules in the various embodiments of this application can be integrated into one unit / module, or each unit / module can exist physically separately, or two or more units / modules can be integrated together. The integrated units / modules described above can be implemented in hardware or as software program modules.

[0274] When an integrated unit / module is implemented in hardware, the hardware can be digital circuits, analog circuits, etc. The physical implementation of the hardware structure includes, but is not limited to, transistors, memristors, etc. Unless otherwise specified, the processor can be any suitable hardware processor, such as a CPU, GPU, FPGA, DSP, and ASIC, etc. Unless otherwise specified, the storage unit can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc.

[0275] If the integrated unit / module is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0276] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as these combinations of technical features do not contradict each other, they should be considered within the scope of this specification.

[0277] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0278] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for verifying a forecast model, characterized in that, The method includes: Obtain the target parameters input by the user, including the target area, meteorological elements, and target time period; Determine multiple forecast data corresponding to multiple observation data and multiple forecast models for the target area within the target time period; Based on the meteorological elements, the multiple observation data and the multiple forecast data are verified and processed to obtain the verification index value corresponding to each forecast model; Based on the test index value, the degradation factor corresponding to each forecast model is determined. The degradation factor is used to represent the driving contribution of the meteorological elements to the degradation of forecast skills under the forecast model. Based on multiple test index values ​​and multiple degradation factors, visualized test results corresponding to multiple forecast models are generated. For any forecast model, based on the meteorological elements, the multiple observation data and the multiple forecast data are tested to obtain the test index value corresponding to the forecast model, including: Based on the meteorological elements, the target algorithm is determined from multiple verification algorithms; If the target algorithm is a dynamic threshold test algorithm, the dynamic threshold corresponding to the time point is determined based on the preset quantile level and multiple observation data corresponding to the first time period. The first time period is determined based on the target time period and the preset sliding window half width. Based on the dynamic threshold, multiple observation data and multiple forecast data corresponding to the first time period are classified and processed to obtain the observation level of multiple observation data and the forecast level of multiple forecast data. Based on the observation level of multiple observation data and the forecast level of multiple forecast data, the test data corresponding to the time point is determined. Based on the test data corresponding to multiple time points, determine the first index value corresponding to the forecast model; The first index value is resampled to obtain the test index value corresponding to the forecast model.

2. The method according to claim 1, characterized in that, The meteorological elements include visibility, precipitation, temperature, zonal wind, meridional wind, and humidity. The multiple testing algorithms include a dynamic threshold testing algorithm and a continuous variable testing algorithm. Based on the meteorological elements, a target algorithm is determined from among the multiple testing algorithms, including: If the meteorological element is visibility or precipitation, then the target algorithm is determined to be a dynamic threshold verification algorithm; If the meteorological elements are temperature, zonal wind, meridional wind, and humidity, then the target algorithm is determined to be a continuous variable testing algorithm.

3. The method according to claim 1, characterized in that, Based on the test data corresponding to multiple time points, the first index value corresponding to the forecast model is determined, including: According to the order of multiple inspection levels from smallest to largest, the level inspection data corresponding to multiple inspection levels are determined sequentially from the inspection data corresponding to multiple time points. The level inspection data includes the number of hits, the number of missed reports, the number of false alarms, and the number of correct rejections corresponding to the same inspection level at multiple time points. For any given inspection level, the inspection score for that inspection level is determined based on the inspection data corresponding to that inspection level. The first indicator value corresponding to the forecasting model is determined based on the inspection scores corresponding to multiple inspection levels and the weight values ​​corresponding to each of the multiple inspection levels.

4. The method according to claim 2, characterized in that, The target algorithm is a continuous variable testing algorithm. Based on the target algorithm, the multiple observation data and the multiple forecast data are tested to obtain the first index value corresponding to the forecast model, including: Obtain the latitude-weighted value corresponding to the target region; Based on the multiple observation data, the multiple forecast data, and the latitude weighted value, the target algorithm determines multiple first index values ​​corresponding to the forecast model. The first index values ​​include anomaly correlation coefficient, deviation, and root mean square error.

5. The method according to any one of claims 1-4, characterized in that, For any forecast model, based on the test index value, determine the degradation factor corresponding to the forecast model, including: Determine the type of indicator corresponding to the test indicator value; Determine the target formula and reference value based on the indicator type; Based on the test index value and the reference value, the degradation index corresponding to the test index value is determined by the target formula. The degradation factor corresponding to the forecast model is determined based on the degradation index and the weight values ​​of multiple stations in the target area.

6. A device for verifying forecasts from a forecasting model, characterized in that, The device includes: The acquisition module is used to acquire target parameters input by the user, including target area, meteorological elements, and target time period; The determination module is used to determine multiple observation data and multiple forecast data corresponding to multiple forecast models for the target area within the target time period; The first verification module is used to verify the multiple observation data and the multiple forecast data according to the meteorological elements, and obtain the verification index value corresponding to each forecast model. The second verification module is used to determine the degradation factor corresponding to each forecast model based on the verification index value. The degradation factor is used to represent the driving contribution of the meteorological elements to the degradation of forecast skills under the forecast model. The display module is used to generate visual test results corresponding to multiple forecast models based on multiple test index values ​​and multiple degradation factors. For any given forecast model, the first verification module is specifically used for: Based on the meteorological elements, the target algorithm is determined from multiple verification algorithms; If the target algorithm is a dynamic threshold test algorithm, the dynamic threshold corresponding to the time point is determined based on the preset quantile level and multiple observation data corresponding to the first time period. The first time period is determined based on the target time period and the preset sliding window half width. Based on the dynamic threshold, multiple observation data and multiple forecast data corresponding to the first time period are classified and processed to obtain the observation level of multiple observation data and the forecast level of multiple forecast data. Based on the observation level of multiple observation data and the forecast level of multiple forecast data, the test data corresponding to the time point is determined. Based on the test data corresponding to multiple time points, determine the first index value corresponding to the forecast model; The first index value is resampled to obtain the test index value corresponding to the forecast model.

7. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 5.

8. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method described in any one of claims 1 to 5.

Citation Information

Patent Citations

  • Multi-mode rainfall forecasting and real-time dynamic inspection and evaluation system

    CN113033957A