Correction Methods for Typhoon Precipitation Forecasts and Analysis of Their Error Sources

Abstract

The present invention relates to a correction method for typhoon precipitation forecast and a method for analyzing the source of its error. The correction method comprises: obtaining a forecast data set including original forecast data of a target typhoon; wherein each forecast data in the forecast data set includes a typhoon forecast path, and the original forecast data also includes original typhoon precipitation forecast data; according to the real-time position of the target typhoon, selecting N typhoon forecast paths with the smallest path error from multiple typhoon forecast paths in the forecast data set and performing arithmetic averaging to obtain an optimal forecast path; wherein N is a preset positive integer; determining a position offset between the typhoon forecast path of the original forecast data and the optimal forecast path; and correcting the original typhoon precipitation forecast data based on the position offset to obtain corrected typhoon precipitation forecast data for the optimal forecast path. With the help of the above technical solution, the embodiment of the present application improves the accuracy of typhoon precipitation forecast data.

Description

Technical Field

[0001] The present invention relates to the technical field of typhoon forecasting, and in particular to a correction method for typhoon precipitation forecasting and a method for analyzing error sources thereof. Background Art

[0002] Typhoons (TCs) are one of the major tropical weather systems affecting my country. Each year, the heavy rainfall and subsequent disasters they trigger cause significant economic losses and casualties, particularly in coastal areas. Despite significant progress in typhoon track forecasting over the past few decades, typhoon precipitation forecasts lag significantly behind track predictions.

[0003] Currently, typhoon precipitation forecasts rely primarily on numerical weather models, but these models still present significant uncertainty. In reality, typhoon precipitation forecasts are not only closely related to the typhoon's path but are also constrained by factors such as the typhoon's structure, underlying surface, and large-scale environmental fields. These factors exhibit complex magnitude variations and distribution patterns, significantly increasing the complexity of forecasts. Given the advancements in global numerical models and ensemble forecasts, typhoon path forecasts are continuously improving. Therefore, a new method to effectively enhance typhoon precipitation forecasts and techniques to investigate the sources of forecast errors are needed. Summary of the Invention

[0004] (1) Technical issues to be resolved

[0005] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a correction method for typhoon precipitation forecast and a method for analyzing the source of its error, which solves the technical problem that the existing typhoon precipitation forecast has low accuracy.

[0006] (2) Technical solution

[0007] In order to achieve the above objectives, the main technical solutions adopted by the present invention include:

[0008] In a first aspect, an embodiment of the present invention provides a method for correcting a typhoon precipitation forecast, the correction method comprising: obtaining a forecast data set including original forecast data of a target typhoon; wherein each forecast data in the forecast data set includes a typhoon forecast path, and the original forecast data also includes original typhoon precipitation forecast data; according to the real-time position of the target typhoon, selecting N typhoon forecast paths with the smallest path error from multiple typhoon forecast paths in the forecast data set and performing arithmetic averaging to obtain an optimal forecast path; wherein N is a preset positive integer; determining a position offset between the typhoon forecast path of the original forecast data and the optimal forecast path; and correcting the original typhoon precipitation forecast data based on the position offset to obtain corrected typhoon precipitation forecast data for the optimal forecast path.

[0009] Therefore, the embodiment of the present invention obtains a forecast data set including original forecast data of the target typhoon, and selects N typhoon forecast paths with the smallest path errors from multiple typhoon forecast paths in the forecast data set according to the real-time position of the target typhoon, and performs arithmetic averaging to obtain an optimal forecast path, and determines the position offset between the typhoon forecast path of the original forecast data and the optimal forecast path, and corrects the original typhoon precipitation forecast data based on the position offset to obtain corrected typhoon precipitation forecast data for the optimal forecast path. In this way, the typhoon forecast path can be corrected first, and considering that the typhoon forecast path has the greatest impact on precipitation, the precipitation forecast data can be further corrected based on the optimal typhoon forecast path, thereby obtaining accurate precipitation forecast data (for example, precipitation area and precipitation amount, etc.), thereby improving the accuracy of the typhoon precipitation forecast data.

[0010] Optionally, determining the position offset between the typhoon forecast path of the original forecast data and the optimal forecast path includes: dividing the typhoon forecast path of the original forecast data and the optimal forecast path into m sub-paths based on the precipitation interval output by the model; where m is a positive integer; and calculating the average position offset between each sub-path of the m sub-paths of the typhoon forecast path of the original forecast data and a sub-path of its corresponding optimal forecast path.

[0011] Therefore, the embodiment of the present application divides the typhoon forecast path and the optimal forecast path of the original forecast data, and separately calculates the average position offset between each sub-path in the m sub-paths of the typhoon forecast path of the original forecast data and its corresponding sub-path of the optimal forecast path, so as to accurately determine the position offset.

[0012] Optionally, the original typhoon precipitation forecast data is corrected based on the position offset to obtain corrected typhoon precipitation forecast data for the optimal forecast path, including: determining precipitation grid data within a preset range around the typhoon forecast path of the original forecast data; moving the position of the corresponding precipitation grid data based on each average position offset and keeping the precipitation value unchanged, and accumulating m precipitation correction areas to obtain the corrected precipitation area of ​​the optimal forecast path and the precipitation within the corrected precipitation area.

[0013] Therefore, the embodiment of the present application moves the position of the corresponding precipitation grid data based on each average position offset and keeps the precipitation value unchanged, and accumulates m precipitation correction areas to obtain the corrected precipitation area of ​​the optimal forecast path and the precipitation in the corrected precipitation area, thereby obtaining accurate precipitation forecast data.

[0014] In a second aspect, an embodiment of the present invention provides a method for analyzing sources of typhoon precipitation forecast errors, characterized in that the analysis method is used to analyze sources of error in original forecast data and revised typhoon precipitation forecast data obtained by correcting the original forecast data, and the revised typhoon precipitation forecast data is obtained by any typhoon precipitation forecast correction method of the first aspect, and the analysis method includes: based on a preset precipitation threshold, determining a forecast precipitation object, an actual precipitation object, and an initial continuous rain area CRA precipitation verification area within a specified area; wherein the forecast precipitation object includes an original forecast precipitation object corresponding to the original forecast data and a revised forecast precipitation object corresponding to the revised typhoon precipitation forecast data, and the initial continuous rain area CRA precipitation verification area is a union of the forecast precipitation object and the actual precipitation object; determining the rainfall center position of the actual precipitation object, and searching for the optimal horizontal displacement distance of the forecast precipitation object within a preset range around the rainfall center position of the actual precipitation object to determine an intermediate matching area of ​​the forecast precipitation object, and comparing the intermediate matching area with the initial CRA precipitation. The union of the verification areas is used as the first CRA precipitation verification area; wherein the optimal horizontal displacement distance is used to maximize the first correlation coefficient between the forecast precipitation field and the actual precipitation field in the first CRA precipitation verification area, and when the first correlation coefficient is maximized, the mean square error between the forecast precipitation field and the actual precipitation field in the first CRA precipitation verification area is minimized; an optimal rotation angle of the intermediate matching area is determined, and the intermediate matching area is rotated based on the optimal rotation angle to obtain a final matching area of ​​the forecast precipitation object, and the union of the final matching area and the initial CRA precipitation verification area is used as the second CRA precipitation verification area; wherein the optimal rotation angle is used to maximize the second correlation coefficient between the forecast precipitation field and the actual precipitation field in the second CRA precipitation verification area, and when the second correlation coefficient is maximized, the mean square error between the forecast precipitation field and the actual precipitation field in the second CRA precipitation verification area is minimized; the error of the typhoon precipitation forecast is analyzed based on the initial CRA precipitation verification area, the first CRA precipitation verification area, and the second CRA precipitation verification area to obtain an error analysis result.

[0015] Therefore, the embodiment of the present application proposes a method for analyzing the sources of typhoon precipitation forecast errors, which can accurately analyze the sources of typhoon precipitation misalarms, thereby providing a basis for improving typhoon precipitation forecasts.

[0016] Optionally, the error analysis result includes a center offset error; the center offset error is obtained by the following formula:

[0017] ;

[0018] in, Used to indicate center offset error; S FIt is used to represent the standard deviation of the forecast precipitation in the initial CRA precipitation verification area; Used to represent the standard deviation of actual precipitation in the initial CRA precipitation test area; It is used to represent the spatial correlation between the forecast precipitation and the actual precipitation in the first CRA precipitation verification area; r is used to represent the spatial correlation between the forecast precipitation and the actual precipitation in the initial CRA precipitation verification area.

[0019] Optionally, the error analysis result also includes a rotation error; the rotation error is obtained by the following formula:

[0020] ;

[0021] in, Used to represent rotation error; Used to represent the spatial correlation between forecast precipitation and actual precipitation in the second CRA precipitation verification area.

[0022] Optionally, the error analysis result also includes a total error; the total error is obtained by the following formula:

[0023] ;

[0024] in, Used to indicate total error; It is used to represent the average value of the forecast precipitation in the second CRA precipitation test area; Used to represent the average value of actual precipitation in the second CRA precipitation test area.

[0025] Optionally, the error analysis result also includes morphological error; the morphological error is obtained by the following formula:

[0026] ;

[0027] in, Used to represent morphological errors.

[0028] Optionally, the error analysis result further includes a movement error, where the movement error is the sum of a center offset error and a rotation error.

[0029] In a third aspect, an embodiment of the present application provides an electronic device having a computer program stored thereon, wherein the computer program is executed by a processor to execute the method described in the first aspect or any optional implementation of the first aspect.

[0030] In a fourth aspect, an embodiment of the present application provides an electronic device having a computer program stored thereon, wherein the computer program is executed by a processor to execute the method described in the second aspect or any optional implementation of the second aspect.

[0031] In order to make the above-mentioned objectives, features and advantages to be achieved by the embodiments of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0033] Figure 1 A flowchart illustrating a method for correcting a typhoon precipitation forecast provided by an embodiment of the present application is shown;

[0034] Figure 2 A flow chart of a method for analyzing sources of typhoon precipitation forecast errors provided by an embodiment of the present application is shown;

[0035] Figure 3a and Figure 3b A schematic diagram showing an improvement in correlation by path correction and CRA shifting provided by an embodiment of the present application is shown;

[0036] Figure 3c and Figure 3d A schematic diagram showing an improvement in root mean square error by path correction and CRA shifting provided by an embodiment of the present application is shown;

[0037] Figure 3e and Figure 3f A schematic diagram showing the improvement of ETS scoring by path correction and CRA shifting provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0038] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation methods in conjunction with the accompanying drawings.

[0039] A recent study of 24-hour precipitation forecasts for Typhoon Lekima showed that while ETS scores for 0-24-hour precipitation forecasts within 72 hours were relatively high, the ETS score dropped rapidly when 24-hour cumulative precipitation reached 100 mm or more. By the time precipitation reached 250 mm or more, the ETS score for 0-24-hour precipitation forecasts fell below 0.1 (or even approached 0). This indicates that typhoon precipitation forecasts are significantly underperforming.

[0040] Previous research has focused primarily on the verification and analysis of typhoon precipitation forecasts, but has not proposed effective and feasible methods for improving them. In fact, it is a common understanding within the research and operational fields that typhoon tracks, as a key factor influencing typhoon precipitation distribution, can significantly impact typhoon precipitation forecast performance, depending on the level of their forecast.

[0041] Based on this, this application proposes a new precipitation forecasting technology based on typhoon path correction based on the diagnostic analysis results of typhoon precipitation forecast errors, thereby making up for the shortcomings of existing typhoon precipitation forecasting technology and accuracy, and at the same time meeting the actual business needs of typhoon precipitation forecasting.

[0042] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. Instead, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0043] See Figure 1 , Figure 1 A flowchart of a method for correcting a typhoon precipitation forecast provided by an embodiment of the present application is shown. It should be understood that the method for correcting a typhoon precipitation forecast can be performed by an electronic device, and the specific device of the electronic device can be configured according to actual needs, and the embodiments of the present application are not limited thereto. For example, the electronic device can be a computer or a server. Specifically, the correction method includes:

[0044] Step S110: Obtain a forecast data set including original forecast data of the target typhoon, wherein each forecast data in the forecast data set includes a typhoon forecast path, and the original forecast data also includes original typhoon precipitation forecast data.

[0045] It should be understood that the specific typhoon of the target typhoon can be set according to actual needs, and the embodiments of the present application are not limited to this.

[0046] It should also be understood that the method for obtaining the forecast data set of the target typhoon including the original forecast data can be set according to actual needs, and the embodiments of the present application are not limited thereto.

[0047] Optionally, raw model data is obtained, and multiple forecast data are extracted from the raw model data. The multiple forecast data may be forecast data for a target typhoon in different regions, and each of the multiple forecast data may include the forecasted typhoon center location (e.g., longitude and latitude) of the target typhoon and its time information (e.g., the forecast location of the typhoon center at a certain moment), as well as a time series set of precipitation data (e.g., the forecasted precipitation within a preset range around the target typhoon at a certain moment). Furthermore, the multiple forecast data may also include raw forecast data, and the raw forecast data may include forecast data to be revised.

[0048] It should be noted here that the time resolution of the multiple forecast data may be determined by the existing forecast business requirements and the time resolution of the original model data.

[0049] In step S120, based on the real-time position of the target typhoon, N typhoon forecast paths with the smallest path errors are selected from the multiple typhoon forecast paths in the forecast dataset, and the arithmetic average is performed to obtain the optimal forecast path. Wherein, N is a preset positive integer, i.e., the specific value of N can be set according to actual needs, and the embodiments of the present application are not limited thereto.

[0050] Specifically, the real-time position of the target typhoon can be obtained from the designated platform, and the N typhoon forecast paths with the smallest path error (or in other words) are selected from the multiple typhoon forecast paths in the forecast data set for arithmetic average (AVE) to obtain the optimal forecast path. The specific formula is as follows:

[0051] ;

[0052] ;

[0053] in, It can be used to represent the distance between the typhoon forecast position PE at the current moment in the typhoon forecast path and the typhoon's current real-time position P0; function It can be used to indicate taking the N typhoon forecast paths with the smallest error distance among multiple typhoon forecast paths; the function AVE can be used to indicate taking the average of the N typhoon forecast paths; It can be used to represent the time difference between the latest forecast time based on the current data set and the start time of the data set; function Can be used to represent the average path of N selected members; function It can represent the forecast path after taking data delay into account.

[0054] Finally, you can Extract the corrected typhoon center position and time information.

[0055] Step S130: determining the position offset between the typhoon forecast path of the original forecast data and the optimal forecast path.

[0056] Specifically, the precipitation interval based on the model output (For example, 6h, etc., which can be set according to actual needs) The typhoon forecast path and the optimal forecast path of the original forecast data are divided into m sub-paths, where m is a positive integer, and the average position offset between each sub-path of the m sub-paths of the typhoon forecast path of the original forecast data and its corresponding sub-path of the optimal forecast path is calculated. .

[0057] For example, the precipitation intervals can be based on the model output The typhoon forecast path and the optimal forecast path of the original forecast data are divided into 20 sub-paths (i.e. each sub-path corresponds to a precipitation interval , that is, a sub-path is a typhoon in a precipitation interval The average position offset between the first sub-path of the typhoon forecast path of the original forecast data and the first sub-path of the optimal forecast path can be calculated. Based on the above steps, the average position offset between the remaining 19 sub-paths can be calculated respectively.

[0058] Step S140 , correcting the original typhoon precipitation forecast data based on the position offset to obtain corrected typhoon precipitation forecast data of the optimal forecast path.

[0059] Specifically, the precipitation grid data within the preset range around the typhoon forecast path of the original forecast data is determined, and the average position offset is used to calculate the precipitation grid data. Move the position of each precipitation grid point while keeping the precipitation value unchanged, and finally accumulate the m precipitation correction areas to obtain the final total precipitation area.

[0060] ;

[0061] ;

[0062] ;

[0063] in, is the surrounding precipitation area corresponding to the original forecast path i of the original forecast data; is the surrounding precipitation area corresponding to the optimal forecast path i; R represents the precipitation amount.

[0064] Therefore, with the help of the above technical solution, the embodiment of the present application can make up for the shortcomings of existing typhoon precipitation forecasting technology and accuracy, while meeting the actual business needs of typhoon precipitation forecasting.

[0065] It should be understood that the above-mentioned correction method for typhoon precipitation forecast is merely exemplary, and those skilled in the art may make various modifications based on the above-mentioned method, and the modified schemes also fall within the scope of protection of this application.

[0066] See Figure 2 , Figure 2 A flowchart of a method for analyzing the sources of typhoon precipitation forecast errors provided by an embodiment of the present application is shown. It should be understood that the analysis method can be performed by an electronic device, and the specific device of the electronic device can be set according to actual needs, and the embodiments of the present application are not limited thereto. For example, the electronic device can be a computer or a server. Specifically, the analysis method includes:

[0067] Step S210 determines, based on a preset precipitation threshold, the forecast precipitation objects, current precipitation objects, and the initial continuous rain area (CRA) precipitation verification area within the designated area. The forecast precipitation objects include original forecast precipitation objects corresponding to the original forecast data (or, original forecast precipitation objects are model output precipitation without any path correction) and revised forecast precipitation objects corresponding to the revised typhoon precipitation forecast data (or, revised forecast precipitation objects are forecast precipitation objects corrected based on the optimal forecast path). Current precipitation objects are precipitation objects corresponding to precipitation observation data (i.e., current precipitation objects are based on instrument monitoring and can be selected from precipitation observation data). The initial continuous rain area (CRA) precipitation verification area is the union of the forecast precipitation objects and current precipitation objects. Furthermore, a precipitation object can refer to a precipitation area within a preset range.

[0068] It should be noted that the following steps S220 to S240 can be performed on the original forecast precipitation object and the revised forecast precipitation object respectively (for example, the forecast precipitation object in step S220 and step S240 can be regarded as the original forecast precipitation object or the revised forecast precipitation object), so as to obtain the analysis results of the original forecast precipitation object and the analysis results of the revised forecast precipitation object. The analysis results of the original forecast precipitation object and the analysis results of the revised forecast precipitation object can be further compared to achieve the purpose of Figure 1 Verification of the advantages of the corrected method for typhoon precipitation forecast shown compared to the existing typhoon precipitation forecast method.

[0069] Specifically, all the predicted precipitation objects in the specified area can be determined according to the preset precipitation threshold P. (in, The grid precipitation values ​​in the grid are all greater than or equal to P), the real-time precipitation object (in, The grid precipitation values ​​in the initial continuous rain area CRA precipitation test area CRA origin =Af ∪A o .

[0070] In step S220, the rainfall center of the live precipitation object is determined, and an optimal horizontal displacement distance of the forecast precipitation object is searched within a preset range around the rainfall center of the live precipitation object to determine an intermediate matching region of the forecast precipitation object. The union of the intermediate matching region and the initial CRA precipitation verification region is used as the first CRA precipitation verification region. The optimal horizontal displacement distance is used to maximize the first correlation coefficient between the forecast precipitation field and the live precipitation field within the first CRA precipitation verification region, and when the first correlation coefficient is maximized, minimize the mean square error between the forecast precipitation field and the live precipitation field within the first CRA precipitation verification region.

[0071] Specifically, calculate the forecast precipitation object The predicted rainfall center position C f , and the live rainfall center position C of the live precipitation object can also be calculated o , and can be located at the center of the actual rainfall C o Search for precipitation objects within the specified surrounding area The optimal horizontal displacement distance , and get the middle matching area A f ´, so that the first CRA precipitation test area CRA moved = A f ´∪CRA origin Correlation coefficient (CORR) between the internal forecast precipitation field and the actual precipitation field fo Maximum, and when the correlation coefficient CORR fo At maximum, the first CRA precipitation inspection area CRA moved The mean square error (MSE) between the predicted precipitation field and the actual precipitation field in the region fo ) minimum.

[0072] Step S230: Determine an optimal rotation angle for the intermediate matching region, rotate the intermediate matching region based on the optimal rotation angle, obtain a final matching region for the forecast precipitation object, and use the union of the final matching region and the initial CRA precipitation verification region as the second CRA precipitation verification region. The optimal rotation angle is designed to maximize the second correlation coefficient between the forecast precipitation field and the actual precipitation field within the second CRA precipitation verification region, and minimize the mean square error between the forecast precipitation field and the actual precipitation field within the second CRA precipitation verification region when the second correlation coefficient is maximized.

[0073] Specifically, in the middle matching area A f ', search for the middle matching area A f Optimal rotation angle , and get the final matching area A f ", making the second CRA precipitation test area CRA shifted =A f "∪CRA origin Correlation coefficient (CORR) between the internal forecast precipitation field and the actual precipitation field fo Maximum, and when the correlation coefficient CORR fo At maximum, the second CRA precipitation inspection area CRA shifted The mean square error (MSE) between the predicted precipitation field and the actual precipitation field in the region fo ) minimum.

[0074] Step S240 : Analyze the error of the typhoon precipitation forecast based on the initial CRA precipitation verification area, the first CRA precipitation verification area, and the second CRA precipitation verification area to obtain an error analysis result.

[0075] Specifically, the mean square error (MSE) of the predicted precipitation can be expressed as total It is decomposed into the following four items: center offset error, rotation error, total error and morphological error. And the calculation formula for each type of error is as follows:

[0076] ;

[0077] ;

[0078] ;

[0079] ;

[0080] in, It is used to indicate the center offset error, which reflects the forecast error caused by the position deviation of the forecast precipitation object; It is used to represent the standard deviation of the forecast precipitation in the initial CRA precipitation verification area; Used to represent the standard deviation of actual precipitation in the initial CRA precipitation test area; It is used to represent the spatial correlation between the forecast precipitation and the actual precipitation in the first CRA precipitation verification area; r is used to represent the spatial correlation between the forecast precipitation and the actual precipitation in the initial CRA precipitation verification area; It is used to represent the rotation error, which reflects the impact of the typhoon's rotation on the predicted precipitation distribution; It is used to express the spatial correlation between the forecast precipitation and the actual precipitation in the second CRA precipitation test area; It is used to express the total amount error, which reflects the difference between the predicted mean precipitation intensity and the actual situation; It is used to represent the average value of the forecast precipitation in the second CRA precipitation test area; Used to represent the average value of actual precipitation in the second CRA precipitation test area; It is used to represent the morphological error, which reflects the difference between the distribution shape of predicted precipitation and the actual precipitation.

[0081] Additionally, the center offset error can be calculated and rotation error The sum is used as the moving error , which reflects the forecast error associated with the precipitation position offset.

[0082] Therefore, based on the comparative study of path correction and CRA shifting for precipitation forecast improvement, the embodiment of the present application proposes a method for analyzing the sources of typhoon precipitation forecast errors, which can accurately analyze the sources of typhoon precipitation misalarms, thereby providing a basis for improving typhoon precipitation forecasts.

[0083] In order to facilitate understanding of the embodiments of the present application, a description is given below using specific embodiments.

[0084] Specifically, the 6-hour interval model precipitation forecast is first corrected (or corrected) based on the optimal path, that is, the precipitation forecast position in the corresponding period is adjusted according to the position offset between the original forecast path (i.e., the typhoon forecast path of the original forecast data) and the optimal path. Then, a comparative analysis is performed with the model forecast precipitation after CRA mobile matching, which can be seen in Figures 3a to 3f Among them, the improvement comparison of precipitation forecast after path correction (3a, 3c, 3e) and CRA shifting (3b, 3d, 3f), (3a, 3c, 3e) corresponds to different precipitation levels, and Figure 3a 、 3c The horizontal axis of 3e is in mm, and (3b, 3d, 3f) correspond to different forecast periods, and Figure 3b 、 Figure 3d and Figure 3f The horizontal axis unit is h. And (3a, 3b) represents the correlation, and its vertical axis unit is dimensionless, and (3c, 3d) the root mean square error, and Figure 3c and Figure 3d The vertical axis unit is , and (3e, 3f) represents the ETS score, and Figure 3e and Figure 3f The vertical axis unit is dimensionless.

[0085] The results show that for all precipitation magnitudes, the correlation (CC) of model precipitation improved in more samples than it decreased after path correction. Similar to the correlation, the root mean square error (RMSE) decreased for most samples after path correction. While the improvement in ETS was less pronounced than in the previous two, the improvement was most pronounced when the precipitation magnitude reached 100 mm, with an extreme improvement approaching 0.4. The improvement in ETS decreased somewhat when the precipitation magnitude reached 250 mm. The improvements across different forecast periods show that path correction does not significantly improve precipitation within a 24-hour timeframe, with the 50th percentile values ​​of the CC, RMSE, and ETS improvements approaching 0. However, the improvement significantly increases after 24 hours. These analyses indicate that path correction can generally reduce model precipitation errors, particularly for typhoon precipitation forecasts beyond 24 hours, and improves precipitation forecasts across all magnitudes. In addition, the distribution of typhoon precipitation is not always directly related to the path. The structure of the typhoon's spiral rain belt and the relative position between the typhoon and the external circulation will have an impact on the former. Therefore, path correction may not necessarily have a positive improvement effect on precipitation forecasts. Of course, the proportion of such cases is relatively small.

[0086] It should be understood that the above-mentioned analysis method of the sources of typhoon precipitation forecast errors is merely exemplary, and those skilled in the art may make various modifications based on the above-mentioned method, and the modified schemes also fall within the scope of protection of this application.

[0087] The present application provides an electronic device having a computer program stored thereon. The computer program is executed by a processor to execute the method described in the embodiment.

[0088] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0089] The present invention is described with reference to flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each process flow and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions.

[0090] It should be noted that, in the claims, any reference signs placed between brackets shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention may be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In claims enumerating several means, several of these means may be embodied by one and the same hardware. The use of the words first, second, third etc. is for convenience only and does not indicate any order. These words may be understood as part of the component name.

[0091] In addition, it should be noted that, in the description of this specification, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0092] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments after learning the basic creative concept. Therefore, the claims should be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0093] Obviously, those skilled in the art may make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if such modifications and variations fall within the scope of the claims and their equivalents, the present invention shall also include such modifications and variations.

Claims

1. A correction method for typhoon precipitation forecast, characterized in that: include: Obtaining a forecast data set including original forecast data of a target typhoon; wherein each forecast data in the forecast data set includes a typhoon forecast path, and the original forecast data also includes original typhoon precipitation forecast data; According to the real-time position of the target typhoon, N typhoon forecast paths with the smallest path errors are selected from the multiple typhoon forecast paths in the forecast data set and arithmetic average is performed to obtain an optimal forecast path; wherein N is a preset positive integer; Determining a position offset between the typhoon forecast path of the original forecast data and the optimal forecast path; Correcting the original typhoon precipitation forecast data based on the position offset to obtain corrected typhoon precipitation forecast data for the optimal forecast path; Wherein, determining the position offset between the typhoon forecast path of the original forecast data and the optimal forecast path includes: Dividing the typhoon forecast path of the original forecast data and the optimal forecast path into m sub-paths based on the precipitation interval output by the model; wherein m is a positive integer; Calculating respectively the average position offset between each sub-path in the m sub-paths of the typhoon forecast path of the original forecast data and a sub-path of its corresponding optimal forecast path; The correcting the original typhoon precipitation forecast data based on the position offset to obtain the corrected typhoon precipitation forecast data of the optimal forecast path includes: Determining precipitation grid data within a preset range around the typhoon forecast path of the original forecast data; Based on each average position offset, the position of the corresponding precipitation grid point data is moved while keeping the precipitation value unchanged, and m precipitation correction areas are accumulated to obtain a corrected precipitation area of ​​the optimal forecast path and the precipitation within the corrected precipitation area.

2. A method for analyzing the sources of error in typhoon precipitation forecasts, characterized in that: The analysis method is used to analyze error sources of original forecast data and revised typhoon precipitation forecast data obtained by correcting the original forecast data, and the revised typhoon precipitation forecast data is obtained by the typhoon precipitation forecast correction method according to claim 1, and the analysis method includes: Based on a preset precipitation threshold, determining a forecast precipitation object, an actual precipitation object, and an initial continuous rain area CRA precipitation verification area within a designated area; wherein the forecast precipitation object includes an original forecast precipitation object corresponding to the original forecast data and a revised forecast precipitation object corresponding to the revised typhoon precipitation forecast data, and the initial continuous rain area CRA precipitation verification area is a union of the forecast precipitation object and the actual precipitation object; Determining the rainfall center position of the live precipitation object, and searching for an optimal horizontal displacement distance of the forecast precipitation object within a preset range around the rainfall center position of the live precipitation object to determine an intermediate matching region of the forecast precipitation object, and using the union of the intermediate matching region and the initial continuous rain area CRA precipitation verification region as a first CRA precipitation verification region; wherein the optimal horizontal displacement distance is used to maximize a first correlation coefficient between the forecast precipitation field and the live precipitation field within the first CRA precipitation verification region, and when the first correlation coefficient is maximized, minimize a mean square error between the forecast precipitation field and the live precipitation field within the first CRA precipitation verification region; Determining an optimal rotation angle for the intermediate matching area, rotating the intermediate matching area based on the optimal rotation angle to obtain a final matching area for the forecast precipitation object, and using the union of the final matching area and the initial continuous rain area CRA precipitation verification area as a second CRA precipitation verification area; wherein the optimal rotation angle is used to maximize a second correlation coefficient between the forecast precipitation field and the actual precipitation field within the second CRA precipitation verification area, and when the second correlation coefficient is maximized, minimize a mean square error between the forecast precipitation field and the actual precipitation field within the second CRA precipitation verification area; An error in typhoon precipitation forecast is analyzed based on the initial continuous rain area CRA precipitation verification area, the first CRA precipitation verification area, and the second CRA precipitation verification area to obtain an error analysis result.

3. The analysis method according to claim 2, characterized in that The error analysis result includes the center offset error; the center offset error is obtained by the following formula: ； in, Used to represent the center offset error; S F It is used to represent the standard deviation of the forecast precipitation in the initial continuous rain area CRA precipitation test area; Used to represent the standard deviation of actual precipitation in the initial continuous rain area CRA precipitation test area; It is used to represent the spatial correlation between the forecast precipitation and the actual precipitation in the first CRA precipitation test area; r is used to represent the spatial correlation between the forecast precipitation and the actual precipitation in the initial continuous rain area CRA precipitation test area.

4. The analysis method according to claim 3, characterized in that The error analysis result also includes a rotation error; the rotation error is obtained by the following formula: ； in, Used to represent the rotation error; Used to represent the spatial correlation between forecast precipitation and actual precipitation in the second CRA precipitation verification area.

5. The analysis method according to claim 4, characterized in that The error analysis result also includes a total error; the total error is obtained by the following formula: ； in, Used to represent the total amount error; It is used to represent the average value of the forecast precipitation in the second CRA precipitation test area; Used to represent the average value of actual precipitation in the second CRA precipitation test area.

6. The analysis method according to claim 5, characterized in that The error analysis result also includes morphological error; the morphological error is obtained by the following formula: ； in, Used to represent the morphological error.

7. The analysis method according to claim 5, characterized in that The error analysis result also includes a movement error, which is the sum of the center offset error and the rotation error.

8. An electronic device, characterized in that, The electronic device stores a computer program, and when the computer program is executed by the processor, the computer program executes the correction method for typhoon precipitation forecast according to claim 1 and the analysis method for the error source of typhoon precipitation forecast according to any one of claims 2 to 7.

Images