Rainfall-increasing operation effect evaluation method and device based on intelligent prediction of natural rainfall

Abstract

The application provides a rain enhancement operation effect evaluation method and device for intelligently predicting natural rainfall, and the method comprises the following steps: determining an operation influence area and a comparison area based on the range of artificial rain enhancement operation; the comparison area is not affected by the catalyst of the artificial rain enhancement operation; respectively performing rainfall amount extrapolation prediction on the comparison area and the operation influence area through a rainfall prediction model to obtain rainfall prediction data of the comparison area and rainfall prediction data of the operation influence area; respectively acquiring real rainfall data of the comparison area and real rainfall data of the operation influence area; obtaining rainfall prediction error of the rainfall prediction model based on the rainfall prediction data of the comparison area and the real rainfall data of the comparison area; and obtaining absolute rainfall increase of the artificial rain enhancement operation based on the rainfall prediction data of the operation influence area, the real rainfall data of the operation influence area and the rainfall prediction error of the rainfall prediction model. The accuracy of effect evaluation is improved.

Description

Technical Field

[0001] The present application relates to the technical field of rain enhancement operation effect evaluation, and in particular to a rain enhancement operation effect evaluation method, device, electronic device and storage medium for intelligently predicting natural rainfall. Background Art

[0002] The vitality of weather modification science depends on the effectiveness of its applications and its role in disaster reduction and mitigation. Therefore, effectiveness evaluation is an essential and unavoidable component of cloud enhancement operations, occupying a crucial position within the overall program. Scientifically and objectively measuring the effectiveness of cloud enhancement operations can help understand cloud response mechanisms under varying operating conditions, analyze variations in rainfall output and its relationship to operational parameters, and optimize key factors such as operation timing, catalyst type, and catalyst dosage. This is crucial for improving cloud and precipitation theory and enhancing the efficiency and service levels of weather modification operations. However, due to the uncertainty of the evaluation targets and the significant natural variability in cloud and precipitation data, effectively verifying the effectiveness of rain enhancement within this variability presents significant challenges. Consequently, effectiveness verification remains a global challenge.

[0003] Traditional effect verification methods (physical testing, statistical testing, and numerical model testing) have many problems and difficulties, mainly manifested in: limitations of observation methods, imperfections in statistical methods, the need to improve the reliability of numerical models, incomplete understanding of cloud and precipitation patterns, and the mechanisms of artificial weather modification, which affect the accuracy of effect evaluation. Summary of the Invention

[0004] The present application aims to solve one of the technical problems in the related art at least to a certain extent.

[0005] To this end, the first purpose of this application is to propose a method for evaluating the effectiveness of rainmaking operations by intelligently predicting natural rainfall, so as to improve the accuracy of the effect evaluation.

[0006] The second purpose of this application is to propose a rain-making operation effect evaluation device that can intelligently predict natural rainfall.

[0007] The third objective of this application is to provide an electronic device.

[0008] The fourth object of this application is to provide a computer-readable storage medium.

[0009] A fifth object of this application is to provide a computer program product.

[0010] To achieve the above objectives, the first embodiment of the present application proposes a method for evaluating the effectiveness of rain enhancement operations by intelligently predicting natural rainfall, comprising:

[0011] Determining an operation-affected area and a comparison area based on the scope of the artificial rainmaking operation; the comparison area is not affected by the catalyst of the artificial rainmaking operation;

[0012] Using a rainfall prediction model, rainfall extrapolation prediction is performed on the comparison area and the operation impact area respectively, to obtain rainfall prediction data for the comparison area and rainfall prediction data for the operation impact area;

[0013] respectively obtaining the actual rainfall data of the comparison area and the actual rainfall data of the operation impact area;

[0014] Obtaining a rainfall prediction error of the rainfall prediction model based on the rainfall prediction data of the comparison area and the actual rainfall data of the comparison area;

[0015] Based on the rainfall prediction data of the operation impact area, the actual rainfall data of the operation impact area and the rainfall prediction error of the rainfall prediction model, the absolute rainfall increase of the artificial rainmaking operation is obtained, and the absolute rainfall increase of the artificial rainmaking operation is used as an evaluation indicator of the rainmaking operation effect.

[0016] In some implementations, before performing rainfall extrapolation prediction on the comparison area and the operation impact area respectively using a rainfall prediction model, the method includes:

[0017] Determining the rainfall type within the scope of the artificial rain enhancement operation; wherein the rainfall type includes stratiform cloud rainfall and convective cloud rainfall, and the rainfall type is divided according to the maximum rainfall intensity;

[0018] The rainfall extrapolation model corresponding to the rainfall type is selected as the rainfall prediction model.

[0019] In some implementations, the method for training the rainfall prediction model includes:

[0020] A sample set is constructed based on minute-grid precipitation data. Each sample in the sample set includes 2L consecutive hours of rainfall data. The K-minute grid-point rainfall data for the first L hours of each sample is used as input parameters, and the K-minute grid-point rainfall data for the last L hours of each sample is used as label data.

[0021] The PredRNN++ model is trained using the sample set to obtain a rainfall extrapolation model.

[0022] In some implementations, the step of performing rainfall extrapolation prediction on the comparison area and the operation-affected area using a rainfall prediction model to obtain rainfall prediction data for the comparison area and rainfall prediction data for the operation-affected area includes:

[0023] Based on the time of the artificial rain enhancement operation, obtaining first rainfall data collected by a rainfall station in the comparison area in a first time period before the artificial rain enhancement operation;

[0024] Based on the first rainfall data, performing rainfall extrapolation prediction on the comparison area using the rainfall prediction model to obtain rainfall prediction data for the comparison area in a second time period;

[0025] Based on the time of the artificial rain enhancement operation, obtaining second rainfall data collected by a rainfall station in the operation impact area in the first time period before the artificial rain enhancement operation;

[0026] Based on the second rainfall data, rainfall extrapolation prediction is performed on the operation impact area using the rainfall prediction model to obtain rainfall prediction data for the operation impact area in the second time period.

[0027] In some implementations, respectively obtaining the actual rainfall data of the comparison area and the actual rainfall data of the operation-affected area includes:

[0028] Obtaining actual rainfall data collected by n rainfall stations in the comparison area;

[0029] Obtain actual rainfall data collected by m rainfall stations in the operation impact area.

[0030] In some implementations, obtaining a rainfall prediction error of the rainfall prediction model based on the rainfall prediction data of the comparison area and the actual rainfall data of the comparison area includes:

[0031] Calculating the difference between the actual rainfall data collected by n rainfall stations in the comparison area and the rainfall forecast data in the comparison area respectively to obtain n differences;

[0032] The average value of the n differences is used as the rainfall prediction error of the rainfall prediction model.

[0033] In some implementations, obtaining the absolute rainfall increase of the artificial rainfall enhancement operation based on rainfall prediction data of the operation influence area, actual rainfall data of the operation influence area, and rainfall prediction error of the rainfall prediction model includes:

[0034] Based on the rainfall forecast data of the operation impact area and the actual rainfall data collected by m rainfall stations in the operation impact area, the average absolute rainfall increase of the m rainfall stations in the operation impact area is obtained;

[0035] Calculating the difference between the average absolute rainfall increase of m rainfall stations in the operation impact area and the rainfall prediction error of the rainfall prediction model to obtain the absolute rainfall increase of a single rainfall station in the operation impact area;

[0036] The absolute amount of rainfall increase of the artificial rainmaking operation is obtained by calculating the product of the absolute amount of rainfall increase of a single rainfall station in the operation influence area and the area of ​​the operation influence area.

[0037] To achieve the above-mentioned purpose, the second embodiment of the present application proposes a rain enhancement operation effect evaluation device for intelligently predicting natural rainfall, comprising:

[0038] A region division module is used to determine an operation-affected area and a comparison area based on the scope of the artificial rain enhancement operation; the comparison area is not affected by the catalyst of the artificial rain enhancement operation;

[0039] A rainfall prediction module is used to perform rainfall extrapolation prediction on the comparison area and the operation impact area respectively through a rainfall prediction model to obtain rainfall prediction data for the comparison area and rainfall prediction data for the operation impact area;

[0040] A rainfall acquisition module, used to respectively acquire the actual rainfall data of the comparison area and the actual rainfall data of the operation impact area;

[0041] an error calculation module, configured to obtain a rainfall prediction error of the rainfall prediction model based on the rainfall prediction data of the comparison area and the actual rainfall data of the comparison area;

[0042] An effect evaluation module is used to obtain the absolute rainfall increase of the artificial rainmaking operation based on the rainfall prediction data of the operation influence area, the actual rainfall data of the operation influence area and the rainfall prediction error of the rainfall prediction model, and use the absolute rainfall increase of the artificial rainmaking operation as an evaluation indicator of the effect of the rainmaking operation.

[0043] In some implementations, the rainfall prediction module is further configured to:

[0044] Determining the rainfall type within the scope of the artificial rain enhancement operation; wherein the rainfall type includes stratiform cloud rainfall and convective cloud rainfall, and the rainfall type is divided according to the maximum rainfall intensity;

[0045] The rainfall extrapolation model corresponding to the rainfall type is selected as the rainfall prediction model.

[0046] In some implementations, the apparatus further includes a model training module configured to:

[0047] A sample set is constructed based on minute-grid precipitation data. Each sample in the sample set includes 2L consecutive hours of rainfall data. The K-minute grid-point rainfall data for the first L hours of each sample is used as input parameters, and the K-minute grid-point rainfall data for the last L hours of each sample is used as label data.

[0048] The PredRNN++ model is trained using the sample set to obtain a rainfall extrapolation model.

[0049] In some implementations, the rainfall prediction module is specifically configured to:

[0050] Based on the time of the artificial rain enhancement operation, obtaining first rainfall data collected by a rainfall station in the comparison area in a first time period before the artificial rain enhancement operation;

[0051] Based on the first rainfall data, performing rainfall extrapolation prediction on the comparison area using the rainfall prediction model to obtain rainfall prediction data for the comparison area in a second time period;

[0052] Based on the time of the artificial rain enhancement operation, obtaining second rainfall data collected by a rainfall station in the operation impact area in the first time period before the artificial rain enhancement operation;

[0053] Based on the second rainfall data, rainfall extrapolation prediction is performed on the operation impact area using the rainfall prediction model to obtain rainfall prediction data for the operation impact area in the second time period.

[0054] In some implementations, the rainfall acquisition module is specifically configured to:

[0055] Obtaining actual rainfall data collected by n rainfall stations in the comparison area;

[0056] Obtain actual rainfall data collected by m rainfall stations in the operation impact area.

[0057] In some implementations, the error calculation module is specifically configured to:

[0058] Calculating the difference between the actual rainfall data collected by n rainfall stations in the comparison area and the rainfall forecast data in the comparison area respectively to obtain n differences;

[0059] The average value of the n differences is used as the rainfall prediction error of the rainfall prediction model.

[0060] In some implementations, the effect evaluation module is specifically configured to:

[0061] Based on the rainfall forecast data of the operation impact area and the actual rainfall data collected by m rainfall stations in the operation impact area, the average absolute rainfall increase of the m rainfall stations in the operation impact area is obtained;

[0062] Calculating the difference between the average absolute rainfall increase of m rainfall stations in the operation impact area and the rainfall prediction error of the rainfall prediction model to obtain the absolute rainfall increase of a single rainfall station in the operation impact area;

[0063] The absolute amount of rainfall increase of the artificial rainmaking operation is obtained by calculating the product of the absolute amount of rainfall increase of a single rainfall station in the operation influence area and the area of ​​the operation influence area.

[0064] To achieve the above-mentioned purpose, the third aspect embodiment of the present application proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the method described in the first aspect.

[0065] To achieve the above-mentioned purpose, the fourth embodiment of the present application proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the method described in the first aspect.

[0066] To achieve the above-mentioned purpose, the fifth embodiment of the present application proposes a computer program product, including a computer program, which implements the method described in the first aspect when executed by a processor.

[0067] The method, device, electronic device and storage medium for evaluating the effect of rain-enhancing operations for intelligently predicting natural rainfall provided by the present application significantly improve the accuracy of rain-enhancing operation effect evaluation by establishing an operation influence area and a comparison area. The comparison area is used to calculate the rainfall extrapolation error of the rainfall prediction model, while the operation influence area eliminates the rainfall extrapolation error when calculating the rainfall enhancement effect. In addition, corresponding rainfall prediction models are trained for different rainfall types, which effectively reduces sample complexity, improves model training efficiency, and the accuracy of rainfall extrapolation prediction. In addition, the present invention realizes accurate extrapolation of 10-minute rainfall grid data for the next two hours through the rainfall prediction model, which is used as a natural rainfall benchmark for evaluating the effect of rain-enhancing operations. Moreover, the PredRNN++ extrapolation model, with its powerful spatiotemporal state transmission capability, not only grasps the development context of precipitation distribution in precipitation prediction, but also accurately captures the dynamic changes of precipitation intensity, laying a solid foundation for testing the effect of rain-enhancing operations. Compared with traditional methods, the rainfall extrapolation forecast provided by the PredRNN++ extrapolation model has higher temporal and spatial resolution, especially in the field of nowcasting, effectively making up for the shortcomings of traditional numerical models in time effectiveness and resolution.

[0068] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0070] Figure 1 A flow chart of a method for intelligently predicting natural rainfall and evaluating the effect of rain enhancement operations provided in an embodiment of the present application;

[0071] Figure 2 A schematic diagram of a flow chart created for the dataset provided in this application example;

[0072] Figure 3 A schematic diagram of the structure of the PredRNN++ extrapolation model provided in the embodiments of the present application;

[0073] Figure 4 A block diagram of a rain enhancement operation effect evaluation device for intelligently predicting natural rainfall provided by an embodiment of the present application;

[0074] Figure 5 A block diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0075] The following describes in detail embodiments of the present application, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0076] The following describes, with reference to the accompanying drawings, a method, device, and equipment for evaluating the effectiveness of rain-enhancement operations for intelligently predicting natural rainfall according to an embodiment of the present application.

[0077] Figure 1 A flow chart of a method for evaluating the effectiveness of rain enhancement operations by intelligently predicting natural rainfall provided in an embodiment of the present application.

[0078] It should be noted that the executor of the method for evaluating the effect of rainmaking operations for intelligently predicting natural rainfall in an embodiment of the present application is the device for evaluating the effect of rainmaking operations for intelligently predicting natural rainfall in an embodiment of the present application. The device for evaluating the effect of rainmaking operations for intelligently predicting natural rainfall can be configured in an electronic device so that the electronic device can perform the function of evaluating the effect of rainmaking operations for intelligently predicting natural rainfall.

[0079] like Figure 1 As shown, the method for evaluating the effect of rain enhancement operation by intelligently predicting natural rainfall includes the following steps:

[0080] Step S101: determining an operation-affected area and a comparison area based on the scope of the artificial rain enhancement operation; the comparison area is not affected by the catalyst of the artificial rain enhancement operation.

[0081] As an implementation method, the principles for determining the scope of the operational impact area and the comparison area in actual rainfall enhancement operations are as follows:

[0082] (1) The prevailing wind direction in the comparison area and the operation impact area should be similar.

[0083] It can be understood that if the prevailing wind directions are different, the correlation coefficient between the two areas will be poor and the analysis sensitivity will be low.

[0084] To calculate the prevailing wind direction at high altitude, the comparison area should be selected on the upwind side or on the side perpendicular to the wind direction (but this does not rule out the possibility of contamination caused by dynamic effects). If the prevailing wind direction of the comparison area and the operation-affected area is inconsistent, more than one comparison area should be selected, and different comparison areas should be selected according to different prevailing wind directions.

[0085] (2) The comparison area and the operation impact area should be as close as possible, and the comparison area should not be affected by the catalyst of the rain enhancement operation.

[0086] Generally, the comparison area and the operation impact area are about 30-50km apart.

[0087] (3) The topography and area of ​​the comparison area are similar to those of the operation impact area.

[0088] (4) The comparison area and the operation impact area are affected by the same weather system and have similar rainfall types. This is reflected in the fact that the rainfall in the comparison area is closely related to the rainfall in the operation impact area. The significance level of the sample correlation coefficient should be above 0.01.

[0089] (5) The comparison area should have a dense network of rainfall stations, just like the operation impact area.

[0090] Usually every 20-50km 2 One, depending on the nature of the precipitation and the degree to which the equipment conditions permit.

[0091] (6) The comparison area and the operation impact area have the same long time series of historical data, so as to conduct data comparison analysis between the comparison area and the operation impact area and train the rainfall prediction model.

[0092] In some embodiments, when determining the operation impact area and the comparison area, the rainfall stations of the operation impact area and the comparison area are determined at the same time.

[0093] For example, during an actual cloud seeding operation, a portion of rain gauges within a city area is designated as the operation's impact zone, totaling m. Another portion of rain gauges is designated as the comparison zone (non-operation-affected stations), totaling n. Based on these rain gauges, the scope of the operation's impact zone and the comparison zone are delineated.

[0094] The principles for determining rainfall stations in the affected area during rainfall enhancement operations are as follows:

[0095] The influence range is defined as the distance over which the wind propagates within a certain period of time above the rainfall stations in the operation's impact zone. For example, the catalyst's impact period is three hours. Therefore, the three-hour propagation distance of the wind (generally 700 hPa) above the rainfall stations in the operation's impact zone is used as the impact range. The number of rainfall stations M in the operation's impact zone is then calculated.

[0096] Step S102 : extrapolating and predicting rainfall for the comparison area and the operation impact area respectively using a rainfall prediction model to obtain rainfall prediction data for the comparison area and rainfall prediction data for the operation impact area.

[0097] The rainfall prediction model is first described in detail below.

[0098] In some embodiments, a training method for a rainfall prediction model includes: constructing a sample set based on minute grid precipitation data, each sample in the sample set includes 2L consecutive hours of rainfall data, wherein the K-minute grid rainfall data for the first L hours of each sample is used as input parameters, and the K-minute grid rainfall data for the last L hours is used as label data; training a PredRNN++ model through the sample set to obtain a rainfall extrapolation model.

[0099] It can be understood that the rainfall prediction model is a rainfall extrapolation model, which is used for rainfall extrapolation prediction.

[0100] For example, minute-grid precipitation data is used as a training sample set for training the rainfall prediction model, with a spatial resolution of 0.01 degrees, and the training sample set is stored in tensor form. According to previous experiments, the rainfall prediction model works better when using the rainfall data of the past two hours to predict the rainfall data of the next two hours. Therefore, L is 2, that is, 2L is 4 hours, K is 10 minutes, and the actual rainfall data of the minute grid every 10 minutes is used as the input data and label data in the training sample set. That is, the continuous rainfall data every 4 hours is used as a sample set, of which 12 (2 hours, 10 minutes sampling frequency, resulting in 12 data) rainfall data in the first 2 hours are input data, and the 12 rainfall data in the last 2 hours are output data.

[0101] It should also be noted that since rain enhancement operations use different intervention methods according to different rainfall types, the samples participating in model training and model testing are classified according to rainfall types.

[0102] As an implementation method, rainfall types are distinguished based on the maximum rainfall intensity of 5 mm / h, at which the reflectivity is approximately equal to 35 dBZ, which is an important reference indicator for stratiform cloud precipitation and convective cloud precipitation.

[0103] In some embodiments, by distinguishing the maximum rainfall intensity, convective cloud precipitation sample sets and stratiform cloud precipitation sample sets are established respectively, and training sets, validation sets and test sets are allocated in proportion to facilitate classification training and optimization of the model.

[0104] For example, the dataset production process is as follows: Figure 2 As shown, it includes obtaining grid rainfall data, continuously verifying the grid rainfall data, decoding and clipping the grid rainfall data after verification, and obtaining a data tensor; at the same time, according to the maximum rainfall intensity, the decoded and clipped grid rainfall data are classified to obtain a stratiform cloud precipitation dataset and a convective cloud precipitation dataset, and a sample set is obtained based on the dataset.

[0105] The rainfall prediction model of the embodiment of the present invention uses PredRNN++ as the basic training model. The basic unit of PredRNN++ is Causal LSTM. The Causal LSTM unit realizes the step-by-step connection of time state, space state and output results, enhances the spatiotemporal prediction capability, and adds a gradient fast transfer unit (GHU) to prevent the disappearance of gradient. The model structure is as follows: Figure 3 As shown, it is the PredRNN++ extrapolation model.

[0106] During model training, the convective cloud precipitation extrapolation model and the stratiform cloud precipitation extrapolation model are trained separately. Since the rainfall intensity of different precipitation types is different, different normalization parameters can be used for training. The model training uses the Adam optimizer for training, and according to the performance of the trained rainfall prediction model on the validation data set, an automatic stop strategy is formulated to prevent overfitting. Among them, in view of the uneven distribution of rainfall intensity, the loss function of the rainfall prediction model is designed by combining MAE and RMAE to increase the complexity of the loss function, thereby improving the generalization ability of the rainfall prediction model. As shown in the following formula (1), the values ​​of the proportional coefficients α and β in formula (1) are determined by the effect test of the rainfall prediction model, and finally the minute rainfall extrapolation model is obtained, that is, the rainfall prediction model of the present invention.

[0107] loss=α·MAE(y',y)+β·RMSE(y',y) (1)

[0108] In the model verification stage, three indicators, namely critical success index (CSI), hit rate (POD), and false alarm rate (FAR), are introduced to verify the rainfall forecast data for the next two hours. The indicator formula is shown in Equation (2), where a represents the number of hits, b represents the number of false alarms, and c represents the number of missed alarms.

[0109]

[0110] For example, the model effect test of the rainfall prediction model was completed using the data set in 2019-2022 when no rain enhancement operations were carried out. According to the sample characteristics of convective cloud precipitation and stratiform cloud precipitation, the impact of rain enhancement operations on rainfall of different magnitudes was tested, and the extrapolation effect of the model was tested at four rainfall intensities of 1mm, 2mm, 5mm and above 5mm.

[0111] In order to test the impact of rain enhancement operations in the next two hours, the extrapolation effect of the model is tested every 10 minutes, so as to realize the extrapolation prediction of rainfall in different rainfall periods. The predicted rainfall is used as the rainfall of natural clouds (without considering the catalytic effect) during the catalytic process of artificial rain enhancement operations, and the next step of rain enhancement effect evaluation is carried out.

[0112] Therefore, in some embodiments, before performing rainfall extrapolation prediction on the comparison area and the operation impact area respectively through the rainfall prediction model, it includes: determining the rainfall type within the range of the artificial rainmaking operation; wherein the rainfall types include stratiform cloud rainfall and convective cloud rainfall, and the rainfall types are divided according to the maximum rainfall intensity; and selecting the rainfall extrapolation model corresponding to the rainfall type as the rainfall prediction model.

[0113] Therefore, by modeling stratiform cloud precipitation and convective cloud precipitation separately during the model training stage, the sample complexity is effectively reduced and the training efficiency is improved.

[0114] After determining the rainfall prediction model, as an implementation method, rainfall extrapolation prediction is performed on the comparison area and the operation impact area respectively through the rainfall prediction model to obtain rainfall prediction data for the comparison area and rainfall prediction data for the operation impact area; including: based on the time of the artificial rainmaking operation, obtaining the first rainfall data collected by the rainfall station in the comparison area in the first time period before the artificial rainmaking operation; based on the first rainfall data, extrapolating the rainfall in the comparison area through the rainfall prediction model to obtain rainfall prediction data for the comparison area in the second time period; based on the time of the artificial rainmaking operation, obtaining the second rainfall data collected by the rainfall station in the operation impact area in the first time period before the artificial rainmaking operation; based on the second rainfall data, extrapolating the rainfall in the operation impact area through the rainfall prediction model to obtain rainfall prediction data for the operation impact area in the second time period.

[0115] For example, according to the actual rain enhancement operation time, through the rainfall prediction model, the 10-minute rainfall data of a single rainfall station in the comparison area within 2 hours before the actual rain enhancement operation is selected as the input parameter of the rainfall prediction model, and the rainfall prediction model is used to calculate the 10-minute rainfall prediction data of the rainfall station in the next 2 hours.

[0116] Step S103 , respectively obtaining the actual rainfall data of the comparison area and the actual rainfall data of the operation impact area.

[0117] As an implementation method, a method for respectively obtaining the real rainfall data of the comparison area and the real rainfall data of the operation impact area includes: obtaining the real rainfall data collected by n rain stations in the comparison area; obtaining the real rainfall data collected by m rain stations in the operation impact area.

[0118] Step S104: obtaining a rainfall prediction error of the rainfall prediction model based on the rainfall prediction data and the actual rainfall data of the comparison area.

[0119] As an implementation method, a method for obtaining the rainfall prediction error of a rainfall prediction model based on the rainfall prediction data and the actual rainfall data in the comparison area includes: calculating the difference between the actual rainfall data collected by n rainfall stations in the comparison area and the rainfall prediction data in the comparison area, to obtain n differences; and taking the average value of the n differences as the rainfall prediction error of the rainfall prediction model.

[0120] For example, according to the actual rainfall enhancement operation time, the rainfall prediction model is used to select the 10-minute rainfall data of a single rainfall station in the comparison area within 2 hours before the actual rainfall enhancement operation starts as the input parameter of the rainfall prediction model, and the rainfall prediction model is used to calculate the 10-minute rainfall prediction data of the rainfall station in the next 2 hours; the rainfall prediction data of the first 10 minutes in the next 2 hours is recorded as X' i The actual rainfall data of the first 10 minutes of the rainfall station in the comparison area is recorded as X i , then the rainfall estimation error for the first 10 minutes at this rainfall station is X' i -X i Then, the rainfall estimation errors of n rainfall stations in the comparison area are counted, and the average rainfall estimation error of the first 10 minutes is calculated as X ave for:

[0121]

[0122] The average value X ave is the 10-minute average rainfall estimation error of multiple rainfall stations in the comparison area.

[0123] Step S105, based on the rainfall prediction data of the operation impact area, the actual rainfall data of the operation impact area and the rainfall prediction error of the rainfall prediction model, the absolute rainfall increase of the artificial rainfall operation is obtained, and the absolute rainfall increase of the artificial rainfall operation is used as an evaluation indicator of the rainfall operation effect.

[0124] As an implementation method, a method for obtaining the absolute rainfall increase of an artificial rainmaking operation based on rainfall forecast data of an operation influence area, actual rainfall data of the operation influence area and rainfall prediction error of a rainfall prediction model comprises: obtaining the average absolute rainfall increase of the m rainfall stations in the operation influence area based on rainfall forecast data of the operation influence area and actual rainfall data collected by m rainfall stations in the operation influence area; calculating the difference between the average absolute rainfall increase of the m rainfall stations in the operation influence area and the rainfall prediction error of the rainfall prediction model to obtain the absolute rainfall increase of a single rainfall station in the operation influence area; calculating the product of the absolute rainfall increase of a single rainfall station in the operation influence area and the area of ​​the operation influence area to obtain the absolute rainfall increase of the artificial rainmaking operation.

[0125] Taking the above example as an example, according to the actual rainfall enhancement operation time, the 10-minute rainfall data of a single rainfall station (for example, the first rainfall station) in the operation impact area within 2 hours before the actual rainfall enhancement operation is selected as the input parameter of the rainfall prediction model. The rainfall prediction model is used to calculate the 10-minute rainfall prediction data of this rainfall station in the next 2 hours. The rainfall prediction data of the first 10 minutes in the next 2 hours is recorded as Y' j The actual rainfall data of the first 10 minutes of the rainfall station is recorded as Y j , then the absolute rainfall increase in the first 10 minutes of the rainfall station is Y j -Y' j Then, the absolute rainfall increase of the m rainfall stations in the operation impact area is counted, and the average value Y of the absolute rainfall increase in the first 10 minutes is calculated. ave for:

[0126]

[0127] The average value Yave is the average value of the absolute rainfall increase within 10 minutes at multiple rainfall stations in the operation impact area.

[0128] According to the scope of the operation influence zone, the area G of the operation influence zone is estimated, and the absolute rainfall increase Ra of the 10-minute rainfall enhancement operation in the operation influence zone is:

[0129] R a =(Y ave -X ave )×G(5)

[0130] The relative rainfall rate R of a 10-minute rainfall enhancement operationr for:

[0131]

[0132] By analogy, the absolute rainfall increase for every 10 minutes in the next 2 hours after the rain enhancement operation can be obtained, and the absolute rainfall increase for the entire rain enhancement operation can be accumulated.

[0133] It should be noted that, combined with the duration of the rain enhancement operation, the absolute rainfall increase of multiple rainfall stations in multiple time periods can be obtained according to formula (5), thereby obtaining the absolute rainfall increase of the entire rain enhancement operation. The absolute rainfall increase of the entire rain enhancement operation can also be obtained by combining the rainfall increase rate obtained according to formula (6) with the actual rainfall data collected in the operation impact area.

[0134] The method for evaluating the effect of rain-enhancing operations by intelligently predicting natural rainfall in the embodiment of the present application, by establishing an operation influence area and a comparison area, the comparison area is used to calculate the rainfall extrapolation error of the rainfall prediction model, and the operation influence area eliminates the rainfall extrapolation error when calculating the rain-enhancing effect, thereby significantly improving the accuracy of the evaluation of the effect of rain-enhancing operations. In addition, corresponding rainfall prediction models are trained for different rainfall types, which effectively reduces the sample complexity, improves the model training efficiency, and the accuracy of rainfall extrapolation prediction. In addition, the present invention realizes the accurate extrapolation of K-minute rainfall grid data for the next two hours through the rainfall prediction model, which is used as the natural rainfall benchmark for evaluating the rain-enhancing effect. And the PredRNN++ extrapolation model, with its powerful spatiotemporal state transmission capability, not only grasps the development context of precipitation distribution in precipitation prediction, but also accurately captures the dynamic changes of precipitation intensity, laying a solid foundation for the inspection of rain-enhancing operation effects. Compared with traditional methods, the rainfall extrapolation forecast provided by the PredRNN++ extrapolation model has higher temporal and spatial resolution, especially in the field of nowcasting, effectively making up for the shortcomings of traditional numerical models in time effectiveness and resolution.

[0135] In order to implement the above embodiment, the present application also proposes a rain enhancement operation effect evaluation device that intelligently predicts natural rainfall. Figure 4 This is a schematic diagram of the structure of a rain enhancement operation effect evaluation device for intelligently predicting natural rainfall provided by an embodiment of the present application. Figure 4 As shown, the rain enhancement operation effect evaluation device for intelligently predicting natural rainfall may include: a region division module 401, a rainfall prediction module 402, a rainfall prediction module 403, an error calculation module 404 and an effect evaluation module 405.

[0136] The region division module 401 is used to determine the operation impact area and the comparison area based on the scope of the artificial rain enhancement operation; the comparison area is not affected by the catalyst of the artificial rain enhancement operation;

[0137] A rainfall prediction module 402 is configured to perform rainfall extrapolation prediction on the comparison area and the operation impact area respectively using a rainfall prediction model to obtain rainfall prediction data for the comparison area and rainfall prediction data for the operation impact area;

[0138] The rainfall prediction module 403 is used to obtain the actual rainfall data of the comparison area and the actual rainfall data of the operation impact area respectively;

[0139] An error calculation module 404 is configured to obtain a rainfall prediction error of the rainfall prediction model based on the rainfall prediction data and the actual rainfall data of the comparison area;

[0140] The effect evaluation module 405 is used to obtain the absolute rainfall increase of the artificial rainmaking operation based on the rainfall prediction data of the operation impact area, the actual rainfall data of the operation impact area and the rainfall prediction error of the rainfall prediction model, and use the absolute rainfall increase of the artificial rainmaking operation as an evaluation indicator of the effect of the rainmaking operation.

[0141] In some implementations, the rainfall prediction module 402 is further configured to:

[0142] Determine the rainfall type within the scope of artificial rain enhancement operations; rainfall types include stratiform cloud rainfall and convective cloud rainfall, and rainfall types are divided according to maximum rainfall intensity;

[0143] The rainfall extrapolation model corresponding to the rainfall type is selected as the rainfall prediction model.

[0144] In some implementations, the apparatus further includes a model training module 406 configured to:

[0145] A sample set is constructed based on minute-grid precipitation data. Each sample in the sample set includes 2L consecutive hours of rainfall data. The K-minute grid-point rainfall data for the first L hours of each sample is used as input parameters, and the K-minute grid-point rainfall data for the last L hours of each sample is used as label data.

[0146] The PredRNN++ model is trained by the sample set to obtain the rainfall extrapolation model.

[0147] In some implementations, the rainfall prediction module 402 is specifically configured to:

[0148] Based on the time of the artificial rain enhancement operation, first rainfall data collected by a rainfall station in the comparison area in a first time period before the artificial rain enhancement operation is obtained;

[0149] Based on the first rainfall data, extrapolating the rainfall prediction of the comparison area through the rainfall prediction model to obtain rainfall prediction data of the comparison area in the second time period;

[0150] Based on the time of the artificial rain enhancement operation, second rainfall data collected by a rainfall station in the operation impact area in a first time period before the artificial rain enhancement operation is obtained;

[0151] Based on the second rainfall data, the rainfall in the operation impact area is extrapolated and predicted using a rainfall prediction model to obtain rainfall prediction data for the operation impact area in the second time period.

[0152] In some implementations, the rainfall prediction module 403 is specifically configured to:

[0153] Obtain the actual rainfall data collected by n rainfall stations in the comparison area;

[0154] Obtain actual rainfall data collected by m rainfall stations in the operation impact area.

[0155] In some implementations, the error calculation module 404 is specifically configured to:

[0156] Calculate the difference between the actual rainfall data collected by n rainfall stations in the comparison area and the rainfall forecast data in the comparison area to obtain n differences;

[0157] The average of the n differences is taken as the rainfall prediction error of the rainfall prediction model.

[0158] In some implementations, the effect evaluation module 405 is specifically configured to:

[0159] Based on the rainfall forecast data of the operation impact area and the actual rainfall data collected by m rainfall stations in the operation impact area, the average absolute rainfall increase of the m rainfall stations in the operation impact area is obtained;

[0160] Calculate the difference between the average absolute rainfall increase of m rainfall stations in the operation impact area and the rainfall prediction error of the rainfall prediction model to obtain the absolute rainfall increase of a single rainfall station in the operation impact area;

[0161] The absolute rainfall increase of the artificial rainmaking operation is obtained by calculating the product of the absolute rainfall increase of a single rainfall station in the operation influence area and the area of ​​the operation influence area.

[0162] It should be noted that the above explanation of the embodiment of the method for evaluating the effect of rainmaking operations by intelligently predicting natural rainfall is also applicable to the device for evaluating the effect of rainmaking operations by intelligently predicting natural rainfall in this embodiment, and will not be repeated here.

[0163] In order to implement the above embodiment, the present application also proposes an electronic device. Figure 5 , Figure 5 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present application. Figure 5As shown, the electronic device 500 includes: a processor 501, and a memory 502 communicatively connected to the processor 501; the memory 502 stores computer-executable instructions; the processor 501 executes the computer-executable instructions stored in the memory to implement the method provided in the aforementioned embodiment.

[0164] In order to implement the above embodiments, the present application also proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the methods provided by the above embodiments.

[0165] In order to implement the above embodiments, the present application also proposes a computer program product, including a computer program, which implements the methods provided by the above embodiments when executed by a processor.

[0166] The collection, storage, use, processing, transmission, provision and disclosure of user personal information involved in this application are in compliance with relevant laws and regulations and do not violate public order and good morals.

[0167] It is important to note that personal information collected from users should be used for legitimate and reasonable purposes and should not be shared or sold beyond these legitimate uses. Furthermore, such collection / sharing should be conducted only after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign an agreement / authorization that includes the relevant user information before using the feature. Furthermore, any necessary steps must be taken to safeguard and secure access to such personal information and ensure that others with access to personal information comply with its privacy policy and procedures.

[0168] This application contemplates providing implementations that allow users to selectively block the use or access of personal information data. Specifically, this disclosure contemplates providing hardware and / or software to prevent or block access to such personal information data. Risks can be minimized by limiting data collection and deleting data once it is no longer needed. Furthermore, where applicable, such personal information can be de-identified to protect user privacy.

[0169] In the descriptions of the foregoing embodiments, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean 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 application. 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 may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are mutually inconsistent.

[0170] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. Throughout the description of this application, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0171] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0172] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and a portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing it in another suitable manner if necessary, and then storing it in a computer memory.

[0173] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0174] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0175] In addition, the functional units in the various embodiments of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0176] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application. Persons skilled in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A method for evaluating the effect of rain enhancement operations by intelligently predicting natural rainfall, characterized in that: The following steps are involved: Determining an operation-affected area and a comparison area based on the scope of the artificial rainmaking operation; the comparison area is not affected by the catalyst of the artificial rainmaking operation; Using a rainfall prediction model, rainfall extrapolation prediction is performed on the comparison area and the operation impact area respectively, to obtain rainfall prediction data for the comparison area and rainfall prediction data for the operation impact area; respectively obtaining the actual rainfall data of the comparison area and the actual rainfall data of the operation impact area; Obtaining a rainfall prediction error of the rainfall prediction model based on the rainfall prediction data of the comparison area and the actual rainfall data of the comparison area; Based on the rainfall prediction data of the operation impact area, the actual rainfall data of the operation impact area and the rainfall prediction error of the rainfall prediction model, the absolute rainfall increase of the artificial rainmaking operation is obtained, and the absolute rainfall increase of the artificial rainmaking operation is used as an evaluation indicator of the rainmaking operation effect.

2. The method according to claim 1, characterized in that Before performing rainfall extrapolation prediction on the comparison area and the operation impact area respectively using a rainfall prediction model, the method includes: Determining the rainfall type within the scope of the artificial rain enhancement operation; wherein the rainfall type includes stratiform cloud rainfall and convective cloud rainfall, and the rainfall type is divided according to the maximum rainfall intensity; The rainfall extrapolation model corresponding to the rainfall type is selected as the rainfall prediction model.

3. The method according to claim 1, characterized in that The training method of the rainfall prediction model comprises: A sample set is constructed based on minute-grid precipitation data. Each sample in the sample set includes 2L consecutive hours of rainfall data. The K-minute grid-point rainfall data for the first L hours of each sample is used as input parameters, and the K-minute grid-point rainfall data for the last L hours of each sample is used as label data. The PredRNN++ model is trained using the sample set to obtain a rainfall extrapolation model.

4. The method according to claim 1, wherein The method of performing rainfall extrapolation prediction on the comparison area and the operation-affected area respectively by using a rainfall prediction model to obtain rainfall prediction data for the comparison area and rainfall prediction data for the operation-affected area comprises: Based on the time of the artificial rain enhancement operation, obtaining first rainfall data collected by a rainfall station in the comparison area in a first time period before the artificial rain enhancement operation; Based on the first rainfall data, performing rainfall extrapolation prediction on the comparison area using the rainfall prediction model to obtain rainfall prediction data for the comparison area in a second time period; Based on the time of the artificial rain enhancement operation, obtaining second rainfall data collected by a rainfall station in the operation impact area in the first time period before the artificial rain enhancement operation; Based on the second rainfall data, rainfall extrapolation prediction is performed on the operation impact area using the rainfall prediction model to obtain rainfall prediction data for the operation impact area in the second time period.

5. The method according to claim 1, wherein The obtaining of the actual rainfall data of the comparison area and the actual rainfall data of the operation impact area respectively includes: Obtaining actual rainfall data collected by n rainfall stations in the comparison area; Obtain actual rainfall data collected by m rainfall stations in the operation impact area.

6. The method according to claim 5, characterized in that The step of obtaining a rainfall prediction error of the rainfall prediction model based on the rainfall prediction data of the comparison area and the actual rainfall data of the comparison area comprises: Calculating the difference between the actual rainfall data collected by n rainfall stations in the comparison area and the rainfall forecast data in the comparison area respectively to obtain n differences; The average value of the n differences is used as the rainfall prediction error of the rainfall prediction model.

7. The method according to claim 6, characterized in that The step of obtaining the absolute rainfall increase amount of the artificial rainfall enhancement operation based on the rainfall prediction data of the operation influence area, the actual rainfall data of the operation influence area, and the rainfall prediction error of the rainfall prediction model comprises: Based on the rainfall forecast data of the operation impact area and the actual rainfall data collected by m rainfall stations in the operation impact area, the average absolute rainfall increase of the m rainfall stations in the operation impact area is obtained; Calculating the difference between the average absolute rainfall increase of m rainfall stations in the operation impact area and the rainfall prediction error of the rainfall prediction model to obtain the absolute rainfall increase of a single rainfall station in the operation impact area; The absolute amount of rainfall increase of the artificial rainmaking operation is obtained by calculating the product of the absolute amount of rainfall increase of a single rainfall station in the operation influence area and the area of ​​the operation influence area.

8. A device for evaluating the effect of rain enhancement operations for intelligently predicting natural rainfall, characterized in that: include: The regional division module is used to determine the operation impact area and comparison area based on the scope of artificial rain enhancement operations; The comparison area is not affected by the catalyst of the artificial rainmaking operation; A rainfall prediction module is used to perform rainfall extrapolation prediction on the comparison area and the operation impact area respectively through a rainfall prediction model to obtain rainfall prediction data for the comparison area and rainfall prediction data for the operation impact area; A rainfall acquisition module, used to respectively acquire the actual rainfall data of the comparison area and the actual rainfall data of the operation impact area; an error calculation module, configured to obtain a rainfall prediction error of the rainfall prediction model based on the rainfall prediction data of the comparison area and the actual rainfall data of the comparison area; An effect evaluation module is used to obtain the absolute rainfall increase of the artificial rainmaking operation based on the rainfall prediction data of the operation influence area, the actual rainfall data of the operation influence area and the rainfall prediction error of the rainfall prediction model, and use the absolute rainfall increase of the artificial rainmaking operation as an evaluation indicator of the effect of the rainmaking operation.

9. An electronic device, characterized in that: include: a processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

Images