Artificial precipitation enhancement effect analysis method and system based on dynamic matching algorithm

Abstract

The application discloses a kind of artificial precipitation enhancement effect analysis method and system based on dynamic matching algorithm, method includes: obtaining multi-source data and topographic correction and abnormal value preprocessing are carried out to meteorological reanalysis data;According to the wind field of operation time and the atmospheric stability correction factor determined by Richardson number, fan-shaped target area and multiple candidate control areas are dynamically constructed, and are discretized into subsectors;The optimal control area is determined by matching the similarity of terrain, precipitation statistics and ecological characteristics;The quantile regression model containing wind speed interaction term is adaptively selected by power analysis to predict the natural precipitation without operation, and the enhancement effect is calculated;Based on the error discovery rate control, the multiple comparison correction is carried out on the results of multiple operations, and the evaluation results are output.The dynamic adaptability of target zoning, the matching accuracy of control area and the reliability of statistical inference are significantly improved.

Description

Technical Field

[0001] This invention belongs to the field of weather modification effect evaluation technology, and in particular relates to a method and system for analyzing the effect of artificial precipitation enhancement based on dynamic matching algorithm. Background Technology

[0002] Artificial precipitation enhancement operations aim to increase surface precipitation by seeding clouds with catalysts to alter cloud microphysical processes. Scientific and objective evaluation of the actual effectiveness of artificial precipitation enhancement operations is crucial for optimizing operational plans and improving water resource management capabilities.

[0003] Existing methods for evaluating the effects of artificial precipitation enhancement primarily employ a statistical comparison scheme using fixed target areas and control areas. The fixed-area scheme struggles to adapt to complex and variable meteorological conditions, neglecting the differences in catalyst diffusion patterns under varying atmospheric stability, leading to inaccurate target area delineation. Furthermore, in selecting control areas, existing methods often rely on subjective experience or simple geographical proximity principles, lacking systematic similarity matching of topographic, ecological, and precipitation-climate characteristics, easily introducing systematic biases. In addition, in the statistical analysis phase, existing methods typically use fixed linear regression models, failing to consider the interaction between wind speed and operational effects, and failing to adaptively select model complexity based on data sample size, resulting in insufficient robustness of statistical inference. For evaluation results from multiple operations, there is also a lack of appropriate multiple comparison corrections, leading to a high error rate in significance determination.

[0004] Therefore, there is an urgent need for an analysis method for artificial precipitation enhancement effects that can dynamically adapt to meteorological conditions, achieve precise control area matching, adaptively select statistical models, and effectively control multiple comparison errors. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for analyzing the effects of artificial precipitation enhancement based on dynamic matching algorithm, so as to solve the problems of static target area delineation, subjective control area selection, fixed statistical model selection, and insufficient control of multiple comparison errors in the existing technology.

[0006] In a first aspect, the present invention provides a method for analyzing the enhancement effect of artificial precipitation based on a dynamic matching algorithm, comprising: Acquire multi-source data from artificial precipitation enhancement operations, including operation logs, meteorological reanalysis data, digital elevation model data, and ecological remote sensing data; The meteorological reanalysis data is subjected to terrain correction and outlier preprocessing to obtain the preprocessed target meteorological data; For current artificial precipitation enhancement operations, a fan-shaped target area and multiple fan-shaped candidate control areas are dynamically constructed based on wind field data and atmospheric stability correction factors at the time of operation. The optimal control area is determined by matching from the multiple fan-shaped candidate control areas through a comprehensive similarity index. The radial angle of the fan-shaped target area is jointly determined by the wind speed and atmospheric stability correction factors, and the fan-shaped target area is discretized into multiple sub-sectors to calculate the regional average precipitation. Using the precipitation data of the optimal control area and the fan-shaped target area during the historical training period, a statistical model is adaptively selected through power analysis, and an enhanced quantile regression model with a wind speed interaction term is fitted to predict the natural precipitation of the fan-shaped target area under no-operation conditions. The precipitation enhancement effect of the current artificial precipitation enhancement operation is calculated based on the actual precipitation and the natural precipitation. The statistical significance of the precipitation enhancement effect is tested, and multiple comparisons are made of the test results of multiple artificial precipitation enhancement operations to output the evaluation results of the current artificial precipitation enhancement operation.

[0007] Secondly, the present invention provides a system for analyzing the effects of artificial precipitation enhancement based on a dynamic matching algorithm, comprising: The acquisition module is configured to acquire multi-source data from artificial precipitation enhancement operations, including operation logs, meteorological reanalysis data, digital elevation model data, and ecological remote sensing data. The preprocessing module is configured to perform terrain correction and outlier preprocessing on the meteorological reanalysis data to obtain preprocessed target meteorological data. The determination module is configured to dynamically construct a fan-shaped target area and multiple fan-shaped candidate control areas based on wind field data and atmospheric stability correction factors at the time of the current artificial precipitation enhancement operation. The optimal control area is determined by matching from the multiple fan-shaped candidate control areas through a comprehensive similarity index. The radial angle of the fan-shaped target area is jointly determined by the wind speed and atmospheric stability correction factors, and the fan-shaped target area is discretized into multiple sub-sectors to calculate the regional average precipitation. The prediction module is configured to use precipitation data from the optimal control area and the fan-shaped target area during the historical training period, adaptively select a statistical model through power analysis, fit an enhanced quantile regression model with a wind speed interaction term, predict the natural precipitation in the fan-shaped target area under no-operation conditions, and calculate the precipitation enhancement effect of the current artificial precipitation enhancement operation based on the actual precipitation and the natural precipitation. The output module is configured to perform a statistical significance test on the precipitation enhancement effect, and to perform multiple comparisons of the test results of multiple artificial precipitation enhancement operations, and output the evaluation result of the current artificial precipitation enhancement operation.

[0008] Thirdly, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the artificial precipitation enhancement effect analysis method based on dynamic matching algorithm according to any embodiment of the present invention.

[0009] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the steps of the method for analyzing the artificial precipitation enhancement effect based on a dynamic matching algorithm according to any embodiment of the present invention.

[0010] This application presents a method and system for analyzing the effects of artificial precipitation enhancement based on a dynamic matching algorithm. In terms of target zoning, it extracts atmospheric boundary layer height and Monin-Obukhov length from meteorological reanalysis data to calculate the Richardson number, determine atmospheric stability, and set correction factors. It then dynamically determines the radial angle and radial radius of the fan-shaped target area in conjunction with wind speed, ensuring the target area morphology matches the diffusion patterns under actual meteorological conditions. Regarding control area matching, it extracts indicators from three dimensions: topographic features, precipitation time-series statistical features, and ecological remote sensing features. Through adaptive weighted fusion, it obtains a comprehensive similarity index, quantitatively selecting the optimal control area from the candidate control area set, systematically reducing evaluation bias introduced by differences in natural background. In terms of model construction, it evaluates the statistical power under effective sample size through power analysis, adaptively selects whether to include the interaction term between wind speed and regional indicator variables, and fits a regression model under enhanced quantiles, balancing model flexibility with statistical robustness under small sample conditions. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A flowchart illustrating an analysis method for enhancing artificial precipitation effects based on a dynamic matching algorithm, provided in an embodiment of the present invention; Figure 2 This is a structural block diagram of an artificial precipitation enhancement effect analysis system based on a dynamic matching algorithm, provided in an embodiment of the present invention. Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] Please see Figure 1 The diagram shows a flowchart of an analysis method for artificial precipitation enhancement based on a dynamic matching algorithm, as presented in this application.

[0015] like Figure 1 As shown, the method for analyzing the enhancement effect of artificial precipitation based on the dynamic matching algorithm specifically includes the following steps: Step S101: Obtain multi-source data from artificial precipitation enhancement operations. The multi-source data includes operation logs, meteorological reanalysis data, digital elevation model data, and ecological remote sensing data.

[0016] In this step, the operation log records the time (accurate to the minute), geographic coordinates of the operation site, and operation type for each artificial precipitation enhancement operation. Meteorological reanalysis data consists of gridded historical meteorological field data, including variables such as precipitation, temperature, wind speed, wind direction, atmospheric boundary layer height, and Monin-Obukhov length. Digital elevation model data provides topographic elevation information. Ecological remote sensing data includes normalized vegetation index and land use type proportions obtained from satellite remote sensing inversion.

[0017] Step S102: Perform terrain correction and outlier preprocessing on the meteorological reanalysis data to obtain preprocessed target meteorological data.

[0018] In this step, the digital elevation model data is acquired, a buffer zone with a preset radius is established with the work point as the center, and the elevation raster values ​​within the buffer zone are extracted. For the precipitation and temperature variables in the meteorological reanalysis data, a local elevation gradient regression model is constructed using the digital elevation model data. The meteorological variable values ​​at each grid point are then corrected for elevation to obtain topographically corrected meteorological reanalysis data. For the meteorological reanalysis data after terrain correction, an outlier detection algorithm based on spatial consistency is used to calculate the time series correlation coefficient and deviation between each grid point and its surrounding neighboring grid points. Grid points with correlation coefficients lower than a preset threshold and deviations exceeding three times the local standard deviation are marked as outliers. After removing outliers, the missing positions are filled using spatiotemporal kriging interpolation to obtain continuous and complete preprocessed target meteorological data in both space and time.

[0019] Step S103: For the current artificial precipitation enhancement operation, a fan-shaped target area and multiple fan-shaped candidate control areas are dynamically constructed based on the wind field data and atmospheric stability correction factor at the time of the operation. The optimal control area is determined by matching from the multiple fan-shaped candidate control areas through a comprehensive similarity index. The radial angle of the fan-shaped target area is jointly determined by the wind speed and atmospheric stability correction factor, and the fan-shaped target area is discretized into multiple sub-sectors to calculate the regional average precipitation.

[0020] In this step, the geographical coordinates of the work site are used as the vertex, and the wind direction in the wind field data at the time of the work is used as the reference to determine the direction of the central axis of the fan. Calculate the radial angle of the sector based on the wind speed and atmospheric stability correction factor at the time of operation; The radial radius of the fan-shaped target area is determined by multiplying the boundary layer average wind speed in the meteorological reanalysis data upstream of the operation site with the operation time. Based on the direction of the central axis of the sector, the radial angle of the sector, and the radial radius, a sector-shaped target area with the work point as the vertex is constructed; The constructed fan-shaped target area is discretized into multiple sub-sectors along the radial and angular directions. Each sub-sector covers a predetermined angular span and radial range, which is used to calculate the regional average precipitation within the sub-sector. Centered on the work point and based on the fan-shaped target area, multiple fan-shaped candidate control areas that do not overlap with the fan-shaped target area are constructed upwind and on both sides. The fan-shaped candidate control areas have the same discretized sub-sector structure and area range.

[0021] The atmospheric stability correction factor is determined as follows: The atmospheric boundary layer height and Moning-Obukhov length at the time of operation are extracted from the meteorological reanalysis data, and the Richardson number at the time of operation is calculated. The Richardson number is used as a quantitative indicator of atmospheric stability. When the Richardson number is less than the preset stability threshold, the atmosphere is determined to be in an unstable state and the atmospheric stability correction factor is set to the first preset value. When the Richardson number is greater than or equal to the stability threshold, the atmosphere is determined to be in a stable state and the atmospheric stability correction factor is set to a second preset value, wherein the first preset value is less than the second preset value.

[0022] The first and second preset values ​​are quantitative settings based on the influence of the atmospheric stability correction factor on the radial angle of the sector. Their determination method is as follows: Historical samples of artificial precipitation enhancement operations were collected, and meteorological reanalysis data and measured surface precipitation distribution data were obtained for each operation. For each operation, the optimal sector radial angle under the given meteorological conditions was calculated based on the actual spatial distribution of the precipitation enhancement area. The ratio of the calculated optimal sector radial angle to the reference angle and wind speed coefficient was used as the measured corresponding value of the atmospheric stability correction factor. All historical operation samples were divided into unstable and stable groups based on whether the Richardson number was less than the stability threshold. The median or mean of the measured corresponding values ​​of the atmospheric stability correction factor within each group was calculated and used as the calibration results for the first and second preset values.

[0023] It should be noted that the statistical features of precipitation time series, topographic features and ecological remote sensing features of the fan-shaped target area during the historical training period are extracted. Among them, the statistical features of precipitation time series include the mean and variance of precipitation, the topographic features include the average elevation, slope and aspect, and the ecological remote sensing features include the normalized vegetation index and the proportion of land use types. For each fan-shaped candidate control area, extract the same category and the same dimension of precipitation time series statistical features, topographic features, and ecological remote sensing features; Calculate the terrain feature similarity between each fan-shaped candidate control region and the fan-shaped target region, and use it as the first similarity. Calculate the statistical similarity of precipitation time series features between each fan-shaped candidate control area and the fan-shaped target area, and use it as the second similarity. Calculate the ecological remote sensing feature similarity between each fan-shaped candidate control area and the fan-shaped target area, and use it as the third similarity. An adaptive weighting method is used to sum the first similarity, second similarity, and third similarity to obtain a comprehensive similarity index for each sector candidate control region. The sector-shaped candidate control area with the highest comprehensive similarity index value is determined as the optimal control area for the current artificial precipitation enhancement operation.

[0024] In one specific embodiment, the process of constructing the fan-shaped target region specifically includes: S1031: Using the geographical coordinates of the work site as the vertex and the wind direction in the wind field data at the time of the work as the reference, the direction of the central axis of the fan-shaped area is determined so that the target area of ​​the fan-shaped area expands along the downwind direction.

[0025] S1032: Calculate the sector radial angle based on the wind speed and atmospheric stability correction factor at the time of operation. The atmospheric stability correction factor is determined as follows: extract the atmospheric boundary layer height and Monin-Obukhov length at the time of operation from meteorological reanalysis data, and calculate the Richardson number at the time of operation; when the Richardson number is less than a preset stability threshold (e.g., 0), the atmosphere is determined to be unstable, and the atmospheric stability correction factor is set to a first preset value (e.g., 0.8); when the Richardson number is greater than or equal to the stability threshold, the atmosphere is determined to be stable, and the atmospheric stability correction factor is set to a second preset value (e.g., 1.2). The first preset value is less than the second preset value, so that under the same wind speed conditions, the sector radial angle is smaller when the atmosphere is unstable and larger when the atmosphere is stable. The formula for calculating the sector radial angle can be expressed as: Radial angle = Reference angle × Wind speed coefficient × Atmospheric stability correction factor, where the reference angle is a preset constant, and the wind speed coefficient decreases as the wind speed increases.

[0026] S1033: Determine the radial radius of the fan-shaped target area based on the product of the boundary layer average wind speed in the meteorological reanalysis data upstream of the operation site and the operation duration. The operation duration refers to the time window from catalyst application to the expected precipitation enhancement effect, typically 3 to 6 hours.

[0027] S1034: Based on the direction of the central axis of the sector, the radial angle of the sector, and the radial radius, construct a sector-shaped target area with the work point as the vertex.

[0028] S1035: Discretize the constructed fan-shaped target area into multiple sub-sectors along the radial and angular directions. For example, divide it into 3 equal segments along the radial direction and 5 equal segments along the angular direction to form 15 sub-sectors. Each sub-sector covers a predetermined angular span and radial range for subsequent calculation of the regional average precipitation within the sub-sector.

[0029] After constructing the sector-shaped target region, multiple sector-shaped candidate control regions are further constructed: S1036: Centered on the operation point and based on the fan-shaped target area, construct multiple fan-shaped candidate control areas that do not overlap with the fan-shaped target area in the upwind direction and on both sides. All fan-shaped candidate control areas have the same discretized sub-sector structure and area range as the fan-shaped target area, but the azimuth angle is rotated and covered by a preset step size (e.g., 10 degrees) to form a set containing dozens of candidate control areas.

[0030] Then, the optimal control region is determined by matching from the candidate control region set using a comprehensive similarity index, specifically including: S1037: Extract the statistical characteristics of precipitation time series, topographic features, and ecological remote sensing features of the fan-shaped target area during the historical training period (e.g., the same period in the 30 days prior to the operation). The statistical characteristics of precipitation time series include the mean and variance of precipitation; the topographic features include the mean elevation, slope, and aspect; and the ecological remote sensing features include the normalized difference vegetation index and the proportion of land use types.

[0031] S1038: For each fan-shaped candidate control area, extract the same category and the same dimension of precipitation time series statistical features, topographic features, and ecological remote sensing features.

[0032] S1039: Calculate the topographic feature similarity between each fan-shaped candidate control area and the fan-shaped target area as the first similarity; calculate the statistical feature similarity of precipitation time series as the second similarity; calculate the ecological remote sensing feature similarity as the third similarity. Similarity calculation can employ methods such as Euclidean distance normalization or cosine similarity.

[0033] S10310: Adaptive weights are used to weight and sum the first, second, and third similarities to obtain a comprehensive similarity index for each sector-shaped candidate control area. The adaptive weights can be obtained by collecting historical artificial precipitation enhancement operation samples and using principal component analysis to calculate the variance contribution rate of each terrain feature, precipitation time series statistical feature, and ecological remote sensing feature. The normalized variance contribution rate is then used as the weighting coefficient for each similarity.

[0034] S10311: The fan-shaped candidate control area with the highest comprehensive similarity index value is determined as the optimal control area for the current artificial precipitation enhancement operation.

[0035] Step S104: Using the precipitation data of the optimal control area and the fan-shaped target area during the historical training period, an adaptive statistical model is selected through power analysis, and an enhanced quantile regression model with wind speed interaction term is fitted to predict the natural precipitation of the fan-shaped target area under no-operation conditions. The precipitation enhancement effect of the current artificial precipitation enhancement operation is calculated based on the actual precipitation and the natural precipitation.

[0036] In this step, the effective sample size during the historical training period is counted, the minimum detectable enhancement effect value is set, and the statistical power is calculated under the effective sample size and the minimum detectable enhancement effect value. The statistical power is compared with the preset power threshold. Specifically, when the statistical power is greater than or equal to the preset power threshold, the full quantile regression model structure is selected, and the explanatory variables of the model include precipitation in the optimal control area, regional indicator variable, wind speed, and the interaction term between regional indicator variable and wind speed; when the statistical power is less than the preset power threshold, the simplified quantile regression model structure is selected, and the explanatory variables of the model only include precipitation in the optimal control area and regional indicator variable. Using daily precipitation data from the optimal control area and sector target area during the historical training period, a historical sample dataset is constructed with the regional daily precipitation as the response variable and the corresponding variables of the selected model structure as the explanatory variables. Under the preset enhanced quantile, quantile regression is fitted to the historical sample dataset to obtain the estimated values ​​of each regression coefficient in the enhanced quantile regression model; At the current operation time, the actual precipitation and actual wind speed of the optimal control area are obtained. The actual precipitation of the optimal control area is used as the precipitation input of the optimal control area. The regional indicator variable is set to the category value corresponding to the target area, and the wind speed is set to the actual wind speed at the current operation time. Substitute these values ​​into the fitted enhanced quantile regression model to calculate the predicted value of natural precipitation in the fan-shaped target area under no-operation conditions.

[0037] Furthermore, after the current operation, the precipitation observation values ​​of all sub-sectors within the fan-shaped target area are obtained. Using the proportion of the area of ​​each sub-sector to the total area of ​​the fan-shaped target area as the weight, the precipitation observation values ​​of all sub-sectors are averaged by area to obtain the actual precipitation of the fan-shaped target area. The absolute enhanced precipitation is obtained by subtracting the predicted natural precipitation under no-operation conditions from the actual precipitation in the fan-shaped target area. Divide the absolute increase in precipitation by the predicted natural precipitation under no-operation conditions, and multiply by 100% to obtain the relative increase percentage. The absolute increase in precipitation and the relative percentage increase are used together as the quantitative result of the precipitation enhancement effect of the current artificial precipitation enhancement operation.

[0038] In one specific embodiment, the effective sample size (i.e., the number of days without missing precipitation data) during the historical training period is counted, a minimum detectable enhancement effect value (e.g., 1 mm / day) is set, and the statistical power under this effective sample size and minimum detectable enhancement effect value is calculated. Compare the statistical power with the preset power threshold (e.g., 0.8): When the statistical power is greater than or equal to 0.8, select the full quantile regression model structure, and the model explanatory variables include precipitation in the optimal control area, regional indicator variable, wind speed, and the interaction term between the regional indicator variable and wind speed; when the statistical power is less than 0.8, select the simplified quantile regression model structure, and the model explanatory variables only include precipitation in the optimal control area and regional indicator variable. The regional indicator variable is a binary variable, with a value of 1 indicating that the sample belongs to the sector target area and a value of 0 indicating that it belongs to the optimal control area. Using daily precipitation data from the optimal control area and sector target area during the historical training period, a historical sample dataset is constructed with the regional daily precipitation as the response variable and the corresponding variables of the selected model structure as the explanatory variables. Under a preset enhanced quantile (e.g., 0.9 quantile, corresponding to the case of enhanced extreme precipitation), quantile regression is fitted to the historical sample dataset to obtain the estimated values ​​of each regression coefficient in the enhanced quantile regression model; At the current operation time, the actual precipitation and actual wind speed of the optimal control area are obtained. The actual precipitation of the optimal control area is used as the precipitation input of the optimal control area. The regional indicator variable is set to the category value corresponding to the target area (i.e., 1). The wind speed is set to the actual wind speed at the current operation time. Substitute it into the fitted enhanced quantile regression model to calculate the predicted value of natural precipitation of the fan-shaped target area under no operation conditions. After obtaining the precipitation observation values ​​of all sub-sectors within the fan-shaped target area, the actual precipitation of the fan-shaped target area is obtained by performing an area-weighted average with the proportion of each sub-sector area to the total area of ​​the fan-shaped target area as the weight. The absolute increase in precipitation is obtained by subtracting the predicted natural precipitation under no-operation conditions from the actual precipitation in the fan-shaped target area; the relative increase in precipitation is obtained by dividing the absolute increase in precipitation by the predicted natural precipitation and then multiplying by 100%. The absolute increase in precipitation and the relative percentage increase are used together as the quantitative result of the precipitation enhancement effect of the current artificial precipitation enhancement operation.

[0039] Step S105: Perform a statistical significance test on the precipitation enhancement effect, and conduct multiple comparisons of the test results of multiple artificial precipitation enhancement operations to output the evaluation result of the current artificial precipitation enhancement operation.

[0040] In this step, the daily precipitation time series data of the optimal control area and the fan-shaped target area during the historical training period are obtained. The block length parameter is set, and the moving block extraction method with replacement is used to extract B bootstrap sample sets from the daily precipitation time series data. The length of each bootstrap sample set is consistent with the length of the historical training period. For each bootstrap sample set, the bootstrap sample set data is used to replace the original historical training period data, the enhanced quantile regression model is refitted, and B virtual enhanced effect values ​​are calculated according to the same calculation process as the actual precipitation enhancement effect, forming the bootstrap null distribution of the virtual enhanced effect values. The actual precipitation enhancement effect value of the current artificial precipitation enhancement operation is compared with the self-established null distribution. The proportion of samples in the self-established null distribution that is greater than or equal to the actual precipitation enhancement effect value is calculated as the p-value of the one-sided test. Summarize the p-values ​​corresponding to each artificial precipitation enhancement operation to obtain a set of M p-values. Sort the M p-values ​​in ascending order to obtain the sorted p-value sequence. Find the largest p-value that meets the preset screening conditions from the sorted p-value sequence, and determine that the artificial precipitation enhancement operations corresponding to the largest p-value and all p-values ​​sorted before the largest p-value are statistically significant, while the artificial precipitation enhancement operations corresponding to all p-values ​​sorted after the largest p-value are not statistically significant, thus obtaining a set of significance determination results for M artificial precipitation enhancement operations. Extract the judgment result corresponding to the current artificial precipitation enhancement operation from the set of significance judgment results, calculate the confidence interval of the precipitation enhancement effect quantification result of the current artificial precipitation enhancement operation using the percentile of the self-generated zero distribution, combine the precipitation enhancement effect quantification result of the current artificial precipitation enhancement operation, the confidence interval and the extracted significance judgment result to output the evaluation result of the current artificial precipitation enhancement operation.

[0041] In one specific embodiment, the daily precipitation time series data of the optimal control area and the fan-shaped target area during the historical training period are obtained, the block length parameter is set (for example, 5 days are taken as a block), and a moving block extraction method with replacement is used to extract B bootstrap sample sets (B is usually 1000) from the daily precipitation time series data. The length of each bootstrap sample set is consistent with the length of the historical training period. For each bootstrap sample set, the bootstrap sample set data is used to replace the original historical training period data, the enhanced quantile regression model is refitted, and B virtual enhancement effect values ​​are calculated. These B virtual enhancement effect values ​​constitute the bootstrap null distribution of precipitation enhancement effect. The actual precipitation enhancement effect value (absolute or relative value) of the current artificial precipitation enhancement operation is compared with the self-established null distribution. The proportion of samples in the self-established null distribution that is greater than or equal to the actual precipitation enhancement effect value is calculated as the one-sided test p-value for a single operation. Summarize the p-values ​​corresponding to each artificial precipitation enhancement operation to obtain a set of M p-values. Sort the M p-values ​​in ascending order to obtain the sorted p-value sequence. The preset screening condition is set as the error detection rate control condition. Specifically, the global error detection rate control level α (e.g., 0.1) is set. For the k-th p-value after sorting, its corresponding adjusted significance threshold is calculated using the formula: (k / M)×α. Then, p-values ​​are checked sequentially from k to M to see if they are less than or equal to the corresponding adjusted significance threshold. The largest index k that meets the condition is found. The artificial precipitation enhancement operations corresponding to the largest index k and all p-values ​​sorted before the largest index k are determined to be statistically significant, while the artificial precipitation enhancement operations corresponding to all p-values ​​sorted after the largest index k are not statistically significant. This yields the significance determination result set for M operations. Extract the judgment results corresponding to the current artificial precipitation enhancement operation from the set of significance judgment results, and at the same time, calculate the confidence interval of the current operation's enhancement effect size using the specified percentiles (e.g., 2.5% and 97.5% percentiles) of the self-generated zero distribution; The quantitative results, confidence intervals, and significance determination results of the current artificial precipitation enhancement operations are combined to output the evaluation results of the current artificial precipitation enhancement operations.

[0042] In summary, the method of this application acquires multi-source data and performs topographic correction and outlier preprocessing on meteorological reanalysis data; dynamically constructs a sector-shaped target area and multiple candidate control areas based on the wind field at the operation time and the atmospheric stability correction factor determined by the Richardson number, and discretizes them into sub-sectors; determines the optimal control area by comprehensively considering topography, precipitation statistics, and ecological feature similarity matching; uses power analysis to adaptively select a quantile regression model containing a wind speed interaction term to predict unoperated natural precipitation and calculate the enhancement effect; and outputs the evaluation result after multiple comparisons and corrections of the results of multiple operations based on error detection rate control; significantly improving the dynamic adaptability of target zoning, the accuracy of control area matching, and the reliability of statistical inference.

[0043] Please see Figure 2 The diagram shows a structural block diagram of an artificial precipitation enhancement effect analysis system based on a dynamic matching algorithm according to this application.

[0044] like Figure 2 As shown, the artificial precipitation enhancement effect analysis system 200 includes an acquisition module 210, a preprocessing module 220, a determination module 230, a prediction module 240, and an output module 250.

[0045] The acquisition module 210 is configured to acquire multi-source data from artificial precipitation enhancement operations, including operation logs, meteorological reanalysis data, digital elevation model data, and ecological remote sensing data. The preprocessing module 220 is configured to perform terrain correction and outlier preprocessing on the meteorological reanalysis data to obtain preprocessed target meteorological data. The determination module 230 is configured to dynamically construct a fan-shaped target area and multiple fan-shaped candidate control areas based on wind field data and atmospheric stability correction factors at the time of the current artificial precipitation enhancement operation, and determine the optimal control area by matching from the multiple fan-shaped candidate control areas using a comprehensive similarity index. The radial angle of the fan-shaped target area is determined by wind speed and atmospheric stability correction factors. Positive factors are jointly determined, and the fan-shaped target area is discretized into multiple sub-sectors to calculate the regional average precipitation; the prediction module 240 is configured to use the precipitation data of the optimal control area and the fan-shaped target area during the historical training period, adaptively select a statistical model through power analysis, fit an enhanced quantile regression model with a wind speed interaction term, predict the natural precipitation of the fan-shaped target area under no-operation conditions, and calculate the precipitation enhancement effect of the current artificial precipitation enhancement operation based on the actual precipitation and the natural precipitation; the output module 250 is configured to perform a statistical significance test on the precipitation enhancement effect, and perform multiple comparisons of the test results of multiple artificial precipitation enhancement operations, and output the evaluation result of the current artificial precipitation enhancement operation.

[0046] It should be understood that Figure 2 The modules and references described in the document Figure 1 The steps described in the text correspond to those in the method described above. Therefore, the operations, features, and corresponding technical effects described above also apply to the method described in the text. Figure 2 The various modules in the document will not be described in detail here.

[0047] In other embodiments, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the artificial precipitation enhancement effect analysis method based on dynamic matching algorithm in any of the above method embodiments. In one embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions, which are configured as follows: Acquire multi-source data from artificial precipitation enhancement operations, including operation logs, meteorological reanalysis data, digital elevation model data, and ecological remote sensing data; The meteorological reanalysis data is subjected to terrain correction and outlier preprocessing to obtain the preprocessed target meteorological data; For current artificial precipitation enhancement operations, a fan-shaped target area and multiple fan-shaped candidate control areas are dynamically constructed based on wind field data and atmospheric stability correction factors at the time of operation. The optimal control area is determined by matching from the multiple fan-shaped candidate control areas through a comprehensive similarity index. The radial angle of the fan-shaped target area is jointly determined by the wind speed and atmospheric stability correction factors, and the fan-shaped target area is discretized into multiple sub-sectors to calculate the regional average precipitation. Using the precipitation data of the optimal control area and the fan-shaped target area during the historical training period, a statistical model is adaptively selected through power analysis, and an enhanced quantile regression model with a wind speed interaction term is fitted to predict the natural precipitation of the fan-shaped target area under no-operation conditions. The precipitation enhancement effect of the current artificial precipitation enhancement operation is calculated based on the actual precipitation and the natural precipitation. The statistical significance of the precipitation enhancement effect is tested, and multiple comparisons are made of the test results of multiple artificial precipitation enhancement operations to output the evaluation results of the current artificial precipitation enhancement operation.

[0048] Computer-readable storage media may include a stored program area and a stored data area, wherein the stored program area may store an operating system and an application program required for at least one function; the stored data area may store data created based on the use of the artificial precipitation enhancement effect analysis system based on a dynamic matching algorithm. Furthermore, the computer-readable storage medium may include high-speed random access memory, and may also include memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the computer-readable storage medium may optionally include memory remotely configured relative to a processor, which can be connected to the artificial precipitation enhancement effect analysis system based on a dynamic matching algorithm via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0049] Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present invention, such as... Figure 3 As shown, the device includes a processor 310 and a memory 320. The electronic device may also include an input device 330 and an output device 340. The processor 310, memory 320, input device 330, and output device 340 can be connected via a bus or other means. Figure 3Taking a bus connection as an example, the memory 320 is the computer-readable storage medium described above. The processor 310 executes various server functions and data processing by running non-volatile software programs, instructions, and modules stored in the memory 320, thereby implementing the artificial precipitation enhancement effect analysis method based on the dynamic matching algorithm described in the above embodiment. The input device 330 can receive input digital or character information and generate key signal inputs related to user settings and function control of the artificial precipitation enhancement effect analysis system based on the dynamic matching algorithm. The output device 340 may include a display screen or other display device.

[0050] The aforementioned electronic device can execute the method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in the embodiments of the present invention.

[0051] In one implementation, the above-described electronic device is applied in an artificial precipitation enhancement effect analysis system based on a dynamic matching algorithm, for a client, and includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Acquire multi-source data from artificial precipitation enhancement operations, including operation logs, meteorological reanalysis data, digital elevation model data, and ecological remote sensing data; The meteorological reanalysis data is subjected to terrain correction and outlier preprocessing to obtain the preprocessed target meteorological data; For current artificial precipitation enhancement operations, a fan-shaped target area and multiple fan-shaped candidate control areas are dynamically constructed based on wind field data and atmospheric stability correction factors at the time of operation. The optimal control area is determined by matching from the multiple fan-shaped candidate control areas through a comprehensive similarity index. The radial angle of the fan-shaped target area is jointly determined by the wind speed and atmospheric stability correction factors, and the fan-shaped target area is discretized into multiple sub-sectors to calculate the regional average precipitation. Using the precipitation data of the optimal control area and the fan-shaped target area during the historical training period, a statistical model is adaptively selected through power analysis, and an enhanced quantile regression model with a wind speed interaction term is fitted to predict the natural precipitation of the fan-shaped target area under no-operation conditions. The precipitation enhancement effect of the current artificial precipitation enhancement operation is calculated based on the actual precipitation and the natural precipitation. The statistical significance of the precipitation enhancement effect is tested, and multiple comparisons are made of the test results of multiple artificial precipitation enhancement operations to output the evaluation results of the current artificial precipitation enhancement operation.

[0052] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0053] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for analyzing the enhancement effect of artificial precipitation based on a dynamic matching algorithm, characterized in that, include: Acquire multi-source data from artificial precipitation enhancement operations, including operation logs, meteorological reanalysis data, digital elevation model data, and ecological remote sensing data; The meteorological reanalysis data is subjected to terrain correction and outlier preprocessing to obtain the preprocessed target meteorological data; For current artificial precipitation enhancement operations, a fan-shaped target area and multiple fan-shaped candidate control areas are dynamically constructed based on wind field data and atmospheric stability correction factors at the time of operation. The optimal control area is determined by matching from the multiple fan-shaped candidate control areas through a comprehensive similarity index. The radial angle of the fan-shaped target area is jointly determined by wind speed and atmospheric stability, and the fan-shaped target area is discretized into multiple sub-sectors to calculate the regional average precipitation. Using precipitation data from the optimal control area and the fan-shaped target area during the historical training period, an enhanced quantile regression model with a wind speed interaction term is fitted through adaptive selection of a statistical model via power analysis to predict the natural precipitation in the fan-shaped target area under no-operation conditions. The precipitation enhancement effect of the current artificial precipitation enhancement operation is then calculated based on the actual precipitation and the natural precipitation, specifically including: The effective sample size during the historical training period is determined, the minimum detectable enhancement effect value is set, and the statistical power is calculated under the effective sample size and the minimum detectable enhancement effect value. The statistical power is compared with the preset power threshold. Specifically, when the statistical power is greater than or equal to the preset power threshold, the full quantile regression model structure is selected, and the explanatory variables of the model include precipitation in the optimal control area, regional indicator variable, wind speed, and the interaction term between regional indicator variable and wind speed; when the statistical power is less than the preset power threshold, the simplified quantile regression model structure is selected, and the explanatory variables of the model only include precipitation in the optimal control area and regional indicator variable. Using daily precipitation data from the optimal control area and sector target area during the historical training period, a historical sample dataset is constructed with the regional daily precipitation as the response variable and the corresponding variables of the selected model structure as the explanatory variables. Under the preset enhanced quantile, quantile regression is fitted to the historical sample dataset to obtain the estimated values ​​of each regression coefficient in the enhanced quantile regression model; At the current operation time, the actual precipitation and actual wind speed of the optimal control area are obtained. The actual precipitation of the optimal control area is used as the precipitation input of the optimal control area. The regional indicator variable is set to the category value corresponding to the target area, and the wind speed is set to the actual wind speed at the current operation time. Substitute them into the fitted enhanced quantile regression model to calculate the predicted value of natural precipitation of the fan-shaped target area under no operation conditions. The statistical significance of the precipitation enhancement effect is tested, and multiple comparisons are made of the test results of multiple artificial precipitation enhancement operations to output the evaluation results of the current artificial precipitation enhancement operation.

2. The method for analyzing the enhancement effect of artificial precipitation based on a dynamic matching algorithm according to claim 1, characterized in that, The process of performing terrain correction and outlier preprocessing on the meteorological reanalysis data to obtain preprocessed target meteorological data includes: The digital elevation model data is acquired, and a buffer zone with a preset radius is established with the work point as the center. The elevation raster values ​​within the buffer zone are then extracted. For the precipitation and temperature variables in the meteorological reanalysis data, a local elevation gradient regression model is constructed using the digital elevation model data. The meteorological variable values ​​at each grid point are then corrected for elevation to obtain topographically corrected meteorological reanalysis data. For the meteorological reanalysis data after terrain correction, an outlier detection algorithm based on spatial consistency is used to calculate the time series correlation coefficient and deviation between each grid point and its surrounding neighboring grid points. Grid points with correlation coefficients lower than a preset threshold and deviations exceeding three times the local standard deviation are marked as outliers. After removing outliers, the missing positions are filled using spatiotemporal kriging interpolation to obtain continuous and complete preprocessed target meteorological data in both space and time.

3. The method for analyzing the effect of artificial precipitation enhancement based on dynamic matching algorithm according to claim 1, characterized in that, The aforementioned method for dynamically constructing a fan-shaped target area and multiple fan-shaped candidate control areas based on wind field data and atmospheric stability correction factors at the time of the operation includes: The direction of the central axis of the sector is determined by taking the geographical coordinates of the work site as the vertex and the wind direction in the wind field data at the time of the work as the reference. Calculate the radial angle of the sector based on the wind speed and atmospheric stability correction factor at the time of operation; The radial radius of the fan-shaped target area is determined by multiplying the boundary layer average wind speed in the meteorological reanalysis data upstream of the operation site with the operation time. Based on the direction of the central axis of the sector, the radial angle of the sector, and the radial radius, a sector-shaped target area with the work point as the vertex is constructed; The constructed fan-shaped target area is discretized into multiple sub-sectors along the radial and angular directions. Each sub-sector covers a predetermined angular span and radial range, which is used to calculate the regional average precipitation within the sub-sector. Centered on the work point and based on the fan-shaped target area, multiple fan-shaped candidate control areas that do not overlap with the fan-shaped target area are constructed upwind and on both sides. The fan-shaped candidate control areas have the same discretized sub-sector structure and area range.

4. The method for analyzing the effect of artificial precipitation enhancement based on dynamic matching algorithm according to claim 3, characterized in that, in, The atmospheric stability correction factor is determined as follows: Extract the atmospheric boundary layer height and Moning-Obukhov length at the time of operation from the meteorological reanalysis data, and calculate the Richardson number at the time of operation; When the Richardson number is less than the preset stability threshold, the atmosphere is determined to be in an unstable state and the atmospheric stability correction factor is set to the first preset value. When the Richardson number is greater than or equal to the stability threshold, the atmosphere is determined to be in a stable state and the atmospheric stability correction factor is set to a second preset value, wherein the first preset value is less than the second preset value.

5. The method for analyzing the enhancement effect of artificial precipitation based on a dynamic matching algorithm according to claim 3, characterized in that, The step of determining the optimal control region by matching from the multiple sector candidate control regions using a comprehensive similarity index includes: Extract the precipitation time series statistical features, topographic features and ecological remote sensing features of the fan-shaped target area during the historical training period. The precipitation time series statistical features include the mean and variance of precipitation, the topographic features include the average elevation, slope and aspect, and the ecological remote sensing features include the normalized vegetation index and the proportion of land use types. For each fan-shaped candidate control area, extract the same category and the same dimension of precipitation time series statistical features, topographic features, and ecological remote sensing features; Calculate the terrain feature similarity between each fan-shaped candidate control region and the fan-shaped target region, and use it as the first similarity. Calculate the statistical similarity of precipitation time series features between each fan-shaped candidate control area and the fan-shaped target area, and use it as the second similarity. Calculate the ecological remote sensing feature similarity between each fan-shaped candidate control area and the fan-shaped target area, and use it as the third similarity. An adaptive weighting method is used to sum the first similarity, second similarity, and third similarity to obtain a comprehensive similarity index for each sector candidate control region. The sector-shaped candidate control area with the highest comprehensive similarity index value is determined as the optimal control area for the current artificial precipitation enhancement operation.

6. The method for analyzing the effect of artificial precipitation enhancement based on dynamic matching algorithm according to claim 1, characterized in that, The calculation of the precipitation enhancement effect of the current artificial precipitation enhancement operation based on the actual precipitation and the natural precipitation includes: After obtaining the precipitation observation values ​​of all sub-sectors within the fan-shaped target area, the actual precipitation of the fan-shaped target area is obtained by taking the area weight of the precipitation observation values ​​of all sub-sectors as the proportion of the area of ​​each sub-sector to the total area of ​​the fan-shaped target area. The absolute enhanced precipitation is obtained by subtracting the predicted natural precipitation under no-operation conditions from the actual precipitation in the fan-shaped target area. Divide the absolute increase in precipitation by the predicted natural precipitation under no-operation conditions, and multiply by 100% to obtain the relative increase percentage. The absolute increase in precipitation and the relative percentage increase are used together as the quantitative results of the precipitation enhancement effect of the current artificial precipitation enhancement operation.

7. The method for analyzing the effect of artificial precipitation enhancement based on dynamic matching algorithm according to claim 1, characterized in that, The statistical significance test of the precipitation enhancement effect is performed, and multiple comparison corrections are applied to the test results of multiple artificial precipitation enhancement operations. The output of the current evaluation result of the artificial precipitation enhancement operation includes: Obtain daily precipitation time series data of the optimal control area and fan-shaped target area during the historical training period, set the block length parameter, and use the moving block extraction method with replacement to extract B bootstrap sample sets from the daily precipitation time series data. The length of each bootstrap sample set is consistent with the length of the historical training period. For each bootstrap sample set, the bootstrap sample set data is used to replace the original historical training period data, the enhanced quantile regression model is refitted, and B virtual enhanced effect values ​​are calculated according to the same calculation process as the actual precipitation enhancement effect, forming the bootstrap null distribution of the virtual enhanced effect values. The actual precipitation enhancement effect value of the current artificial precipitation enhancement operation is compared with the self-established null distribution. The proportion of samples in the self-established null distribution that is greater than or equal to the actual precipitation enhancement effect value is calculated as the p-value of the one-sided test. Summarize the p-values ​​corresponding to each artificial precipitation enhancement operation to obtain a set containing M p-values. Sort the M p-values ​​in ascending order to obtain the sorted p-value sequence. Find the largest p-value that meets the preset screening conditions from the sorted p-value sequence, and determine that the artificial precipitation enhancement operations corresponding to the largest p-value and all p-values ​​sorted before the largest p-value are statistically significant, while the artificial precipitation enhancement operations corresponding to all p-values ​​sorted after the largest p-value are not statistically significant, thus obtaining a set of significance determination results for M artificial precipitation enhancement operations. Extract the judgment result corresponding to the current artificial precipitation enhancement operation from the set of significance judgment results, calculate the confidence interval of the enhancement effect of the current artificial precipitation enhancement operation using the percentile of the self-generated zero distribution, combine the estimated value of the enhancement effect of the current artificial precipitation enhancement operation, the confidence interval and the extracted significance judgment result, and output the evaluation result of the current artificial precipitation enhancement operation.

8. A system for analyzing the effects of artificial precipitation enhancement based on a dynamic matching algorithm, characterized in that, include: The acquisition module is configured to acquire multi-source data from artificial precipitation enhancement operations, including operation logs, meteorological reanalysis data, digital elevation model data, and ecological remote sensing data. The preprocessing module is configured to perform terrain correction and outlier preprocessing on the meteorological reanalysis data to obtain preprocessed target meteorological data. The determination module is configured to dynamically construct a fan-shaped target area and multiple fan-shaped candidate control areas based on wind field data and atmospheric stability correction factors at the time of the current artificial precipitation enhancement operation. The optimal control area is determined by matching from the multiple fan-shaped candidate control areas through a comprehensive similarity index. The radial angle of the fan-shaped target area is jointly determined by the wind speed and atmospheric stability, and the fan-shaped target area is discretized into multiple sub-sectors to calculate the regional average precipitation. The prediction module is configured to utilize precipitation data from the optimal control area and the fan-shaped target area during the historical training period, adaptively select a statistical model through power analysis, fit an enhanced quantile regression model with a wind speed interaction term, predict the natural precipitation in the fan-shaped target area under no-operation conditions, and calculate the precipitation enhancement effect of the current artificial precipitation enhancement operation based on the actual precipitation and the natural precipitation. Specifically, this includes: The effective sample size during the historical training period is determined, the minimum detectable enhancement effect value is set, and the statistical power is calculated under the effective sample size and the minimum detectable enhancement effect value. The statistical power is compared with the preset power threshold. Specifically, when the statistical power is greater than or equal to the preset power threshold, the full quantile regression model structure is selected, and the explanatory variables of the model include precipitation in the optimal control area, regional indicator variable, wind speed, and the interaction term between regional indicator variable and wind speed; when the statistical power is less than the preset power threshold, the simplified quantile regression model structure is selected, and the explanatory variables of the model only include precipitation in the optimal control area and regional indicator variable. Using daily precipitation data from the optimal control area and sector target area during the historical training period, a historical sample dataset is constructed with the regional daily precipitation as the response variable and the corresponding variables of the selected model structure as the explanatory variables. Under the preset enhanced quantile, quantile regression is fitted to the historical sample dataset to obtain the estimated values ​​of each regression coefficient in the enhanced quantile regression model; At the current operation time, the actual precipitation and actual wind speed of the optimal control area are obtained. The actual precipitation of the optimal control area is used as the precipitation input of the optimal control area. The regional indicator variable is set to the category value corresponding to the target area, and the wind speed is set to the actual wind speed at the current operation time. Substitute them into the fitted enhanced quantile regression model to calculate the predicted value of natural precipitation of the fan-shaped target area under no operation conditions. The output module is configured to perform a statistical significance test on the precipitation enhancement effect, and to perform multiple comparisons of the test results of multiple artificial precipitation enhancement operations, and output the evaluation result of the current artificial precipitation enhancement operation.

Images