Method for quantitatively evaluating comprehensive risk of meteorological disasters based on main grain crops

Abstract

This invention discloses a quantitative assessment method for comprehensive meteorological disaster risk based on major grain crops, belonging to the field of comprehensive meteorological disaster risk assessment technology. This method, based on major grain crops, achieves several key improvements: First, by integrating 18 multi-dimensional indicators, it overcomes the limitations of single meteorological disaster assessment, enabling the quantification of comprehensive risks from multiple disasters such as drought and floods, thus alleviating the difficulty of traditional methods in addressing cascading disasters. Second, it is adaptable to major grain crops such as wheat, rice, and corn, accurately capturing risk differences among different crops by combining the trends, fluctuations, and extreme values ​​of crop meteorological yields. Third, it employs standardized processing and weighted calculation to eliminate differences in indicator units, improving the comparability and objectivity of the assessment results. Fourth, it innovates a dual-level classification method to ensure the scientific rigor and stability of risk level division. Fifth, it can directly serve regional agricultural meteorological disaster prevention decision-making through quantitative results and visual presentation.

Description

Technical Field

[0001] This invention relates to the field of comprehensive meteorological disaster risk assessment technology, specifically to a quantitative assessment method for comprehensive meteorological disaster risk based on major food crops. Background Technology

[0002] Existing quantitative assessment methods for comprehensive meteorological disaster risks based on major food crops still have the following drawbacks in practical application:

[0003] As global climate change intensifies, the frequency of extreme weather events is increasing year by year, causing increasingly severe losses to agricultural production. A report by the Food and Agriculture Organization of the United Nations shows that between 1991 and 2021, global losses of crops and livestock due to meteorological disasters totaled US$3.8 trillion, accounting for more than 5% of the annual global agricultural output. This figure is significantly underestimated due to inadequate monitoring systems. In my country, the production security of major food crops such as winter wheat, rice, and corn is directly related to the national food strategy. The "Full-Chain Agricultural Meteorological Disaster Risk Early Warning Project for Major Food Crops," launched in 2025 alone, needs to address more than 10 types of disaster threats, including drought and high-temperature heat damage, highlighting the urgency of accurate risk assessment.

[0004] Currently, agricultural meteorological disaster risk assessment technology has developed in a multidisciplinary manner. At the data acquisition level, satellite remote sensing, UAV synthetic aperture radar, and ground-based intelligent monitoring equipment constitute a multi-scale observation network. Among them, UAV technology, through its fully polarized operating mode, achieves meter-level resolution for pre- and post-disaster comparative monitoring, effectively compensating for the long revisit cycle of traditional satellite remote sensing. In terms of analytical methods, machine learning algorithms are widely used to establish rainfall models and crop response mechanisms. For example, prediction models trained based on historical meteorological data can improve the accuracy of waterlogging disaster assessment. In terms of operational applications, a dynamic monitoring system for soybean drought and low-temperature damage has been established in Northeast China, achieving daily identification of disaster processes through daily-scale data rolling calculations.

[0005] However, existing technological systems still have significant limitations: First, their comprehensive assessment capabilities for multiple disasters are insufficient. Most methods focus on single disaster types, such as waterlogging warnings for rice-wheat rotation farmland or specific meteorological factor analysis during the tobacco growing season, making it difficult to address the cascading risks of concurrent drought and high temperatures, and the proliferation of pests and diseases after floods in actual production. Second, there is a contradiction between crop specificity and assessment universality. Existing technologies are mostly targeted at specific crops or regions, lacking a unified assessment framework that can simultaneously cover staple crops such as wheat, rice, and corn, and failing to fully reflect the differences in vulnerability of different crops during their growing seasons. Third, the accuracy of quantitative assessment is limited. Traditional methods often rely on empirical models or single data sources. For example, drought indices based on monthly meteorological data are difficult to reflect the instantaneous stress during the critical growing season of crops, and the fusion mechanism of multi-source data (meteorological observations, crop growth, and soil characteristics) has not yet been standardized. This results in significant defects in the global agricultural disaster data system, with different loss statistical standards in different regions, making horizontal comparison and vertical tracing of multiple disaster risks extremely difficult. Summary of the Invention

[0006] The purpose of this invention is to provide a quantitative assessment method for the comprehensive risk of meteorological disasters based on major food crops, in order to solve the aforementioned problems.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a quantitative assessment method for comprehensive meteorological disaster risks based on major grain crops, comprising the following steps:

[0008] Step S1: Detrendization is performed on the crop yield series, decomposing the crop yield series into trend yield and meteorological yield. The specific formula is as follows:

[0009] (1);

[0010] (2);

[0011] In the formula, For crop yield per unit area, The trend output determined by socio-economic and technological development. Meteorological output affected by meteorological factors This is relative meteorological yield, reflecting the intensity of the impact of meteorological factors on yield;

[0012] Step S2: Calculate risk indicators, including average production reduction rate, maximum production reduction rate, average volatility coefficient, volatility tendency rate, yield variation coefficient, production reduction variation coefficient, and probability of different production reduction rates occurring;

[0013] Step S3: Standardize the risk indicators and calculate the comprehensive risk level, including standardizing the risk indicators and selecting 18 indicators to calculate the comprehensive risk level. The 18 indicators are the trend yield tendency rate, average yield reduction rate, maximum yield reduction rate, yield reduction variation coefficient, volatility tendency rate, average volatility coefficient, yield variation coefficient, average yield reduction rate and average volatility coefficient in the 1980s, 1990s, 2000s and 2010s, and the probability of the yield reduction rate occurring in the intervals of [0, 5%], (5%, 10%), and (30%, 100%).

[0014] Step S4: Classify the overall risk level.

[0015] Furthermore, the formula for calculating the average production reduction rate in step S2 is as follows:

[0016] ;

[0017] In the formula, ADR is the average production reduction rate. Let be the relative meteorological yield in year i, and n be the number of years of reduced yield. The average reduction rate from 1984 to 1990, 1991 to 2000, 2001 to 2010, and 2011 to 2020 represents the interdecadal average reduction rate for the 1980s, 1990s, and the first decade and 2010 of the 21st century, respectively.

[0018] The formula for calculating the maximum production reduction rate is: ;

[0019] In the formula, MR represents the maximum production reduction rate. This is a relative meteorological yield series;

[0020] The formula for calculating the average volatility coefficient is:

[0021] ;

[0022] In the formula, BR is the average volatility coefficient. Let n be the relative meteorological yield in year i, and n be the year. We also calculate the interdecadal average fluctuation coefficient.

[0023] The formula for calculating the volatility tendency rate is:

[0024] ;

[0025] In the formula, slope is the volatility tendency rate. This is a relative meteorological yield series, where year is the annual series;

[0026] The formula for calculating the coefficient of variation of yield per unit area is:

[0027] ;

[0028] In the formula, The coefficient of variation for yield per unit area. For the crop yield per unit area in year i, Let n be the average yield, and n be the number of years.

[0029] The formula for calculating the coefficient of variation of production reduction is:

[0030] ;

[0031] In the formula, Cv is the coefficient of variation for production reduction. Let be the relative meteorological output in the i-th year of reduced production, ADR be the average reduction rate, and n be the number of years of reduced production.

[0032] The formula for calculating the probability of different production reduction rates is as follows:

[0033] ;

[0034] In the formula, To reduce production within the range The probability of it occurring in time For relative meteorological yield, To determine the probability of a certain rate of production reduction, each risk indicator is classified using the percentile method: low risk zone (0-20%), low-to-medium risk zone (20%-40%), medium risk zone (40%-60%), medium-to-high risk zone (60%-80%), and high risk zone (80-100%). For interdecadal variations, a combination of equal intervals and percentiles is used for classification.

[0035] Furthermore, the standardization of risk indicators in step S3 adopts the deviation standardization method, and the specific formula is as follows:

[0036] ;

[0037] In the formula, The original data, and They represent The maximum and minimum values, The standardized data has a value range of [0, 1.0]; when The risk is highest when the value is 1.0. A value of 0 indicates the lowest risk; if Follow When increasing from 0 to 1.0, a positive normalization calculation formula is used; if Follow When production decreases and increases from 0 to 1.0, a negative standardization formula is used. This formula includes the trend yield tendency rate, average production reduction rate, maximum production reduction rate, production reduction coefficient of variation, the average production reduction rate for the 1980s, 1990s, 2000s, and 2010s, and the probability of production reduction rates falling within the ranges of [0, 5%] and (5%, 10%), all calculated using negative standardization. Conversely, the volatility tendency rate, average volatility coefficient, yield coefficient of variation, the average volatility coefficient for the 1980s, 1990s, 2000s, and 2010s, and the probability of production reductions exceeding 30%, are calculated using positive standardization.

[0038] The formula for calculating the overall risk level is:

[0039] ;

[0040] In the formula, To assess the overall risk level, For the first The risk level of each risk indicator For the first The weights of the risk indicators are defined as follows, where n is the number of indicators, and equal weights (1 / 18) are used here.

[0041] Furthermore, in step S4, the standard deviation grading method and K-means clustering algorithm are used to classify the comprehensive risk level; the standard deviation grading method is based on the sample population, and the sample population is divided into 5 levels according to the sample mean (μ) and standard deviation (σ). The grading criteria for each risk level are as follows:

[0042] [μ+1.5σ,1] represents high risk;

[0043] [μ+0.5σ,μ+1.5σ] represents medium to high risk;

[0044] [μ-0.5σ,μ+0.5σ] represents medium risk;

[0045] [μ-1.5σ, μ-0.5σ] represent low to medium risk;

[0046] [0, μ-1.5σ] represents low risk;

[0047] Where μ is the mean of the risk index; σ is the standard deviation of the risk index;

[0048] The specific steps of the K-means clustering algorithm are as follows:

[0049] The first step is to standardize the raw data and determine the initial parameters;

[0050] Step 2: Randomly select K sample points from the dataset as initial cluster centers, and perform initial classification on the sample points, assigning the sample points to the cluster with the smallest distance;

[0051] Step 3: Calculate the centroid of each class, update the cluster center of each cluster, and then calculate the distance of each sample to the new agglomeration center and assign it to the class of the nearest agglomeration center. The process terminates when the calculated center is exactly the same as the original agglomeration center; otherwise, the process is repeated. Step 4: Repeat the iterative process, and each iteration reduces the corresponding classification function. When the agglomeration centers of the two iterations are exactly the same, the calculation process converges, and the classification function tends to a constant value, resulting in the ideal classification result.

[0052] Compared with existing technologies, the quantitative assessment method for comprehensive meteorological disaster risks based on major food crops provided by this invention has the following beneficial effects:

[0053] This quantitative assessment method for comprehensive meteorological disaster risks based on major grain crops achieves several key advantages: First, by integrating 18 multi-dimensional indicators, it overcomes the limitations of single-disaster assessment, enabling the quantification of comprehensive risks from multiple disasters such as drought and floods, thus alleviating the difficulty of traditional methods in addressing cascading disasters. Second, it is adaptable to major grain crops such as wheat, rice, and corn, accurately capturing risk differences among different crops by combining the trends, fluctuations, and extreme values ​​of crop meteorological yields. Third, it employs standardized processing and weighted calculations to eliminate differences in indicator units, improving the comparability and objectivity of assessment results. Fourth, it innovates a dual-level method (standard deviation + K-means clustering) to ensure the scientific rigor and stability of risk level classification. Fifth, it can directly serve regional agricultural meteorological disaster prevention decision-making, providing precise basis for crop layout optimization and disaster emergency plan formulation through quantitative results and visualization, effectively enhancing the meteorological risk prevention and control capabilities for grain production. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0055] Figure 1 This is a schematic diagram of the quantitative assessment process for comprehensive meteorological disaster risks of grain crops according to the present invention. Detailed Implementation

[0056] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0057] Please see Figure 1A quantitative assessment method for the comprehensive meteorological disaster risk of major food crops includes the following steps:

[0058] Step S1: Detrendization is performed on the crop yield series, decomposing the crop yield series into trend yield and meteorological yield. The specific formula is as follows:

[0059] (1);

[0060] (2);

[0061] In the formula, For crop yield per unit area, The trend output determined by socio-economic and technological development. Meteorological output affected by meteorological factors This is relative meteorological yield, reflecting the intensity of the impact of meteorological factors on yield;

[0062] Step S2: Calculate risk indicators, including average production reduction rate, maximum production reduction rate, average volatility coefficient, volatility tendency rate, yield variation coefficient, production reduction variation coefficient, and probability of different production reduction rates occurring;

[0063] Step S3: Standardize the risk indicators and calculate the comprehensive risk level, including standardizing the risk indicators and selecting 18 indicators to calculate the comprehensive risk level. The 18 indicators are the trend yield tendency rate, average yield reduction rate, maximum yield reduction rate, yield reduction variation coefficient, volatility tendency rate, average volatility coefficient, yield variation coefficient, average yield reduction rate and average volatility coefficient in the 1980s, 1990s, 2000s and 2010s, and the probability of the yield reduction rate occurring in the intervals of [0, 5%], (5%, 10%), and (30%, 100%).

[0064] Step S4: Classify the overall risk level.

[0065] The formula for calculating the average production reduction rate in step S2 is:

[0066] ;

[0067] In the formula, ADR is the average production reduction rate. Let be the relative meteorological yield in year i, and n be the number of years of reduced yield. The average reduction rate from 1984 to 1990, 1991 to 2000, 2001 to 2010, and 2011 to 2020 represents the interdecadal average reduction rate for the 1980s, 1990s, and the first decade and 2010 of the 21st century, respectively.

[0068] The formula for calculating the maximum production reduction rate is: ;

[0069] In the formula, MR represents the maximum production reduction rate; The highest yield reduction rate is a relative meteorological yield series, reflecting the severity of crop disasters in the region and is the highest yield reduction in history.

[0070] The formula for calculating the average volatility coefficient is:

[0071] ;

[0072] In the formula, BR is the average volatility coefficient. Let n be the relative meteorological yield in year i, and n be the year. We also calculate the interdecadal average fluctuation coefficient. The average fluctuation coefficient represents the average situation of yield fluctuation caused by meteorological factors and reflects the instability of yield under the influence of meteorological factors.

[0073] The formula for calculating the volatility tendency rate is:

[0074] ;

[0075] In the formula, slope is the volatility tendency rate. This is a relative meteorological yield series, where year is the annual series;

[0076] The formula for calculating the coefficient of variation of yield per unit area is:

[0077] ;

[0078] In the formula, The coefficient of variation for yield per unit area. For the crop yield per unit area in year i, Let n be the average yield, n be the annual series, and the yield variation coefficient reflects the fluctuation of yield per unit area. It is the ratio of the standard deviation of the actual yield per unit area to the average value, reflecting the dispersion of the yield data and characterizing the fluctuation and variation of yield per unit area. The larger the variation coefficient, the more obvious the yield fluctuation.

[0079] The formula for calculating the coefficient of variation of production reduction is:

[0080] ;

[0081] In the formula, Cv is the coefficient of variation for production reduction. Let be the relative meteorological yield in the i-th year of reduced yield, ADR be the average yield reduction rate, n be the number of years of reduced yield, and the coefficient of variation of yield reduction reflects the fluctuation of yield reduction and characterizes the degree of variation of crop yield reduction rate.

[0082] The formula for calculating the probability of different production reduction rates is as follows:

[0083] ;

[0084] In the formula, To reduce production within the range The probability of it occurring in time For relative meteorological yield, To determine the probability of a certain rate of production reduction, each risk indicator is classified using the percentile method: low risk zone (0-20%), low-to-medium risk zone (20%-40%), medium risk zone (40%-60%), medium-to-high risk zone (60%-80%), and high risk zone (80-100%). For interdecadal variations, a combination of equal intervals and percentiles is used for classification.

[0085] In step S3, the standardization of risk indicators uses the deviation standardization method, and the specific formula is as follows:

[0086] ;

[0087] In the formula, The original data, and They represent The maximum and minimum values, The standardized data has a value range of [0, 1.0]; when The risk is highest when the value is 1.0. A value of 0 indicates the lowest risk; if Follow When increasing from 0 to 1.0, a positive normalization calculation formula is used; if Follow When production decreases and increases from 0 to 1.0, a negative standardization formula is used. This formula includes the trend yield tendency rate, average production reduction rate, maximum production reduction rate, production reduction coefficient of variation, the average production reduction rate for the 1980s, 1990s, 2000s, and 2010s, and the probability of production reduction rates falling within the ranges of [0, 5%] and (5%, 10%), all calculated using negative standardization. Conversely, the volatility tendency rate, average volatility coefficient, yield coefficient of variation, the average volatility coefficient for the 1980s, 1990s, 2000s, and 2010s, and the probability of production reductions exceeding 30%, are calculated using positive standardization.

[0088] The formula for calculating the overall risk level is:

[0089] ;

[0090] In the formula, To assess the overall risk level, For the first The risk level of each risk indicator For the first The weights of the risk indicators are assigned, where n is the number of indicators, and equal weights (1 / 18) are used here. Based on the characteristics of each indicator, 18 indicators are selected, including the trend yield tendency rate, average yield reduction rate, maximum yield reduction rate, yield reduction variation coefficient, volatility tendency rate, average volatility coefficient, yield variation coefficient, average yield reduction rate and average volatility coefficient in the 1980s, 1990s, 2000s and 2010s, and the probability of yield reduction rate occurring in the intervals of [0, 5%], (5%, 10%), and (30%, 100%). These indicators are used to calculate the comprehensive risk level. After standardization, the weighted average of each indicator is taken as the comprehensive risk level of each county.

[0091] In step S4, the standard deviation grading method and K-means clustering algorithm are used to classify the comprehensive risk level. The standard deviation grading method is based on the sample population and divides the sample population into 5 levels according to the sample mean (μ) and standard deviation (σ). The grading criteria for each risk level are as follows:

[0092] [μ+1.5σ,1] represents high risk;

[0093] [μ+0.5σ,μ+1.5σ] represents medium to high risk;

[0094] [μ-0.5σ,μ+0.5σ] represents medium risk;

[0095] [μ-1.5σ, μ-0.5σ] represent low to medium risk;

[0096] [0, μ-1.5σ] represents low risk;

[0097] Where μ is the mean of the risk index; σ is the standard deviation of the risk index;

[0098] The specific steps of the K-means clustering algorithm are as follows:

[0099] The first step is to standardize the raw data and determine the initial parameters;

[0100] Step 2: Randomly select K sample points from the dataset as initial cluster centers, and perform initial classification on the sample points, assigning the sample points to the cluster with the smallest distance;

[0101] Step 3: Calculate the centroid of each class, update the cluster center of each cluster, and then calculate the distance of each sample to the new agglomeration center and assign it to the class of the nearest agglomeration center. The process terminates when the calculated center is exactly the same as the original agglomeration center; otherwise, the process is repeated. Step 4: Repeat the iterative process, and each iteration reduces the corresponding classification function. When the agglomeration centers of the two iterations are exactly the same, the calculation process converges, and the classification function tends to a constant value, resulting in the ideal classification result.

[0102] Example 1: Comprehensive Risk Assessment of Meteorological Disasters on Grain Crops in Gansu Province

[0103] Regional characteristics: Gansu Province is located in the arid and semi-arid region of Northwest my country, in the upper reaches of the Yellow River, at the intersection of the Qinghai-Tibet Plateau, Inner Mongolia Plateau, and Loess Plateau, with an area of ​​45.37 × 10⁻⁶. 4 km 2 Gansu Province has a long and narrow shape, stretching 1655 km east to west and 530 km north to south, with an altitude ranging from 618 to 5482 meters. The terrain slopes from southwest to northeast, exhibiting a complex and diverse topography, with mountains, plateaus, plains, river valleys, deserts, and Gobi interspersed throughout, making its ecological environment extremely fragile. The province can be roughly divided into six major regions: the Longnan Mountains, the Longzhong Loess Plateau, the Gannan Plateau, the Qilian Mountains, and the Hexi Corridor. The climate is complex and diverse, ranging from subtropical humid to arid zones, with multiple climatic zones coexisting. The average annual temperature is 0℃ to 14℃, annual sunshine hours are 1700 to 3300 hours, and average annual rainfall is 42 to 760 mm, with significant regional variations and uneven seasonal distribution. Agriculture is the dominant industry in Gansu Province, but the region is prone to meteorological disasters, accounting for 88.5% of all natural disasters, higher than the national average, severely impacting the development of agriculture and the socio-economic situation in Gansu.

[0104] Key indicator calculations: 1) Grain output in most cities and counties of Gansu Province shows an increasing trend, with Wuwei and Linxia Prefecture showing the largest growth rates, exceeding 120 kg / hm² per year. 2 2) In Gansu Province, 39.24% of counties showed a positive relative trend in yield fluctuations due to meteorological factors, indicating that grain production was still unstable and significantly affected by meteorological conditions. 3) Before 2000, the average yield reduction rate in most counties of Gansu Province was still below 20%. From the 1980s to the 1990s, the average yield reduction rate increased in 56.96% of counties, but the spatial distribution of severely affected and lightly affected areas remained relatively unchanged. From 2000 to 2010, the average yield reduction rate increased significantly, with 37.97% of counties experiencing an average yield reduction rate exceeding 20%. The number of severely affected grain-producing areas increased significantly and showed a clear shift. After 2010, the average yield reduction rate decreased significantly, with 87.34% of counties experiencing a significant decrease in average yield reduction, and the damage in most counties improved significantly. In the 1980s, 1990s, and 2000s and 2010s, the average reduction rate of grain production in Gansu Province was 5.45%, 6.59%, 8.09%, and 4.23%, respectively, with average fluctuation coefficients of 5.00%, 5.51%, 7.05%, and 3.82%. Clearly, the average reduction rate in most cities and prefectures of Gansu Province only decreased significantly after 2010, and grain production only then tended to stabilize. 4) Counties prone to severe disasters are mainly distributed in the Hexi Corridor, with a probability of over 8.11%. However, 27.85% of counties do not experience severe or major disasters, mainly distributed in southwestern Gansu.

[0105] The risk levels for grain production vary significantly across counties within Gansu Province's prefecture-level administrative regions, with both high-risk and low-risk counties existing. Overall, Huating County and Jinchang City are classified as high-risk areas, while Cheng County, Diebu County, Jinta County, and Xiahe County are classified as low-risk areas.

[0106] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A quantitative assessment method for comprehensive meteorological disaster risks based on major grain crops, characterized in that, Includes the following steps: Step S1: Detrendization is performed on the crop yield series, decomposing the crop yield series into trend yield and meteorological yield. The specific formula is as follows: （1）； （2）； In the formula, For crop yield per unit area, The trend output determined by socio-economic and technological development. Meteorological output affected by meteorological factors This is relative meteorological yield, reflecting the intensity of the impact of meteorological factors on yield; Step S2: Calculate risk indicators, including average production reduction rate, maximum production reduction rate, average volatility coefficient, volatility tendency rate, yield variation coefficient, production reduction variation coefficient, and probability of different production reduction rates occurring; The formula for calculating the average production reduction rate in step S2 is: ； In the formula, ADR is the average production reduction rate. Let be the relative meteorological yield in year i, and n be the number of years of reduced yield. The average reduction rate from 1984 to 1990, 1991 to 2000, 2001 to 2010, and 2011 to 2020 represents the interdecadal average reduction rate for the 1980s, 1990s, and the first decade and 2010 of the 21st century, respectively. The formula for calculating the maximum production reduction rate is: ; In the formula, MR represents the maximum production reduction rate; This is a relative meteorological yield series; The formula for calculating the average volatility coefficient is: ； In the formula, BR is the average volatility coefficient. Let n be the relative meteorological yield in year i, and n be the year. We also calculate the interdecadal average fluctuation coefficient. The formula for calculating the volatility tendency rate is: ； In the formula, slope is the volatility tendency rate. This is a relative meteorological yield series, where year is the annual series; The formula for calculating the coefficient of variation of yield per unit area is: ； In the formula, The coefficient of variation for yield per unit area is . For the crop yield per unit area in year i, Let n be the average yield, and n be the number of years. The formula for calculating the coefficient of variation of production reduction is: ； In the formula, Cv is the coefficient of variation for production reduction. Let be the relative meteorological output in the i-th year of reduced production, ADR be the average reduction rate, and n be the number of years of reduced production. The formula for calculating the probability of different production reduction rates is as follows: ； In the formula, To reduce production within the range The probability of it occurring in time For relative meteorological yield, To determine the probability of a certain rate of production reduction, each risk indicator is classified using a percentile method: low risk zone (0-20%), low-to-medium risk zone (20%-40%), medium risk zone (40%-60%), medium-to-high risk zone (60%-80%), and high risk zone (80-100%). For interdecadal variations, a combination of equal intervals and percentiles is used for classification. Step S3: Standardize the risk indicators and calculate the comprehensive risk level, including standardizing the risk indicators and selecting 18 indicators to calculate the comprehensive risk level. The 18 indicators are the trend yield tendency rate, average yield reduction rate, maximum yield reduction rate, yield reduction variation coefficient, volatility tendency rate, average volatility coefficient, yield variation coefficient, average yield reduction rate and average volatility coefficient in the 1980s, 1990s, 2000s and 2010s, and the probability of the yield reduction rate occurring in the intervals of [0, 5%], (5%, 10%), and (30%, 100%). Step S4: Classify the overall risk level.

2. The method for quantitative assessment of comprehensive meteorological disaster risk based on major grain crops according to claim 1, characterized in that, In step S3, the standardization of risk indicators uses the deviation standardization method, and the specific formula is as follows: ； In the formula, The original data, and They represent The maximum and minimum values, The standardized data has a value range of [0, 1.0]; when The risk is highest when the value is 1.

0. A value of 0 indicates the lowest risk; if Follow When increasing from 0 to 1.0, a positive normalization calculation formula is used; if Follow When production decreases and increases from 0 to 1.0, a negative standardization formula is used. This formula includes the trend yield tendency rate, average production reduction rate, maximum production reduction rate, production reduction coefficient of variation, the average production reduction rate for the 1980s, 1990s, 2000s, and 2010s, and the probability of production reduction rates falling within the ranges of [0, 5%] and (5%, 10%), all calculated using negative standardization. Conversely, the volatility tendency rate, average volatility coefficient, yield coefficient of variation, the average volatility coefficient for the 1980s, 1990s, 2000s, and 2010s, and the probability of production reductions exceeding 30%, are calculated using positive standardization. The formula for calculating the overall risk level is: ； In the formula, To assess the overall risk level, For the first The risk level of each risk indicator For the first The weights of the risk indicators are defined, where n is the number of indicators, and equal weights are used here.

3. The method for quantitative assessment of comprehensive meteorological disaster risk based on major grain crops according to claim 1, characterized in that, Step S4 employs the standard deviation grading method and K-means clustering algorithm to comprehensively classify the risk level. The standard deviation grading method is based on the population sample, dividing it into 5 levels according to the sample mean and standard deviation. The grading criteria for each risk level are as follows: [μ+1.5σ,1] represents high risk; [μ+0.5σ,μ+1.5σ] represents medium to high risk; [μ-0.5σ,μ+0.5σ] represents medium risk; [μ-1.5σ, μ-0.5σ] represent low to medium risk; [0, μ-1.5σ] represents low risk; Where μ is the mean of the risk index; σ is the standard deviation of the risk index; The specific steps of the K-means clustering algorithm are as follows: The first step is to standardize the raw data and determine the initial parameters; Step 2: Randomly select K sample points from the dataset as initial cluster centers, and perform initial classification on the sample points, assigning the sample points to the cluster with the smallest distance; Step 3: Calculate the centroid of each class, update the cluster center of each cluster, calculate the distance of each sample to the new agglomeration center, and assign it to the class to which the nearest agglomeration center belongs; the process terminates when the calculated center is exactly the same as the original agglomeration center, otherwise the process will be repeated. Step 4: Repeat the iterative process, reducing the corresponding classification function with each iteration; when the aggregation centers of the two iterations are exactly the same, the calculation process converges, the classification function tends to a constant value, and the ideal classification result is obtained.

Images