Agricultural drought comprehensive risk assessment method based on grid scale and risk index

Abstract

The invention discloses an agricultural drought comprehensive risk assessment method based on a grid scale and a risk index, and the method comprises the steps: obtaining agricultural drought yield data under a preset grid scale, and selecting an agricultural drought comprehensive risk assessment index according to an impact factor; according to the method, crop drought comprehensive risk indexes are obtained through a constructed agricultural drought comprehensive risk evaluation model, meteorological per unit yield characterization of crops is obtained by adopting an HP filtering method according to actual per unit yield and trend per unit yield of the crops so as to construct an optimal joint distribution function, and meanwhile, the optimal joint distribution function is constructed based on a Monte Carlo simulation method. And according to the nonlinear relationship between the drought comprehensive risk and the crop meteorological yield measured by the optimal joint distribution function, a determination model for realizing the pure rate of agricultural drought yield insurance is obtained, so that the pure rate of the crop yield insurance is obtained, the difference between the calculated insurance rate and the actual insurance rate is greatly reduced, and the accuracy of the agricultural drought yield insurance is improved. The drought comprehensive risk threshold value used for adjusting the pure rate of the agricultural drought yield insurance is obtained through the obtained determination model, so that comprehensive evaluation of the crop yield loss is realized according to the drought comprehensive risk threshold value, pure rate optimization and early warning of the crop yield insurance are realized, and a theoretical basis is provided for scientific formulation of the insurance rate.

Description

Technical Field

[0001] The present invention relates to the technical field of pure premium rates for agricultural drought yield insurance, and particularly to a comprehensive risk assessment method for agricultural drought based on grid scale and risk index. Background Art

[0002] In recent years, affected by global climate change, the climate has tended to be warm and dry, and drought disasters have occurred frequently, bringing a non-negligible impact on agricultural production. Most of the existing pure premium rate determination schemes for agricultural drought yield insurance are the "one rate for one province" system, and most of them use linear correlation relationships to construct the dependence structure between crop yields and weather indices, thereby establishing a linear compensation function for compensation. However, the existing comprehensive assessment methods for agricultural drought yields have the following problems:

[0003] (1) Most previous studies were "one rate for one province", and the insurance rate could not be determined according to the actual disaster risk, resulting in a large difference between the obtained insurance rate and the actual insurance rate, and a low accuracy rate of the comprehensive assessment results of agricultural drought yields;

[0004] (2) Previous studies used linear correlation relationships to construct the dependence structure between crop meteorological yields and weather indices, that is, the functional relationship of mutual dependence between variables, thereby establishing a linear compensation function for compensation. A linear function means that there is a linear relationship between variables in proportion. There is not only a linear relationship but also a non-linear relationship between the comprehensive drought risk index and crop meteorological yields. Therefore, assuming that there is only a linear relationship between the comprehensive drought risk index and crop meteorological yields and establishing a linear compensation function, this assumption is too strict and one-sided, and thus it cannot comprehensively and effectively achieve a comprehensive assessment and risk prediction of agricultural drought yields. Summary of the Invention

[0005] The present invention provides a comprehensive risk assessment method for agricultural drought based on grid scale and risk index to overcome the above technical problems.

[0006] To achieve the above object, the technical solution of the present invention is as follows:

[0007] A comprehensive risk assessment method for agricultural drought based on grid scale and risk index, comprising the following steps:

[0008] S1: Obtain agricultural drought yield data under a preset grid scale, and based on the theory of natural disaster risk formation, select comprehensive risk evaluation indicators for agricultural drought according to influencing factors to construct a comprehensive risk evaluation model for agricultural drought;

[0009] And the influencing factors include the danger of disaster-causing factors, the vulnerability of disaster-bearing bodies, the exposure of disaster-forming environments, and the disaster prevention and mitigation capabilities;

[0010] S2: Obtain the comprehensive risk index of crop drought based on the comprehensive risk assessment model of agricultural drought, and use the HP filtering method to obtain the meteorological yield representation of the crop according to the actual yield and trend yield of the crop;

[0011] S3: Construct an optimal joint distribution function according to the comprehensive risk index of crop drought and the yield representation;

[0012] S4: Based on the Monte Carlo simulation method, obtain a calibration model for realizing the pure premium rate of agricultural drought yield insurance according to the optimal joint distribution function;

[0013] S5: Based on the calibration model, obtain a comprehensive risk threshold of drought for adjusting the pure premium rate of agricultural drought yield insurance, so as to realize the comprehensive risk assessment of crop yield loss according to the comprehensive risk threshold of drought.

[0014] Further, the S1 specifically includes the following steps:

[0015] S11: Obtain the agricultural drought yield data under the preset grid scale, and based on the theory of natural disaster risk formation, select the comprehensive risk assessment indicators of agricultural drought according to the influencing factors, including drought frequency, drought intensity, efficiency index, yield variation coefficient, exposure index, total power of agricultural machinery, ratio of effective irrigation area, and per capita disposable income of farmers;

[0016] S12: Use the subjective analytic hierarchy process, take the influencing factors as the target layer, and take each comprehensive risk assessment indicator of agricultural drought as the criterion layer to obtain a hierarchical structure model;

[0017] S13: Evaluate the hierarchical structure model according to expert experience, and use the 1-9 scale method to quantify and evaluate the importance of the elements in the same layer relative to the elements in the upper layer, and define a subjective evaluation judgment matrix for the weights of the comprehensive risk assessment indicators of agricultural drought;

[0018] S14: Normalize the elements in the subjective evaluation judgment matrix, and obtain the eigenvector of the indicator weights of each comprehensive risk assessment indicator of agricultural drought based on the eigenvalue method;

[0019] S15: Normalize the eigenvector of the indicator weights, and obtain the maximum eigenvalue in the subjective evaluation judgment matrix, and conduct a consistency test on the subjective evaluation judgment matrix according to the maximum eigenvalue;

[0020] If it is verified that the subjective evaluation judgment matrix has consistency, confirm the first indicator weight according to the subjective evaluation judgment matrix; otherwise, obtain a new subjective evaluation judgment matrix through expert experience evaluation again, and repeat the execution from S12 to S14;

[0021] S16: Determine the second index weight of each agricultural drought comprehensive risk assessment index based on the objective entropy weight method;

[0022] S17: Based on the combined weighting method, obtain the combined weight of each agricultural drought comprehensive risk assessment index according to the first index weight and the second index weight, and the expression of the combined weighting method is

[0023]

[0024] In the formula: represents the analytic hierarchy process weight coefficient of index j; represents the entropy weight method coefficient of index j; w j represents the combined weight coefficient of index j; m represents the number of agricultural drought comprehensive risk assessment indexes;

[0025] S18: Obtain each impact factor index according to the combined weight, and its expression is

[0026]

[0027] In the formula: G represents the factor index of hazard, vulnerability, exposure or disaster prevention and mitigation ability; A i represents the quantization value of the i-th index and 0 ≤ A i ≤ 1; w i represents the weight coefficient calculated by the combined weighting method for the i-th index and w i > 0, n represents the number of system evaluation indexes;

[0028] S19: Take the meteorological single yield of crops as the reference sequence, and take the hazard factor index, exposure factor index, vulnerability factor index and disaster prevention and mitigation ability factor index as the comparison sequences;

[0029] And perform dimensionless processing on the reference sequence and the comparison sequences;

[0030] Based on the grey relational analysis method, obtain the grey relational coefficients of the reference sequence and the comparison sequences after dimensionless processing;

[0031] According to the grey relational coefficients and the combined weights of each agricultural drought comprehensive risk assessment index, obtain the grey relational weights of each impact factor;

[0032] S20: According to the impact factor indexes and the grey relational weights of the corresponding impact factors, obtain the agricultural drought comprehensive risk assessment model, and its expression is

[0033]

[0034] Where: X is the drought disaster risk index used to represent the degree of drought disaster risk; H, E, V, and C respectively represent the hazard factor index, exposure factor index, vulnerability factor index, and disaster prevention and mitigation capacity factor index of drought disaster risk; W H , W V , W E , W C respectively represent the weights of the hazard factor, exposure factor, vulnerability factor, and disaster prevention and mitigation capacity factor obtained by grey relational analysis.

[0035] Furthermore, the index selection formula for the comprehensive risk assessment index of agricultural drought disasters in S1 is

[0036] Drought frequency index DF: Where: n represents the number of years with drought disasters in a certain time period; N represents the total number of years in the statistical period;

[0037] Drought intensity index DS: Where: n represents the number of years with a certain level of drought; SPEI i represents the SPEI value of a certain level of drought occurring in a certain time period;

[0038] The efficiency index EI refers to the comparison of the average crop yield of each city with the average crop yield of the whole province: Where: Y A represents the average crop yield of a certain city; Y P represents the average crop yield per unit area of the whole province;

[0039] Yield coefficient of variation index F: Where: n represents the total number of years; Y i represents the crop yield per unit area of a certain city in the i-th year; Y represents the average crop yield of a certain city in n years;

[0040] Exposure index E: Where: A i represents the crop sown area of a certain city in the i-th year; AT i represents the average crop sown area of each city in the i-th year;

[0041] Total power of agricultural machinery index P: Where: P i represents the total power of agricultural machinery of a certain city in the i-th year

[0042] Effective irrigation area ratio index N: Where: m represents the average value of the effective irrigation area of a certain city over the years; M represents the average value of the total crop sown area of a certain city over the years;

[0043] Per capita disposable income index of farmers NI: Where: NIi represents the per capita disposable income of farmers in a certain city in the \(i\)-th year.

[0044] Furthermore, the expression for obtaining the meteorological yield representation of crops according to the actual yield and trend yield of crops by using the HP filtering method in S2 is

[0045] y a = y t + y + ε

[0046] In the formula: y a represents the actual yield of crops; y t represents the trend yield of crops, y represents the meteorological yield of crops; ε represents the random yield of crops.

[0047] Furthermore, the specific steps of S3 are as follows:

[0048] S31: Define the comprehensive drought risk index of crops and the meteorological yield representation as a joint distribution function of a two-dimensional variable, and its expression is

[0049] F(x, y) = C[F(x), F(y)]

[0050]

[0051] In the formula: F(x, y) represents the joint distribution function; F(x) represents the optimal marginal distribution function with respect to the comprehensive drought risk index of x; F(y) represents the optimal marginal distribution function with respect to the meteorological yield y of crops; u1 = F(x), u2 = F(y), and C(u1, u2) represents the Copula distribution function;

[0052] S32: Select several alternative marginal distribution functions, perform distribution fitting on the comprehensive drought risk index of crops and the meteorological yield representation, and confirm the alternative marginal distribution function with the best distribution fitting effect through the goodness-of-fit test KS as the optimal marginal distribution function;

[0053] And the alternative marginal distribution functions at least include the normal distribution function, the lognormal distribution function, the Weibull distribution function, the Logistic distribution function, and the Gamma distribution function;

[0054] S33: Select several Copula alternative functions with respect to the optimal marginal distribution function, perform standardization processing on the comprehensive drought risk index of crops and the meteorological yield representation, and perform distribution fitting on the standardized comprehensive drought risk index of crops and the meteorological yield representation;

[0055] And the Copula alternative functions at least include Gaussian function, Student't function, Clayton function, Gumbel function, and Frank Copula function;

[0056] Confirm the Copula alternative function with the best distribution fitting effect through the Akaike information criterion AIC as the optimal Copula function;

[0057] S34: Confirm the optimal joint distribution function according to the optimal Copula function and the optimal marginal distribution function.

[0058] Further, the S4 specifically includes the following steps:

[0059] S41: Based on the Monte Carlo simulation method, obtain the inverse functions of the optimal marginal distribution functions of the drought comprehensive risk index x and the meteorological yield per unit area of crops according to the optimal joint distribution function, and its expression is

[0060] x = F -1 (u1), y = F -1 (u2)

[0061] And use them as the new drought comprehensive risk index and the new meteorological yield per unit area of crops data;

[0062] S42: According to the inverse function of the optimal marginal distribution function of the meteorological yield per unit area of crops y, obtain the calibration model for realizing the pure premium rate of agricultural drought yield insurance, and its expression is

[0063]

[0064] In the formula: R represents the pure premium rate of agricultural drought yield insurance; f(y) represents the inverse function of the optimal marginal distribution function of the meteorological yield per unit area of crops y; Y represents the meteorological yield per unit area of crops; a represents the guarantee level of the meteorological yield per unit area of crops; represents the expected value of the meteorological yield per unit area of crops.

[0065] Further, in S5, obtain the drought comprehensive risk threshold for adjusting the pure premium rate of agricultural drought yield insurance based on the calibration model, and its expression is

[0066]

[0067] In the formula: C xy represents the optimal joint distribution function of the drought comprehensive risk index and the meteorological yield per unit area of crops; represents the expected yield per unit area of crops at the guarantee level a corresponding drought comprehensive risk index distribution function; x0 represents the drought comprehensive risk threshold.

[0068] Beneficial effects: The present invention provides a comprehensive risk assessment method for agricultural drought disasters based on grid scale and risk index. By obtaining agricultural drought disaster yield data at a preset grid scale, selecting comprehensive risk evaluation indicators for agricultural drought disasters according to influencing factors, and using high-resolution grid data to refine the risk assessment of crop drought disasters and the determination of the pure premium rate of yield insurance, it breaks the limitation of the traditional "one rate for one province"; by constructing a comprehensive risk evaluation model for agricultural drought disasters to obtain the comprehensive risk index of crop drought disasters, and using the HP filtering method to obtain the meteorological yield representation of crops based on the actual and trend yields of crops to construct the optimal joint distribution function. At the same time, based on the Monte Carlo simulation method, the non-linear relationship between the comprehensive risk of drought disasters and the meteorological yield of crops is measured according to the optimal joint distribution function, and a determination model for the pure premium rate of agricultural drought disaster yield insurance is obtained to obtain the pure premium rate of crop yield insurance, which greatly reduces the difference between the calculated insurance rate and the actual insurance rate. And the comprehensive risk threshold of drought disasters for adjusting the pure premium rate of agricultural drought disaster yield insurance is obtained according to the obtained determination model, so as to realize the comprehensive assessment of crop yield losses based on the comprehensive risk threshold of drought disasters, realize the optimization and early warning of the pure premium rate of crop yield insurance, and provide a theoretical basis for scientifically formulating insurance rates. Brief Description of the Drawings

[0069] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0070] Figure 1 It is a flowchart of the comprehensive risk assessment method for agricultural drought disasters based on grid scale and risk index of the present invention;

[0071] Figure 2 It is the core block diagram of the comprehensive risk assessment method for agricultural drought disasters based on grid scale and risk index in this embodiment. Detailed Embodiments

[0072] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0073] This embodiment provides a comprehensive risk assessment method for agricultural drought disasters based on grid scale and risk index, asFigures 1 to 2 As shown in the figure, it includes the following steps:

[0074] S1: Obtain the agricultural drought production data under the preset grid scale, and based on the theory of natural disaster risk formation, select the comprehensive risk assessment indicators for agricultural drought according to the influencing factors to construct a comprehensive risk assessment model for agricultural drought; and the influencing factors include the hazard of the disaster-causing factor, the vulnerability of the disaster-bearing body, the exposure of the disaster-forming environment, and the disaster prevention and mitigation ability.

[0075] Specifically, in this embodiment, the comprehensive risk assessment of crop drought disasters and the determination of the pure premium rate of crop yield insurance are carried out based on 5km×5km high-resolution grid data. Starting from the "four-factor" theory of natural disaster risk formation, with the main line of "before the disaster - during the disaster - after the disaster", through multi-source data fusion technology, indicators representing hazard, vulnerability, exposure, and disaster prevention and mitigation ability are selected, and the natural disaster risk index method is used to evaluate the comprehensive risk of crop drought disasters, which specifically includes the following steps:

[0076] S11: Obtain the agricultural drought production data under the preset grid scale, and based on the theory of natural disaster risk formation, select the comprehensive risk assessment indicators for agricultural drought, including drought frequency, drought intensity, efficiency index, yield variation coefficient, exposure index, total power of agricultural machinery, ratio of effective irrigation area, and per capita disposable income of farmers.

[0077] And the formula for selecting the comprehensive risk assessment indicators of agricultural drought is

[0078] Drought frequency index DF: In the formula: n represents the number of years of drought disasters occurring within a certain time period; N represents the total number of years in the statistical period;

[0079] Drought intensity index DS: In the formula: n represents the number of years of drought of a certain grade; SPEI i represents the SPEI value of drought of a certain grade occurring within a certain time period;

[0080] Efficiency index EI refers to the comparison of the average crop yield of each city with the average crop yield of the whole province: In the formula: Y A represents the average crop yield of a certain city; Y P represents the average unit yield of crops in the whole province;

[0081] Yield variation coefficient index F: In the formula: n represents the total number of years; Y i represents the single crop yield of a certain city in the i-th year; Y represents the average crop yield of a certain city in n years;

[0082] Exposure index E: Where: A i represents the sown area of crops in a certain city in the \(i\)-th year; \(A_T\ i represents the average sown area of crops in each city in the \(i\)-th year;

[0083] Total power index \(P\) of agricultural machinery: Where: \(P\ i represents the total power of agricultural machinery in a certain city in the \(i\)-th year

[0084] Effective irrigation area ratio index \(N\): Where: \(m\) represents the average value of the effective irrigation area in a certain city over the years; \(M\) represents the average value of the total sown area of crops in a certain city over the years;

[0085] Per capita disposable income index \(NI\) of farmers: Where: \(NI\ i represents the per capita disposable income of farmers in a certain city in the \(i\)-th year;

[0086] Based on the theory of natural disaster risk formation, this embodiment starts from four factors: the danger of disaster-causing factors, the vulnerability of disaster-bearing bodies, the exposure of disaster-forming environments, and the disaster prevention and mitigation capabilities. Combining the principles of systematicness, representativeness, availability, and dynamics in index selection, eight sub-indices including drought frequency, drought intensity, efficiency index, yield variation coefficient, exposure index, total power of agricultural machinery, effective irrigation area ratio, and per capita disposable income of farmers are selected to construct an index system, and a comprehensive risk assessment index system for crop drought disasters is obtained. Table 1 shows the selection table of crop drought disaster risk assessment indices;

[0087] Table 1. Selection Table of Crop Drought Disaster Risk Assessment Indices

[0088]

[0089]

[0090] S12: Using the subjective analytic hierarchy process, taking the influencing factors as the target layer and each comprehensive risk assessment index of agricultural drought as the criterion layer to obtain a hierarchical structure model;

[0091] In order to avoid objectivity and uniqueness in the evaluation process, this embodiment uses the subjective analytic hierarchy process to determine the weights of eight sub-indices including drought frequency, drought intensity, efficiency index, yield variation coefficient, exposure index, total power of agricultural machinery, effective irrigation area ratio, and per capita disposable income of farmers;

[0092] S13: Evaluate the hierarchical structure model according to expert experience, and use the 1-9 scale method to quantify and evaluate the importance of elements in the same layer relative to the elements in the upper layer, and define a subjective evaluation judgment matrix for the weights of comprehensive risk assessment indices of agricultural drought;

[0093] S14: Normalize the elements in the subjective evaluation judgment matrix, and obtain the index weight eigenvector of each agricultural drought comprehensive risk evaluation index based on the eigenvalue method;

[0094] S15: Normalize the index weight eigenvector, and obtain the maximum eigenvalue in the subjective evaluation judgment matrix, and conduct a consistency test on the subjective evaluation judgment matrix according to the maximum eigenvalue;

[0095] If it is verified that the subjective evaluation judgment matrix is consistent, confirm the first index weight according to the subjective evaluation judgment matrix; otherwise, obtain a new subjective evaluation judgment matrix through expert experience evaluation again, and repeat the execution from S12 to S14; Table 2 shows the weight results of each evaluation index based on the subjective analytic hierarchy process;

[0096] Table 2. Weight Results of Each Evaluation Index Based on the Analytic Hierarchy Process

[0097]

[0098] S16: Determine the second index weight of each agricultural drought comprehensive risk evaluation index based on the objective entropy weight method;

[0099] In this embodiment, in order to avoid subjective randomness in the evaluation process, the objective entropy weight method is used to determine the weights of 8 sub-indicators including drought frequency, drought intensity, efficiency index, yield variation coefficient, exposure index, total power of agricultural machinery, ratio of effective irrigation area, and per capita disposable income of farmers;

[0100] The objective entropy weight method in this embodiment is implemented through Matlab programming, and the running code is as follows:

[0101]

[0102] Table 3 shows the weight results of each evaluation index based on the objective entropy weight method;

[0103] Table 3. Weight Results of Each Evaluation Index Based on the Entropy Weight Method

[0104]

[0105] S17: In this embodiment, in order to avoid subjective randomness and objective uniqueness in the evaluation process, the combination weighting method is used to weight the weights obtained by the analytic hierarchy process and the weights obtained by the entropy weight method, so as to determine the combined weights of 8 sub-indicators. Based on the combination weighting method, obtain the combined weights of each agricultural drought comprehensive risk evaluation index according to the first index weight and the second index weight, and the expression of the combination weighting method is

[0106]

[0107] Wherein: represents the analytic hierarchy process weight coefficient of index j; represents the entropy weight method coefficient of index j; w j represents the combined weight coefficient of index j; m represents the number of agricultural drought comprehensive risk assessment indicators; Table 4 shows the weight results of each evaluation indicator based on the combined weighting method;

[0108] Table 4. Weight Results of Each Evaluation Indicator Based on the Combined Weighting Method

[0109]

[0110] S18: The weighted comprehensive evaluation method is based on the influence degree of each evaluation indicator on the total goal (four influencing factors), and a corresponding weight coefficient is pre-allocated, and then multiplied by the quantified value of the corresponding evaluated indicator and added together; in this embodiment, the weighted comprehensive evaluation method is used to calculate the hazard, exposure, vulnerability, and disaster prevention and mitigation capacity factor index values of drought disaster risk, that is, the index of each influencing factor is obtained according to the combined weight, and its expression is

[0111]

[0112] Wherein: G represents the factor index of hazard, vulnerability, exposure or disaster prevention and mitigation capacity; A i represents the quantified value of the i-th indicator and 0 ≤ A i ≤ 1; w i represents the weight coefficient obtained by the combined weighting method for the i-th indicator and w i > 0, n represents the number of system evaluation indicators;

[0113] S19: Take the meteorological single yield of crops as the reference sequence, and take the hazard factor index, exposure factor index, vulnerability factor index, and disaster prevention and mitigation capacity factor index as the comparison sequences;

[0114] And perform dimensionless processing on the reference sequence and the comparison sequences;

[0115] Based on the grey relational analysis method, obtain the grey relational coefficients of the reference sequence and the comparison sequences after dimensionless processing;

[0116] According to the grey relational coefficients and the combined weights of each agricultural drought comprehensive risk assessment indicator, obtain the grey relational weights of each influencing factor;

[0117] In this embodiment, the grey relational analysis of Matlab programming software is used. The meteorological yield of crops is used as the reference sequence, and the hazard index, exposure index, vulnerability index and disaster prevention and mitigation capacity index are used as the comparison sequences. Finally, the weights of the four factor indexes of hazard, vulnerability, exposure and disaster prevention and mitigation capacity are obtained. The Matlab implementation code is as follows:

[0118] clc; clear;

[0119] % Import the data in the format of a numerical matrix and rename it as A

[0120] data = A

[0121] r = size(data,1);

[0122] c = size(data,2);

[0123] avg = repmat(mean(data),r,1);

[0124] data = data. / avg;

[0125] Y = data(:,1); % Reference sequence

[0126] X = data(:,2:c); % Comparison sequence

[0127] Y2 = repmat(Y,1,c-1); % Copy the reference sequence to the right for c-1 columns

[0128] absXi_Y = abs(X - Y2)

[0129] a = min(min(absXi_Y)) % Minimum value

[0130] b = max(max(absXi_Y)) % Maximum value

[0131] p = 0.5; % Discrimination coefficient is taken as 0.5

[0132] gamma = (a + p * b). / (absXi_Y + p * b) % Calculate the grey relational degree between each index in the comparison sequence and the reference sequence

[0133] disp("The grey relational degrees of each index in the comparison sequence are respectively:");

[0134] ans = mean(gamma);

[0135] S20: Obtain the comprehensive risk assessment model of agricultural drought according to the grey relational degree weights of the above-mentioned influence factor indexes and the corresponding influence factors. Its expression is

[0136]

[0137] Where: X is the drought disaster risk index, which is used to represent the degree of drought disaster risk. The larger its value, the greater the degree of drought disaster risk; H, E, V, and C respectively represent the hazard factor index, exposure factor index, vulnerability factor index, and disaster prevention and mitigation capacity factor index of drought disaster risk; W H ,W V ,W E ,W C respectively represent the weights of the hazard factor, exposure factor, vulnerability factor, and disaster prevention and mitigation capacity factor obtained by using grey relational analysis;

[0138] S2: Obtain the comprehensive risk index of crop drought disaster according to the agricultural drought comprehensive risk assessment model, and use the HP filtering method to obtain the meteorological yield characterization of the crop according to the actual yield and trend yield of the crop;

[0139] In a specific embodiment, the expression for obtaining the meteorological yield characterization of the crop according to the actual yield and trend yield of the crop by using the HP filtering method is

[0140] y a =y t +y+ε

[0141] Where: y a represents the actual yield of the crop; y t represents the trend yield of the crop, y represents the meteorological yield of the crop; ε represents the random yield of the crop;

[0142] The yield in this embodiment refers to the yield per unit area. The meteorological yield refers to the fluctuation of the crop yield caused by the difference in interannual meteorological conditions. This fluctuating yield component is called the meteorological yield. The actual yield of the crop is formed under the comprehensive influence of various natural and non-natural factors. Usually, the actual yield of the crop can be decomposed into trend yield, meteorological yield, and random yield. The concept of meteorological yield emphasizes the direct impact of meteorological conditions on crop yield, that is, the change of meteorological conditions in different years will lead to the corresponding change of crop yield;

[0143] S3: Construct an optimal joint distribution function according to the comprehensive risk index of crop drought disaster and the yield characterization, which specifically includes the following steps:

[0144] S31: According to Sklar's theorem, for the joint distribution of N random variables, it can be decomposed into the marginal distributions of these N variables and a Copula function, thus separating the randomness and coupling of the variables. Among them, the randomness of each random variable is described by the marginal distribution, and the coupling characteristics between random variables are described by the Copula function. The Copula function is a multi-dimensional joint distribution function with a uniform distribution in the domain [0, 1]. Let the comprehensive drought risk index be x and the meteorological yield per unit area of crops be y. If F is the joint distribution function of a two-dimensional variable, then there exists a two-dimensional Copula function C that can "connect" the marginal distribution and the joint distribution, that is, define the comprehensive drought risk index of crops and the meteorological yield per unit area as the joint distribution function of a two-dimensional variable, and its expression is

[0145] F(x, y) = C[F(x), F(y)]

[0146]

[0147] where: F(x, y) represents the joint distribution function; F(x) represents the optimal marginal distribution function for the comprehensive drought risk index of x; F(y) represents the optimal marginal distribution function for the meteorological yield per unit area of crops y; u1 = F(x), u2 = F(y), and C(u1, u2) represents the Copula distribution function;

[0148] S32: Select several alternative marginal distribution functions, perform distribution fitting on the comprehensive drought risk index of crops and the meteorological yield per unit area, and confirm the alternative marginal distribution function with the best distribution fitting effect through the Kolmogorov-Smirnov (KS) goodness-of-fit test as the optimal marginal distribution function;

[0149] And the alternative marginal distribution functions at least include the normal distribution function, the lognormal distribution function, the Weibull distribution function, the Logistic distribution function, and the Gamma distribution function;

[0150] S33: After determining the optimal marginal fitting distributions of the comprehensive drought risk index and the meteorological yield per unit area of crops, select several Copula alternative functions for the optimal marginal distribution function, perform standardization (normalization) on the comprehensive drought risk index of crops and the meteorological yield per unit area, and perform distribution fitting on the standardized comprehensive drought risk index of crops and the meteorological yield per unit area;

[0151] And the Copula alternative functions at least include the Gaussian function, the Student's t function, the Clayton function, the Gumbel function, and the Frank Copula function;

[0152] The Copula alternative function with the best distribution fitting effect is confirmed by the Akaike Information Criterion (AIC) as the optimal Copula function;

[0153] S34: Confirm the optimal joint distribution function according to the optimal Copula function and the optimal marginal distribution function;

[0154] S4: Based on the Monte Carlo simulation method, obtain the calibration model for realizing the pure premium rate of agricultural drought yield insurance according to the optimal joint distribution function, which specifically includes the following steps:

[0155] S41: After constructing the bivariate joint distribution function of the comprehensive drought risk index and the meteorological yield of crops based on the Copula function, combine the Monte Carlo simulation method with the Copula function. The simulation steps of Monte Carlo are as follows:

[0156] (1) First, using the selected optimal Copula function as the random number generator, generate 10,000 groups of random variables u1 and u2 each that follow a uniform distribution on [0, 1];

[0157] (2) Based on the Monte Carlo simulation method, obtain the inverse functions of the optimal marginal distribution functions of the comprehensive drought risk index x and the meteorological yield of crops y according to the optimal joint distribution function. Its expression is

[0158] x = F -1 (u1), y = F -1 (u2)

[0159] And use them as the new comprehensive drought risk index and new meteorological yield data of crops;

[0160] S42: Obtain the calibration model for realizing the pure premium rate of agricultural drought yield insurance according to the inverse function of the optimal marginal distribution function of the meteorological yield of crops y. Its expression is

[0161]

[0162] In the formula: R represents the pure premium rate of agricultural drought yield insurance; f(y) represents the inverse function of the optimal marginal distribution function of the meteorological yield of crops y; Y represents the meteorological yield of crops; a represents the guarantee level of the meteorological yield of crops; represents the expected value of the meteorological yield of crops;

[0163] And when the meteorological yield of crops Y is lower than the expected yield under a certain guarantee level the insurance company needs to compensate the insured farmers. In this embodiment, the guarantee level α starts from 80%, and five different guarantee degrees of 80%, 85%, 90%, 95%, and 100% are used. The expected yield of crops It is the average of the simulated data of the meteorological yield of crops.

[0164] S5: Obtain the comprehensive drought risk threshold for adjusting the pure rate of agricultural drought yield insurance based on the calibration model, so as to realize the comprehensive risk assessment of crop yield loss according to the comprehensive drought risk threshold. Specifically:

[0165] (1) According to the Copula function C that connects the comprehensive drought risk index and the meteorological yield of crops xy , and the random variable u2 of the meteorological yield of crops simulated based on the Monte Carlo method, calculate and obtain the distribution function of the comprehensive drought risk index u1 = C xy u2;

[0166] (2) Based on the calibration model, obtain the comprehensive drought risk threshold for adjusting the pure rate of agricultural drought yield insurance according to the inverse function of the optimal marginal distribution function of the comprehensive drought risk index x, so as to realize the early warning of crop yield loss according to the comprehensive drought risk threshold;

[0167] The expression of the comprehensive drought risk threshold is

[0168]

[0169] In the formula: C xy represents the optimal joint distribution function of the comprehensive drought risk index and the meteorological yield of crops; represents the expected yield of crops at the guarantee level a The corresponding distribution function of the comprehensive drought risk index; x0 represents the comprehensive drought risk threshold, that is The corresponding distribution function of the comprehensive drought risk index The original function value x0 is calculated by the inverse function of, which is the comprehensive drought risk threshold.

[0170] In this embodiment, the expected yield of crops at a certain guarantee level is obtained through the calibration model of the pure rate of agricultural drought yield insurance And obtain about The corresponding distribution function of the comprehensive drought risk index Finally, based on The comprehensive drought risk threshold x0 is obtained.

[0171] This embodiment conducts a comprehensive risk assessment of crop drought disasters and determines the pure premium rate of yield insurance based on 5km×5km high-resolution grid data; by formulating the pure premium rate in a differentiated and refined manner at a high-resolution spatial scale, it breaks the traditional pricing method of "one rate for one province", making the agricultural risks in different regions within a province heterogeneous. Regions with large planting areas and good production conditions bear lower premiums, while regions with small planting areas and poor production conditions bear higher premiums, thus ensuring the fairness of premium-bearing among different regions and effectively avoiding the risks of moral hazard and adverse selection. By using the two-dimensional Copula function to measure the non-linear relationship between the comprehensive drought risk index and the meteorological single yield of crops, a calibration model for realizing the pure premium rate of agricultural drought yield insurance is obtained to obtain the pure premium rate of crop yield insurance, greatly reducing the difference between the calculated insurance rate and the actual insurance rate. And the comprehensive drought risk threshold for adjusting the pure premium rate of agricultural drought yield insurance is obtained through the obtained calibration model to realize the comprehensive risk assessment of crop yield losses according to the comprehensive drought risk threshold, achieving the optimization and early warning of the pure premium rate of crop yield insurance and providing a theoretical basis for scientifically formulating insurance rates.

[0172] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An integrated risk assessment method for agricultural drought disasters based on grid scale and risk index, characterized in that, It includes the following steps: S1: Obtain the agricultural drought yield data under the preset grid scale, and based on the theory of natural disaster risk formation, select the comprehensive risk assessment indicators for agricultural drought according to the influencing factors to construct a comprehensive risk assessment model for agricultural drought; And the influencing factors include the danger of the disaster-causing factor, the vulnerability of the disaster-bearing body, the exposure of the disaster-forming environment, and the disaster prevention and mitigation ability; S2: Obtain the comprehensive risk index of crop drought according to the comprehensive risk assessment model of agricultural drought, and use the HP filtering method to obtain the meteorological yield representation of the crop according to the actual yield and trend yield of the crop; S3: Construct an optimal joint distribution function according to the comprehensive risk index of crop drought and the yield representation; S4: Based on the Monte Carlo simulation method, obtain a calibration model for realizing the pure premium rate of agricultural drought yield insurance according to the optimal joint distribution function; S5: Based on the calibration model, obtain a comprehensive drought risk threshold for adjusting the pure premium rate of agricultural drought yield insurance, so as to realize the comprehensive risk assessment of crop yield loss according to the comprehensive drought risk threshold.

2. The comprehensive risk assessment method for agricultural drought disasters based on grid scale and risk index according to claim 1, characterized in that The specific steps of S1 include the following: S11: Obtain the agricultural drought yield data under the preset grid scale, and based on the theory of natural disaster risk formation, select the comprehensive risk assessment indicators for agricultural drought according to the influencing factors, including drought frequency, drought intensity, efficiency index, yield variation coefficient, exposure index, total power of agricultural machinery, ratio of effective irrigation area, and per capita disposable income of farmers; S12: Use the subjective analytic hierarchy process, take the influencing factors as the target layer, and take each comprehensive risk assessment indicator of agricultural drought as the criterion layer to obtain a hierarchical structure model; S13: Evaluate the hierarchical structure model according to expert experience, and use the 1-9 scale method to quantify and evaluate the importance of elements in the same layer relative to the elements in the upper layer, and define a subjective evaluation judgment matrix for the weights of the comprehensive risk assessment indicators of agricultural drought; S14: Normalize the elements in the subjective evaluation judgment matrix, and obtain the index weight eigenvector of each comprehensive risk assessment indicator of agricultural drought based on the eigenvalue method; S15: Normalize the index weight eigenvector, and obtain the maximum eigenvalue in the subjective evaluation judgment matrix, and conduct a consistency test on the subjective evaluation judgment matrix according to the maximum eigenvalue; If it is verified that the subjective evaluation judgment matrix has consistency, confirm the first index weight according to the subjective evaluation judgment matrix; otherwise, obtain a new subjective evaluation judgment matrix through expert experience evaluation again, and repeat the execution from S12 to S14; S16: Determine the second index weight of each comprehensive risk assessment indicator of agricultural drought based on the objective entropy weight method; S17: Based on the combined weighting method, obtain the combined weight of each comprehensive risk assessment indicator of agricultural drought according to the first index weight and the second index weight, and the expression of the combined weighting method is In the formula: represents the analytic hierarchy process weight coefficient of index j; represents the entropy weight method coefficient of index j; w j represents the combined weight coefficient of index j; m represents the number of agricultural drought comprehensive risk assessment indicators; S18: Obtain the index of each influencing factor according to the combined weight, and its expression is Where: G represents the factor index of hazard, vulnerability, exposure or disaster prevention and mitigation capacity; A i represents the quantified value of the i-th index and 0 ≤ A i ≤ 1; w i represents the weight coefficient obtained by the combined weighting method for the i-th index and w i > 0, n represents the number of system evaluation indicators; S19: Take the meteorological yield of the crop as the reference sequence, and take the danger factor index, exposure factor index, vulnerability factor index, and disaster prevention and mitigation ability factor index as the comparison sequences; And dimensionless processing is performed on the reference sequence and the comparison sequence; Based on the grey relational analysis method, the grey relational coefficients of the reference sequence and the comparison sequence after dimensionless processing are obtained; According to the grey relational coefficients and the combined weights of each agricultural drought comprehensive risk evaluation index, the grey relational degree weights of each influencing factor are obtained; S20: According to the grey relational degree weights of each influencing factor index and the corresponding influencing factors, an agricultural drought comprehensive risk evaluation model is obtained, and its expression is Where: X is the drought disaster risk index used to represent the degree of drought disaster risk; H, E, V, and C respectively represent the hazard factor index, exposure factor index, vulnerability factor index, and disaster prevention and mitigation capacity factor index of drought disaster risk; W H , W V , W E , W C respectively represent the weights of the hazard factor, exposure factor, vulnerability factor, and disaster prevention and mitigation capacity factor obtained by using grey relational analysis.

3. The comprehensive risk assessment method for agricultural drought disasters based on grid scale and risk index according to claim 2, characterized in that, The index selection formula for the agricultural drought comprehensive risk evaluation index in S1 is Drought frequency index DF: Where: n represents the number of years with drought disasters in a certain time period; N represents the total number of years in the statistical period; Drought intensity index DS: Where: n represents the number of years with a certain level of drought; SPEI i represents the SPEI value of a certain level of drought occurring within a certain time period; The efficiency index EI refers to the comparison of the average crop yield in each city with the average crop yield in the whole province: In the formula: Y A represents the average crop yield in a certain city; Y P represents the average crop yield per unit area in the whole province; Coefficient of variation index F of yield: Where: n represents the total number of years; Y i represents the crop yield per unit area in a certain city in the i-th year; Y represents the average crop yield per unit area in a certain city over n years; Exposure index indicator E: In the formula: A i represents the sown area of crops in a certain city in the i-th year; AT i represents the average sown area of crops in each city in the i-th year; Total power index P of agricultural machinery: Where: P i represents the total power of agricultural machinery in a certain city in the i-th year Effective irrigation area ratio index N: Where: m represents the average value of the effective irrigation area in a certain city over the years; M represents the average value of the total sown area of crops in a certain city over the years; Per capita disposable income index of farmers NI: Where: NI i represents the per capita disposable income of farmers in a certain city in the i-th year.

4. The comprehensive risk assessment method for agricultural drought disasters based on grid scale and risk index according to claim 2, characterized in that In S2, the expression for obtaining the meteorological yield representation of crops based on the actual yield and trend yield of crops using the HP filtering method is y a = y t + y + ε Where: y a represents the actual single crop yield; y t represents the trend single crop yield, y represents the meteorological single crop yield; ε represents the random single crop yield.

5. The comprehensive risk assessment method for agricultural drought disasters based on grid scale and risk index according to claim 4, characterized in that, The specific steps of S3 are as follows: S31: Define the joint distribution function of the comprehensive risk index of crop drought and the meteorological single yield as a two-dimensional variable, and its expression is F(x,y) = C[F(x),F(y)] In the formula: F(x,y) represents the joint distribution function; F(x) represents the optimal marginal distribution function for the comprehensive risk index of drought being x; F(y) represents the optimal marginal distribution function for the meteorological single yield y of crops; u1 = F(x), u2 = F(y), and C(u1,u2) represents the Copula distribution function; S32: Select several alternative marginal distribution functions, perform distribution fitting on the comprehensive risk index of crop drought and the meteorological single yield characterization, and confirm the alternative marginal distribution function with the best distribution fitting effect through the Kolmogorov-Smirnov (KS) goodness-of-fit test as the optimal marginal distribution function; And the alternative marginal distribution functions at least include the normal distribution function, the lognormal distribution function, the Weibull distribution function, the Logistic distribution function, and the Gamma distribution function; S33: Select several Copula alternative functions for the optimal marginal distribution function, perform standardization processing on the comprehensive risk index of crop drought and the meteorological single yield characterization, and perform distribution fitting on the standardized comprehensive risk index of crop drought and the meteorological single yield characterization; And the Copula alternative functions at least include the Gaussian function, the Student's t function, the Clayton function, the Gumbel function, and the Frank Copula function; Confirm the Copula alternative function with the best distribution fitting effect through the Akaike information criterion (AIC) as the optimal Copula function; S34: According to the optimal Copula function and the optimal marginal distribution function, confirm the optimal joint distribution function.

6. The comprehensive risk assessment method for agricultural drought disasters based on grid scale and risk index according to claim 5, characterized in that The specific steps of S4 are as follows: S41: Based on the Monte Carlo simulation method, obtain the inverse functions of the optimal marginal distribution functions of the comprehensive drought risk index x and the meteorological yield per unit area of crops y according to the optimal joint distribution function, and its expression is x = F -1 (u1), y = F -1 (u2) And use it as the new comprehensive risk index of drought and the new meteorological single yield data of crops; S42: According to the inverse function of the optimal marginal distribution function of the meteorological single yield y of crops, obtain the calibration model for realizing the pure premium rate of agricultural drought yield insurance, and its expression is Where: R represents the pure premium rate of agricultural drought yield insurance; f(y) represents the inverse function of the optimal marginal distribution function of the meteorological yield y of crops; Y represents the meteorological yield of crops; a represents the guarantee level of the meteorological yield of crops; represents the expected value of the meteorological yield of crops.

7. The comprehensive risk assessment method for agricultural drought disasters based on grid scale and risk index according to claim 5, characterized in that In S5, the drought comprehensive risk threshold for adjusting the pure premium rate of agricultural drought yield insurance is obtained based on the calibration model, and its expression is In the formula: C xy represents the optimal joint distribution function of the comprehensive drought risk index and the meteorological yield per unit area of crops; represents the expected yield per unit area of crops under the guarantee level a corresponding to the distribution function of the comprehensive drought risk index; x0 represents the comprehensive drought risk threshold.

Images