Agricultural drought risk quantitative evaluation method and system considering drought process

Abstract

The invention provides an agricultural drought risk quantitative evaluation method and system considering a drought process, and relates to the technical field of agriculture and climate risk management. According to the method, drought accumulation duration and intensity of all grid points in a research area are determined by using preset drought indexes, a multi-dimensional drought feature joint probability model of a Copula function is constructed based on the data, drought grades and conversion thresholds are divided, and the farmland drought area is counted in combination with land utilization data. By establishing the Markov decision process and optimizing the drought management policy, the prediction precision of the drought risk is effectively improved, the agricultural loss is remarkably reduced, and the resource utilization efficiency is improved. The method provides a reliable basis for scientific decision making.

Description

Technical Field

[0001] The present invention relates to the technical field of agriculture and climate risk management, and in particular to a method and system for quantitatively assessing agricultural drought risk taking drought processes into consideration. Background Art

[0002] To date, there is no unified scientific definition of drought. The U.S. National Weather Service categorizes droughts into meteorological, hydrological, agricultural, and socioeconomic types based on their causes and impacts. Global warming is significantly changing the frequency, intensity, and extent of droughts, posing significant challenges to the assessment of drought risk, particularly agricultural drought risk.

[0003] According to disaster system theory, drought risk is a comprehensive system that integrates drought hazard, exposure of hazard-bearing bodies, vulnerability of the disaster-prone environment, and disaster prevention and mitigation capabilities. It can also be understood as a combination of the likelihood of a drought event occurring and the negative consequences it will cause. Current research on drought risk focuses primarily on two aspects: 1. The risk of a drought event occurring, and 2. The risk of a drought event potentially resulting in certain severe consequences. Due to the complex causes and regional complexity of drought risk, there is currently no unified drought risk assessment model.

[0004] The risk of a drought event primarily refers to drought hazard, which represents the potential risk of a drought impact. It can be considered the conditional expectation of adverse consequences over a period of time and is an essential factor in drought risk analysis. Droughts often persist for extended periods, and their severity and scope vary over time. Therefore, agricultural drought risk assessment methods typically use a representative agricultural drought index. Based on the joint probability of various drought characteristic variables (duration, intensity, and scope), this index analyzes the spatiotemporal variations of agricultural drought to assess the agricultural drought hazard in a given region. This assessment model assumes a positive correlation between drought hazard and disaster severity: the higher the probability of drought, the more severe the drought. However, these traditional drought risk assessment methods use a frequency analysis of historical drought events to infer the potential risk level of a particular region in the future. These methods fail to consider the impact of ongoing droughts and, in other words, fail to reflect the ongoing drought process.

[0005] The risk of potential impacts from drought primarily refers to the combined drought risk, which quantifies the likelihood that environmental and socioeconomic systems will suffer negative consequences from drought. A widely used agricultural drought risk assessment process involves first selecting multiple indicators representing various risk factors, including drought hazard, exposure, vulnerability, and drought resilience, based on the combined effects of four influencing factors within the drought hazard system. Second, statistical methods (such as fuzzy clustering, analytic hierarchy process, and machine learning algorithms) are used to determine the weights of different drought risk indicators. Finally, a drought risk assessment model is constructed to assess the spatiotemporal variations in agricultural drought hazard, vulnerability, and the combined drought risk. However, the results obtained from this method only reflect the potential impact of local drought in a relative manner and lack specific, quantifiable expression. Summary of the Invention

[0006] In order to overcome the shortcomings of the existing technology, the purpose of the present invention is to provide a quantitative assessment method and system for agricultural drought risk taking into account the drought process. By establishing a comprehensive drought assessment and management model, the prediction accuracy of drought risk and the optimization ability of management strategies are effectively improved, thereby significantly reducing agricultural losses and improving resource utilization efficiency.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] A quantitative assessment method for agricultural drought risk considering drought processes, including:

[0009] The cumulative duration and cumulative intensity of drought at all grid points in the study area are determined using pre-set drought indicators;

[0010] Based on the cumulative duration and the cumulative intensity, a joint probability model of multidimensional drought characteristics is constructed using a Copula function;

[0011] Different levels of drought conversion thresholds are divided according to the joint probability model; the drought conversion thresholds are used to divide drought levels;

[0012] Combining the preset land use type data with the drought conversion threshold to calculate the drought-affected cultivated land area under each drought level in the study area;

[0013] A Markov decision process is established based on the drought level, and a drought management policy is optimized according to the Markov decision process.

[0014] Preferably, the cumulative duration and cumulative intensity of drought at all grid points in the study area are determined by using preset drought indicators, including:

[0015] Determining the drought event persistence rule of the study area according to the preset drought index;

[0016] According to the drought event persistence rule, the cumulative duration and cumulative intensity of drought in the study area are calculated respectively; the cumulative duration D c The calculation formula is D c = c; where c represents the cumulative drought duration from the time scale of the beginning of a drought to a certain moment in the development process; the cumulative intensity S c The calculation formula is: Drought threshold; wherein index is the quantitative value of the drought index.

[0017] Preferably, the drought event persistence rule includes:

[0018] At the i-th time scale, if the quantitative value of the drought indicator is less than the drought threshold, it is considered that drought has occurred in the study area and the drought duration begins to accumulate;

[0019] When drought does not occur at the i+1 time scale but occurs at the i+2 time scale, considering that drought recovery takes time, the i-th time scale, the i+1-th time scale, and the i+2 time scale are considered to be continuously developing drought events.

[0020] Preferably, the expression of the joint probability model is:

[0021] P(Dc≥d∩Sc≥s)=1-F D (d)-F S (s)+C DS (F D (d),F S (s))

[0022] Where P(Dc≥d∩Sc≥s) represents the joint probability that the cumulative drought duration Dc is greater than or equal to the drought duration limit d and the cumulative drought intensity Sc is greater than or equal to the drought intensity limit s; F D (d) means D c The one-dimensional distribution function of the sequence is F S (s) means S c The sequence obeys the one-dimensional distribution function, C DS Indicates D c Sequence and S c The optimal Copula function that the sequence obeys.

[0023] Preferably, the method for classifying drought levels includes:

[0024] When the drought conversion threshold P>P 51.5％ When the drought level is determined to be mild drought;

[0025] When P 21.7％ <P≤P 51.5％When the drought level is determined to be moderate drought;

[0026] When P 7.5％ <P≤P 21.7％ When the drought level is determined to be severe drought;

[0027] When P≤P 7.5％ When the drought level is determined to be extreme drought.

[0028] Preferably, the land use type data includes: forest land, grassland, cultivated land, paddy field, water area, urban construction land and unused land.

[0029] Preferably, establishing a Markov decision process based on the drought level, and optimizing the drought management policy according to the Markov decision process, comprises:

[0030] A state space S is defined, wherein the state space S includes different drought levels; S={mild drought, moderate drought, severe drought, extreme drought};

[0031] Based on historical drought event data, the probability of transitioning from one drought level to another is calculated to form a transition probability matrix P;

[0032] Determine an initial state distribution vector π based on the most recent drought-affected cultivated land area; the initial state distribution vector π represents the probability of each drought level in the current state;

[0033] Based on the transition probability matrix P and the initial state distribution π, the Markov process is applied to deduce the state distribution of future time steps to generate a prediction sequence of future drought levels;

[0034] Formulate management strategies appropriate to current and expected future conditions based on the forecast sequence;

[0035] Using a dynamic programming algorithm, based on the transition probability matrix P, the initial state distribution π and the management strategy, to evaluate the expected benefits under different management strategies;

[0036] The management strategy with the largest expected benefit in each state is selected as the drought management policy.

[0037] Preferably, the calculation formula of the dynamic programming algorithm is:

[0038]

[0039] Among them, V(s t ) is the current state s t The value of , that is, the expected benefit of this state, Indicates that among all the optional strategies a t Choose the value V(st ) maximizing strategy, R(s t ,a t ) is in state s t Take strategy a t The immediate profit at the time of t+1 ∣s t ,a t ) is from state s t Go to the next state s t+1 The transition probability, given in state s t Take strategy a t ,V(s t+1 ) is the next state s t+1 value.

[0040] Preferably, the management strategy includes:

[0041] In the case of mild drought, implement moderate irrigation and allocate water resources to maintain soil moisture;

[0042] In moderate drought conditions, increase irrigation frequency and provide crop adjustment advice to farmers to reduce water consumption;

[0043] In severe drought conditions, implement soil moisture conservation strategies and introduce drought-tolerant crops to reduce crop losses;

[0044] In the face of severe drought, water resource management policies are fully implemented and emergency response measures are initiated to mitigate the impact.

[0045] A quantitative assessment system for agricultural drought risk considering drought processes, including:

[0046] The cumulative analysis unit is used to determine the cumulative duration and cumulative intensity of drought at all grid points in the study area using pre-set drought indicators;

[0047] A model building unit, configured to build a joint probability model of multidimensional drought characteristics using a Copula function based on the cumulative duration and the cumulative severity;

[0048] A threshold division unit, configured to divide drought conversion thresholds of different levels according to the joint probability model; the drought conversion thresholds are used to divide drought levels;

[0049] an area statistics unit, configured to combine preset land use type data with the drought conversion threshold to count drought-affected cultivated land areas at each drought level within the study area;

[0050] A policy optimization unit is used to establish a Markov decision process based on the drought level and optimize the drought management policy according to the Markov decision process.

[0051] The present invention discloses the following technical effects:

[0052] The present invention provides a method and system for quantitatively assessing agricultural drought risk that takes into account the drought process, which mainly achieves effective drought assessment and management through a series of scientific steps. First, the cumulative duration and cumulative intensity of all grid points in the study area are quantified through preset drought indicators to establish basic data; then, a joint probability model of multidimensional drought characteristics is constructed using the Copula function to accurately divide the drought conversion threshold in order to classify different levels of drought states; the area of ​​cultivated land affected by drought is statistically calculated in combination with land use type data, providing a key basis for decision-making; finally, by forming a Markov decision process, not only the drought management policy is optimized, but also the ability to respond to future drought risks is enhanced. In summary, the present invention significantly improves the accuracy of drought risk assessment, optimizes resource allocation, reduces agricultural losses, and provides effective support for sustainable agricultural development. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0054] Figure 1 A schematic diagram of a method flow chart provided in an embodiment of the present invention;

[0055] Figure 2 Schematic diagram of some technical routes provided by embodiments of the present invention;

[0056] Figure 3 A schematic diagram of the cumulative probability of different drought levels provided by an embodiment of the present invention;

[0057] Figure 4 A schematic diagram of the drought development process and distribution of each level in a province provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0059] The purpose of the present invention is to provide a method and system for quantitatively assessing agricultural drought risk taking into account the drought process, which significantly improves the accuracy of drought risk assessment, optimizes resource allocation, reduces agricultural losses, and provides effective support for sustainable agricultural development.

[0060] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0061] Figure 1 A flow chart of a method provided in an embodiment of the present invention is shown in FIG. Figure 1 As shown, the present invention provides a method for quantitatively assessing agricultural drought risk taking into account the drought process, comprising:

[0062] Step 100: Determine the cumulative duration and cumulative intensity of drought at all grid points in the study area using a preset drought index;

[0063] Step 200: Based on the cumulative duration and cumulative intensity, a joint probability model of multidimensional drought characteristics is constructed using a Copula function;

[0064] Step 300: Different levels of drought conversion thresholds are divided according to the joint probability model; the drought conversion thresholds are used to divide drought levels;

[0065] Step 400: combining the preset land use type data with the drought conversion threshold to calculate the drought-affected cultivated land area at each drought level in the study area;

[0066] Step 500: Establish a Markov decision process based on the drought level, and optimize the drought management policy according to the Markov decision process.

[0067] Specifically, such as Figure 2 As shown in the figure, the quantitative agricultural drought risk assessment method proposed in this embodiment, which considers the drought process, mainly includes three parts: first, identifying characteristics of the drought development process, such as the cumulative drought duration and cumulative drought intensity; second, constructing a multidimensional joint probability model of drought characteristics and defining transition thresholds for different drought levels during the drought development process; and third, quantitatively assessing agricultural drought risk. By overlaying land use type data, the area of ​​different drought levels within the farmland is statistically calculated.

[0068] Furthermore, this embodiment takes into account the temporal and spatial cumulative effects of drought during its evolution. This study uses drought indicators to identify the cumulative duration and cumulative intensity of drought at all grid points in the study area. The details are as follows:

[0069] First, the drought event persistence rule is determined: ① At the i-th time scale, if the drought index is less than the drought threshold, drought is considered to have occurred in the study area and the drought duration begins to accumulate; ② When drought does not occur at the i+1 time scale, but drought occurs at the i+2 time scale, considering that drought recovery takes time, the i-th time scale, the i+1 time scale, and the i+2 time scale are considered to be continuously developing drought events (that is, regional droughts within one time scale are considered to be continuous drought events).

[0070] Secondly, according to the above rules, the cumulative duration of regional drought (D c ) and cumulative severity (S c )sequence:

[0071] D c =c

[0072]

[0073] Where c represents the cumulative drought duration from the beginning of a drought to a certain point in its development, S c It is the sum of the absolute values ​​of regional drought indicators within the cumulative drought duration range. The drought threshold is the light drought threshold corresponding to the selected drought indicator.

[0074] Optionally, this embodiment comprehensively considers the above drought characteristics and selects the Copula function to construct a joint probability model. The Copulas function can flexibly simulate the dependency structure between random variables based on the marginal distribution of random variables. It has the advantage of connecting the distribution of multi-dimensional variables and has fewer restrictions on the marginal distribution functions obeyed by multiple variables. Therefore, it is often used to calculate the joint probability of multi-dimensional drought characteristics. According to the existing theorem, assuming X1, X2, ..., X m are multiple continuous random variables whose marginal distribution functions are F1, F2, ..., F m , if F(X1,X2,…,X m ) is the joint distribution function of d-dimensional random variables, then there exists a copula function C such that:

[0075] F(x1,x2,…,x m )=C(F1(x1),F2(x2),…,Fm(x m ))=C(u1,u2,…,u m )

[0076] Among them, θ is an undetermined parameter.

[0077] This embodiment focuses on the two-dimensional joint probability P of the cumulative continuous drought duration and the cumulative drought intensity sequence being greater than or equal to a certain value at the same time:

[0078] P(Dc≥d∩Sc≥s)=1-F D (d)-F S (s)+C DS (F D (d),F S (s))

[0079] Where, F D (d) means D c The one-dimensional distribution function of the sequence is F S (s) means S c The sequence obeys the one-dimensional distribution function, as shown in Table 1, C DS Indicates D c Sequence and S c The optimal Copula function obeyed by the sequence is shown in Table 2. The smaller P is, the smaller the probability of the corresponding drought event occurring in the past and the more extreme it is.

[0080] The study used the Kolmogorov Smirnov (KS) method to test the optimal one-dimensional distribution function of Dc and Sc sequences respectively (the minimum value of the KS statistic after passing the 0.05 confidence level).

[0081] Table 1 Marginal distribution function information of one-dimensional variables

[0082]

[0083] Commonly used two-dimensional copula functions include the Clayton Copula, the Gumbel-Hougaard Copula, and the Frank Copula. In Table 2, d = 2. Copula parameters were estimated using the maximum likelihood method. The optimal copula function was tested based on the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) statistics of the empirical and theoretical probabilities of random variables. The smaller the AIC and BIC statistics, the better the copula function fit.

[0084] Table 2 Multidimensional variable Copula function information

[0085]

[0086] Drought events have the characteristics of spreading and developing from point to surface and from mild to severe in space. Based on the joint probability P value obtained above, the percentile method is used to divide the severity levels into mild drought, moderate drought, severe drought, and extreme drought, so as to dynamically grasp the evolution process of drought in real time.

[0087] Based on the thresholds for different drought levels using the commonly used Standard Precipitation Index (SPI), it can be seen that among all water and drought natural disasters, the total proportion of drought, floods, and normal events is 100%. Among them, the proportion of mild drought is considered to be about 15%, moderate drought is about 9.2%, severe drought is about 4.4%, and extreme drought is about 2.3%. Then, taking drought events as the total, the proportion of mild drought is about 48.5% (15 / (15+9.2+4.4+2.3)), the proportion of moderate drought is about 29.8%, the proportion of severe drought is about 14.2%, and the proportion of extreme drought is about 7.5%. Figure 3 As shown in Table 3, based on the percentile method, the joint probability P is used as the research object and arranged in ascending order. The P values ​​corresponding to the 7.5%, 21.7%, and 51.5% percentiles are calculated, and the thresholds for mild, moderate, severe, and extremely severe levels in the development of regional drought are finally determined, as shown in Table 3.

[0088] Table 3 Proportion of different drought levels

[0089]

[0090] Land use data was reclassified into seven categories: forest land, grassland, cultivated land, paddy fields, water bodies, urban construction land, and unused land. The distribution of different drought levels within the study area was analyzed monthly as the drought progressed. The classified land use data was then overlaid with the data to calculate the area of ​​cultivated land affected by drought, enabling a quantitative assessment of agricultural drought risk.

[0091] Taking the drought event in a certain province in 2010 as an example, this example quantitatively assessed the scope of agricultural impact. The results are as follows: Figure 4 and as shown in Table 4.

[0092] Table 4 Agricultural area affected by different drought levels in a certain province in 2010 (km 2 )

[0093]

[0094]

[0095] Furthermore, in this example, a state space S is first defined, encompassing different drought levels, specifically S = {mild drought, moderate drought, severe drought, extreme drought}. This example systematizes the different drought stages, providing a clear framework for subsequent analysis. Each state not only reflects the severity of the current drought but also lays the foundation for drought risk assessment and management strategy development.

[0096] Next, based on historical drought event data, the transition probabilities between different drought levels are calculated to form a transition probability matrix P. In specific implementation, the transition probabilities between states are obtained by statistically analyzing the changes in drought levels in the historical data (in this example, the ratio of cases that transitioned from mild to moderate drought to the total number of mild drought cases). This matrix P will be used to describe the state transition characteristics of the system, thereby reflecting possible changes in drought levels in real time.

[0097] To analyze the impact of the current drought level, an initial state distribution vector π is determined based on the most recent drought-affected cultivated land area. This vector represents the probability distribution of each drought level under the current drought state. Specifically, this embodiment calculates the corresponding initial state distribution by analyzing the current drought-affected cultivated land area (e.g., the ratio of the total affected cultivated land area to the corresponding area of ​​each drought level). This step ensures that the model uses the most recent, accurate state data during analysis.

[0098] Using the transition probability matrix P and the initial state distribution vector π, a Markov process is applied to infer the distribution of future states. Specifically, the method uses weights to calculate the probability changes of each drought level under the current state, thereby generating a series of drought level predictions for future time steps. This embodiment ensures that the changing trends of future drought conditions can be quantified, facilitating the development of forward-looking management strategies.

[0099] Based on the generated series of future drought severity predictions, management strategies are developed that are appropriate for current and future conditions. This strategy implements targeted response measures (such as irrigation adjustments and crop selection) for different drought levels, ensuring that agricultural production stability is effectively improved under each drought scenario. Specific measures will be formulated based on the effectiveness of historical management plans and the economic value of crops to achieve optimal resource allocation.

[0100] Using a dynamic programming algorithm, based on the transition probability matrix P, the initial state distribution π, and the management strategy, the expected returns under different management strategies are gradually evaluated. By analyzing the returns of the strategies implemented in each state, the expected returns are compared with the potential risks. This step enables managers to identify the most economically effective strategies under different scenarios, providing a scientific basis for decision-making.

[0101] Finally, from all the evaluation results, the management strategy with the highest expected benefit under each state is selected to form a comprehensive drought management policy. By taking the optimal decision for each drought level, we ensure that management measures maximize economic benefits and resource utilization efficiency when faced with drought risks. This final strategy will be repeatedly parameterized to ensure its adaptability to the changing drought environment.

[0102] Furthermore, this embodiment implements a moderate irrigation strategy in mild drought conditions to allocate water resources and maintain soil moisture. This embodiment primarily relies on real-time monitoring of soil moisture, using soil moisture sensors combined with meteorological forecast data to assess soil moisture content and drought threats. Based on the monitoring results, a reasonable irrigation program is developed, using efficient irrigation methods such as drip irrigation or permeation irrigation to replenish water in a timely manner, ensuring that crop roots can still effectively absorb the required water under drought conditions. Furthermore, farmland management measures such as the use of mulch can reduce evaporation losses, thereby improving soil moisture retention and ensuring healthy crop growth.

[0103] During moderate droughts, irrigation frequency will be increased and farmers will be provided with crop adjustment recommendations to reduce water consumption. Specifically, by carefully assessing crop water needs, more frequent and precise irrigation plans will be developed to ensure adequate water for crops during critical growth periods. At the same time, agricultural consulting services will be provided, recommending drought-tolerant crops or adjusting current crop varieties to optimize water use efficiency. This can be achieved through the joint efforts of agricultural cooperatives and local governments, enabling timely dissemination of agricultural information and helping farmers develop more scientific water management strategies, thereby reducing water resource consumption.

[0104] In the case of severe and extreme droughts, strategies to retain soil moisture are implemented, and emergency response measures are taken for crops to reduce crop losses. For severe droughts, priority is given to covering the soil surface with mulch (such as straw or plastic film) to reduce soil moisture evaporation, while introducing drought-resistant crops to ensure that crops can survive and grow in a potential water shortage environment. In the case of extreme droughts, water resources management policies are fully implemented and emergency response mechanisms are activated to mitigate the impact of drought. This includes allocating water resources according to priority to ensure that the needs of important crops and drinking water are met, while organizing farmers and relevant institutions to carry out water resources planning and use training, and improving public awareness of drought and response capabilities to maintain the stability of agricultural production under persistent drought conditions.

[0105] Furthermore, this embodiment constructs a state-benefit table for the preset transition probability matrix and initial state distribution, and associates the management strategy under each drought state with its corresponding expected benefit. On this basis, starting from the current state, the state benefits of each future time step are gradually derived, and by traversing all possible state transitions, the expected benefit of each future state under a given management strategy is calculated. At this time, the optimal strategy for each state is dynamically updated in combination with historical data and model prediction information to ensure that the management strategy selected at each time node can maximize the benefits. Finally, the optimal decision results of all time steps are sorted out, and a decision plan containing the optimal management strategy for each drought level and its corresponding benefits is output to ensure that the plan is feasible and effective, thereby providing a scientific basis for the formulation of drought management policies.

[0106] Corresponding to the above method, this embodiment further provides a quantitative assessment system for agricultural drought risk taking into account the drought process, including:

[0107] The cumulative analysis unit is used to determine the cumulative duration and cumulative intensity of drought at all grid points in the study area using pre-set drought indicators;

[0108] A model building unit, configured to build a joint probability model of multidimensional drought characteristics using a Copula function based on the cumulative duration and the cumulative severity;

[0109] A threshold division unit, configured to divide drought conversion thresholds of different levels according to the joint probability model; the drought conversion thresholds are used to divide drought levels;

[0110] an area statistics unit, configured to combine preset land use type data with the drought conversion threshold to count drought-affected cultivated land areas at each drought level within the study area;

[0111] A policy optimization unit is used to establish a Markov decision process based on the drought level and optimize the drought management policy according to the Markov decision process.

[0112] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0113] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A quantitative assessment method for agricultural drought risk considering drought processes, characterized in that: include: The cumulative duration and cumulative intensity of drought at all grid points in the study area are determined using pre-set drought indicators; Based on the cumulative duration and the cumulative intensity, a joint probability model of multidimensional drought characteristics is constructed using a Copula function; Different levels of drought conversion thresholds are divided according to the joint probability model; the drought conversion thresholds are used to divide drought levels; Combining the preset land use type data with the drought conversion threshold to calculate the drought-affected cultivated land area under each drought level in the study area; A Markov decision process is established based on the drought level, and a drought management policy is optimized according to the Markov decision process.

2. The method for quantitatively assessing agricultural drought risk considering drought processes according to claim 1, characterized in that: The cumulative duration and cumulative intensity of drought at all grid points in the study area are determined using pre-set drought indicators, including: Determining the drought event persistence rule of the study area according to the preset drought index; According to the drought event persistence rule, the cumulative duration and cumulative intensity of drought in the study area are calculated respectively; the cumulative duration D c The calculation formula is D c = c; where c represents the cumulative drought duration from the time scale of the beginning of a drought to a certain moment in the development process; the cumulative intensity S c The calculation formula is: if index≤drought threshold; where index is the quantitative value of the drought index.

3. The method for quantitatively assessing agricultural drought risk considering drought processes according to claim 2, characterized in that: The drought event persistence rules include: At the i-th time scale, if the quantitative value of the drought indicator is less than the drought threshold, it is considered that drought has occurred in the study area and the drought duration begins to accumulate; When drought does not occur at the i+1 time scale but occurs at the i+2 time scale, considering that drought recovery takes time, the i-th time scale, the i+1-th time scale, and the i+2 time scale are considered to be continuously developing drought events.

4. The method for quantitatively assessing agricultural drought risk considering drought processes according to claim 2, characterized in that: The expression of the joint probability model is: P(Dc≥d∩Sc≥s)=1-F D (d)-F S (s)+C DS (F D (d),F S (s)) Where P(Dc≥d∩Sc≥s) represents the joint probability that the cumulative drought duration Dc is greater than or equal to the drought duration limit d and the cumulative drought intensity Sc is greater than or equal to the drought intensity limit s; F D (d) means D c The one-dimensional distribution function of the sequence is F S (s) means S c The sequence obeys the one-dimensional distribution function, C DS Indicates D c Sequence and S c The optimal Copula function that the sequence obeys.

5. The method for quantitatively assessing agricultural drought risk considering drought processes according to claim 1, characterized in that: The method for classifying the drought levels includes: When the drought conversion threshold P>P 51.5％ When the drought level is determined to be mild drought; When P 21.7％ <P≤P 51.5％ When the drought level is determined to be moderate drought; When P 7.5％ <P≤P 21.7％ When the drought level is determined to be severe drought; When P≤P 7.5％ When the drought level is determined to be extreme drought.

6. The method for quantitatively assessing agricultural drought risk considering drought processes according to claim 1, characterized in that: The land use type data include: forest land, grassland, cultivated land, paddy field, water area, urban construction land and unused land.

7. The method for quantitatively assessing agricultural drought risk considering drought processes according to claim 5, characterized in that: A Markov decision process is established based on the drought level, and a drought management policy is optimized according to the Markov decision process, including: A state space S is defined, wherein the state space S includes different drought levels; S={mild drought, moderate drought, severe drought, extreme drought}; Based on historical drought event data, the probability of transitioning from one drought level to another is calculated to form a transition probability matrix P; Determine an initial state distribution vector π based on the most recent drought-affected cultivated land area; the initial state distribution vector π represents the probability of each drought level in the current state; Based on the transition probability matrix P and the initial state distribution π, the Markov process is applied to deduce the state distribution of future time steps to generate a prediction sequence of future drought levels; Formulate management strategies appropriate to current and expected future conditions based on the forecast sequence; Using a dynamic programming algorithm, based on the transition probability matrix P, the initial state distribution π and the management strategy, to evaluate the expected benefits under different management strategies; The management strategy with the largest expected benefit in each state is selected as the drought management policy.

8. The method for quantitatively assessing agricultural drought risk considering drought processes according to claim 7, characterized in that: The calculation formula of the dynamic programming algorithm is: Among them, V(s t ) is the current state s t The value of, that is, the expected benefit of the state, max at Indicates that among all the optional strategies a t Choose the value V(s t ) maximizing strategy, R(s t ,a t ) is in state s t Take strategy a t The immediate profit at the time of t+1 ∣s t ,a t ) is from state s t Go to the next state s t+1 The transition probability, given in state s t Take strategy a t ,V(s t+1 ) is the next state s t+1 value.

9. The method for quantitatively assessing agricultural drought risk considering drought processes according to claim 7, characterized in that: The management strategies include: In the case of mild drought, implement moderate irrigation and allocate water resources to maintain soil moisture; In moderate drought conditions, increase irrigation frequency and provide crop adjustment advice to farmers to reduce water consumption; In severe drought conditions, implement soil moisture conservation strategies and introduce drought-tolerant crops to reduce crop losses; In the face of severe drought, water resource management policies are fully implemented and emergency response measures are initiated to mitigate the impact.

10. A quantitative assessment system for agricultural drought risk taking into account drought processes, characterized in that: include: The cumulative analysis unit is used to determine the cumulative duration and cumulative intensity of drought at all grid points in the study area using pre-set drought indicators; A model building unit, configured to build a joint probability model of multidimensional drought characteristics using a Copula function based on the cumulative duration and the cumulative severity; A threshold division unit, configured to divide drought conversion thresholds into different levels according to the joint probability model; The drought conversion threshold is used to classify drought levels; an area statistics unit, configured to combine preset land use type data with the drought conversion threshold to count drought-affected cultivated land areas at each drought level within the study area; A policy optimization unit is used to establish a Markov decision process based on the drought level and optimize the drought management policy according to the Markov decision process.

Images