A quantitative assessment method for drought disaster risk integrating irrigation disaster reduction effects

Abstract

This invention discloses a quantitative drought disaster risk assessment method that integrates irrigation disaster mitigation effects. This method, which belongs to the field of disaster risk assessment, includes the following steps: acquiring different meteorological data at different scales of a three-dimensional spatial grid in a target region and calculating a drought index; identifying drought events using the ST-KMeans clustering method; screening and merging drought events, calculating the spatial overlap between clustered drought regions and historical disaster-affected regions, and integrating risk loss data; generating a dynamic dataset of drought event characteristics to measure drought events; and extracting irrigation response information as a drought adaptation measure, combining land cover and utilization data and farmland irrigated area data in the target region, to quantify irrigation disaster mitigation capacity and conduct risk assessment. This method integrates irrigation information with drought event identification to quantitatively assess the drought mitigation effects of irrigation at different temporal and spatial scales, providing a scientific basis for drought disaster risk management and agricultural adaptation policies.

Description

Technical Field

[0001] The present invention relates to the technical field of disaster risk assessment, and in particular to a drought disaster risk quantitative assessment method integrating irrigation disaster reduction effects. Background Art

[0002] Drought, a major meteorological disaster, is becoming more frequent and more frequent in the context of global warming, posing a persistent threat to agricultural production systems. As a widely used drought adaptation measure, assessing the mitigation effects of irrigation is crucial for optimizing water resource allocation and improving drought resilience. It can mitigate crop yield losses to a certain extent. Therefore, disaster risk assessments based on irrigation regulation responses are urgently needed to provide technical support for agricultural disaster mitigation, water resource allocation, and climate adaptation policies.

[0003] However, existing technologies for assessing drought risk still face the following challenges: 1) Current assessments of irrigation disaster mitigation effectiveness often rely on macroscopic or empirical statistical methods, lacking multi-source data integration and mechanism-driven approaches, making it difficult to reveal the specific intervention effects of irrigation on the evolution of drought events. 2) Existing technical approaches lag behind the demand for spatiotemporal clustering analysis of drought events. Although three-dimensional clustering algorithms and drought volume structure methods have been used for drought event identification, they have not yet been effectively coupled with the mitigation effects of irrigation measures. 3) Existing disaster loss models lack the ability to dynamically characterize the response to irrigation regulation. Mainstream quantitative models (such as linear regression and Copula functions) are often based on static parameter assumptions and fail to consider the dynamic feedback loop between irrigation measures and the spatiotemporal evolution of drought. In particular, for large-scale, long-term drought monitoring and disaster assessment, there is a lack of a systematic analytical approach that integrates the spatiotemporal evolution of drought, the distribution characteristics of irrigation measures, and their interactions.

[0004] In view of the above situation, there is an urgent need to develop a quantitative assessment method for drought disaster risk that integrates the irrigation disaster reduction effect, which can integrate irrigation information on the basis of drought event identification and quantitatively evaluate the drought mitigation effect of irrigation at different time and spatial scales, thereby providing a scientific basis for drought disaster risk management and agricultural adaptation policies. Summary of the Invention

[0005] The problem to be solved by the present invention is to provide a quantitative assessment method for drought disaster risk that integrates the irrigation disaster reduction effect, quantitatively evaluate the regulatory impact of irrigation on drought disaster prevention and reduction, and provide scientific and technological support for precise drought prevention and assessment and mitigation of the adverse effects of socioeconomic drought.

[0006] The present invention adopts the following technical solution: a quantitative assessment method for drought disaster risk integrating irrigation disaster reduction effects, comprising the following steps:

[0007] Step 1: Obtain drought index data: Obtain different meteorological data at three-dimensional spatial grid scales within the target area during the monitoring period, perform standardization, and calculate the drought index;

[0008] Step 2: Spatiotemporal clustering of drought events: Use the ST-KMeans clustering method to perform dynamic detection of drought event clusters and identify drought events;

[0009] Step 3: Risk loss data integration and drought event index construction: Drought events are screened and merged. The spatial overlap between clustered drought areas and historical disaster areas is calculated based on historical drought disaster data to integrate risk loss data. The retained drought events are numbered and recorded in temporal and spatial order. The characteristic variables are calculated using the intensity-area-duration method to generate a dynamic dataset of drought event characteristics and measure drought events.

[0010] Step 4: Extracting irrigation response information: Based on the obtained drought event characteristic dynamic dataset, combined with the target area land cover and utilization data and farmland irrigation area data, extract irrigation response information as a drought adaptation measure;

[0011] Step 5: Drought risk assessment and quantification of irrigation disaster reduction effects: Based on the historical drought disaster data from the disaster risk survey and the county boundary vector data, the effectiveness of irrigation disaster reduction capacity and risk assessment are quantified based on the irrigation response information.

[0012] Preferably, in step 1, the standardization process unifies the numerical range of the acquired meteorological data to [0, 1] or [-1, 1], so that the meteorological data becomes dimensionless data.

[0013] The drought index is based on standardized meteorological data. A three-dimensional sliding window is used to remove outliers and smooth the data in the longitude, latitude, and time dimensions. The smoothed data is then interpolated to fill in missing values, resulting in the pre-processed three-dimensional grid SAPEI drought index data. The calculation formula is as follows:

[0014] ;

[0015] Where, 、 、 、 、 、 is a constant term;

[0016] Intermediate variables The calculation formula is as follows:

[0017] ;

[0018] Where, is the cumulative probability density, expressed as follows:

[0019] ;

[0020] Where, is the probability distribution function, and the formula is as follows:

[0021] ;

[0022] Where, 、 、 are the scale, shape and initial state parameters, respectively, which are obtained by linear moment method fitting.

[0023] Preferably, in step 2, the ST-KMeans clustering method is used to identify drought events, including the following sub-steps:

[0024] Step 2.1. Determine the drought index level and corresponding SAPEI value of the selected core points based on the SAPEI drought index classification, and obtain a three-dimensional point set containing all drought pixels in the format of [time index t, latitude lat, longitude lon, SAPEI value];

[0025] Step 2.2: Construct a spatiotemporal feature sample set. Convert each drought pixel into a feature vector: [x position, y position, time index t, drought intensity]. The feature vectors of all drought pixels form a sample set, where x and y are latitude and longitude or grid index, and t is the day number.

[0026] Step 2.3, normalize the temporal and spatial three-dimensional features of the spatiotemporal feature sample set respectively;

[0027] Step 2.4: Aggregate the standardized drought grids into several event clusters, each cluster representing a drought event;

[0028] Step 2.5: Determine the number of event clusters. Use ST-KMeans, add temporal continuity or spatial connectivity constraints, and use three-dimensional Euclidean distance to determine the distance function of spatiotemporal clustering. Adjust the weight of the time dimension according to the research focus to obtain several drought event clusters and form a three-dimensional drought event dataset.

[0029] Preferably, in step 3, the risk loss data docking includes the following sub-steps:

[0030] Step 3.1, screening drought events: If the drought patch meets any drought event screening criteria, the patch is deemed invalid;

[0031] Criterion 311: Determine whether the minimum duration of a drought event meets the set time threshold requirements and filter out short-term or transient drought signals;

[0032] Criterion 312: Determine whether the spatial coverage area of ​​the drought patch is greater than the set patch area threshold. If so, the drought patch is considered invalid and the local small-scale drought is excluded. Otherwise, the drought patch is retained.

[0033] Guideline 313: Screen for drought events that evolve continuously in time and space, including: gradual migration or diffusion of drought centers and exclusion of isolated patches.

[0034] Step 3.2, merging drought events: Based on the patch screening results, drought patches are merged. Based on the reference patch, other patches are traversed in temporal and spatial order to determine whether they meet the patch merging criteria:

[0035] Criterion 321: If the time interval between two drought patches is less than a preset interval threshold, merge them; the interval threshold is adjusted based on the climate characteristics of the target area and the typical duration of drought events;

[0036] Guideline 322: If drought patches partially overlap in time, indicating that the drought events are continuous, they should be directly merged;

[0037] Rule 323: If the spatial overlap of drought patches exceeds a preset ratio, they are considered to be a continuation of the same event and should be merged;

[0038] Guideline 324: If the drought intensity changes between drought patches show a continuous evolution, gradually increasing or decreasing, they should be merged.

[0039] Step 3.3, Risk loss data docking: Convert the spatiotemporal extent of each drought event recorded in the historical drought disaster data into a 0.25° grid format consistent with the SAPEI data, mark the affected areas, 1 for affected and 0 for non-affected, and calculate the spatial overlap and temporal matching rate of the clustered drought areas and the historical drought disaster areas.

[0040] Preferably, in step 3, a drought index is constructed based on the following drought characteristics to measure drought events:

[0041] Feature 1. Drought duration : refers to the number of time steps between the start time of drought and the end time of drought, which is used to indicate the duration of drought;

[0042] Feature 2: Drought severity : refers to the absolute value of the accumulated SAPEI values ​​of all drought grids during the drought event;

[0043] Feature 3: Drought intensity : refers to the average SAPEI value of all grids during the drought period, which is the severity of drought and drought duration The ratio of

[0044] Feature 4: Drought Area : Refers to the maximum scope of a drought event, expressed as the maximum area that was in a drought state during the drought event, which is the sum of the areas of all grids when the coverage area is the largest.

[0045] Preferably, in step 4, extracting irrigation response information includes the following sub-steps:

[0046] Step 4.1: Process the spatiotemporal range of annual land cover and land use data and farmland irrigation area data into a 0.25° grid format consistent with the drought event data;

[0047] Step 4.2: Screen land use types that may cause direct economic losses due to drought, including cultivated land, economic forest land, and grasslands of varying degrees of coverage. Different land use types are assigned different weight coefficients based on economic output and vulnerability to obtain the first land use weight matrix.

[0048] Step 4.3: Based on the spatial correspondence, the high-resolution annual irrigation area data of the region is statistically mapped to the corresponding 0.25° grid to obtain the second weight matrix of irrigation area. The first weight matrix and the second weight matrix are multiplied at the corresponding time and space grid points to obtain the third weight matrix.

[0049] Step 4.4: Extract the drought intensity and other characteristics of all drought events in the drought event feature dynamic dataset to obtain a first hazard factor intensity matrix. Multiply the first hazard factor intensity matrix and the third weight matrix accordingly to obtain a second hazard factor intensity matrix weakened by disaster prevention and mitigation capabilities.

[0050] Step 4.5: Sum the second hazard factor intensity matrix within the year to obtain the response information of irrigation drought disaster reduction.

[0051] Preferably, in step 5, the drought risk assessment and irrigation disaster reduction effect quantification include the following sub-steps:

[0052] Step 5.1: Reselect and redistribute the second hazard factor intensity matrix based on the administrative boundaries of each county. Calculate the weight of the intensity value of the second hazard factor intensity matrix at each grid point within the county relative to the sum of the intensities of all grid points within the county. This yields the fourth weight matrix for the redistribution of the total economic loss within the county.

[0053] Step 5.2: Multiply the fourth weight matrix by the county's total annual economic losses in time and space to obtain drought economic loss data in raster format, and perform data verification at the spatial and provincial / municipal scales;

[0054] Step 5.3: Fit the grid-based drought economic losses with a logistic function to obtain the first drought economic vulnerability curve considering irrigation adaptability measures, and assess the risks of different drought intensities under the condition of considering irrigation adaptability measures;

[0055] Step 5.4: Remove the irrigation weight from the obtained second hazard intensity matrix to obtain the original third hazard intensity matrix. The disaster reduction effect of irrigation is not considered when calculating the redistribution weights.

[0056] Step 5.5: Calculate the weight of the intensity value of the third hazard factor intensity matrix at each grid point in the county relative to the sum of the intensities of all grid points in the county, and obtain the fifth weight matrix for redistribution of economic loss grid points in the county;

[0057] Step 5.6: Multiply the fifth weight matrix by the county's annual total economic losses to obtain drought economic loss data in raster format for data validation at the spatial and provincial / municipal scales.

[0058] Step 5.7: Fit the grid-based drought economic losses by constructing a spatial dynamic explicit model to obtain the second drought economic vulnerability curve without considering irrigation adaptation measures, and assess the risk of different drought intensities without considering adaptation measures;

[0059] Step 5.8: Under the condition of the same intensity of the hazard factors, the loss value of the second vulnerability curve is subtracted from the loss value of the first vulnerability curve to obtain the quantitative value of the disaster reduction effect of irrigation as a drought adaptation measure.

[0060] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0061] 1. The quantitative drought disaster risk assessment method of the present invention can integrate irrigation information based on drought event identification, and quantitatively evaluate the regulatory impact of irrigation on drought disaster prevention and mitigation at different temporal and spatial scales, thereby providing scientific and technological support for precise drought prevention and the assessment and mitigation of the adverse effects of socioeconomic drought. It is of great significance for the formulation of relevant response policies under future climate change, drought prevention and mitigation, and ecological and environmental protection.

[0062] 2. The quantitative assessment method for drought disaster risk of the present invention, by introducing a three-dimensional spatiotemporal clustering algorithm (ST-KMeans method), effectively overcomes the problem that traditional two-dimensional static indicators are insufficient in capturing the continuity and evolution trajectory of drought processes, thereby improving the accuracy and completeness of drought event identification. At the same time, combined with irrigation distribution data from remote sensing inversion or statistical surveys, it realizes the identification and classification of irrigated and non-irrigated areas at the drought event level, effectively improving the spatial accuracy of irrigation intervention analysis and avoiding the estimation errors caused by traditional administrative district-based statistical averaging methods.

[0063] 3. The quantitative assessment method for drought disaster risk of the present invention constructs a response function model with drought intensity, duration, and spatial impact area as input variables and disaster losses as output variables. It can reveal the comprehensive effects of drought-causing factors and realize loss prediction and sensitivity analysis at multiple scales. It further quantifies the loss mitigation effect of irrigation measures under drought conditions based on the drought event comparison mechanism under "irrigation-no irrigation" conditions, thus realizing a process-based and quantitative assessment of irrigation regulation capacity. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 This is the overall flow chart of the drought disaster risk quantitative assessment method of the present invention;

[0065] Figure 2 This is the drought event identification step based on the clustering algorithm of the present invention;

[0066] Figure 3 This is a schematic diagram of the relationship between drought intensity and economic losses according to an embodiment of the present invention;

[0067] Figure 4 This is a diagram showing the verification results of the irrigation disaster reduction effectiveness according to an embodiment of the present invention;

[0068] Figure 5 This is a diagram of drought disaster risk assessment and effectiveness quantification results according to an embodiment of the present invention. DETAILED DESCRIPTION

[0069] In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions of the application are further elaborated in detail below with reference to the accompanying drawings. The described embodiments are only a part of the embodiments involved in the present invention. All non-innovative embodiments of other researchers in this field on this embodiment fall within the scope of protection of the present invention. At the same time, the step numbers in the embodiments of the present invention are only set for the convenience of explanation and description, and the order between the steps is not limited in any way. The execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0070] In one embodiment of the present invention, a quantitative assessment method for drought disaster risk integrating irrigation disaster reduction effect is provided. Figure 1 As shown, it mainly includes the following parts:

[0071] 1. Obtain drought index data within the monitoring period;

[0072] 2. Dynamic detection of drought events by spatiotemporal clustering;

[0073] 3. Connect risk loss data and construct drought event indicators;

[0074] 4. Superimpose irrigation distribution data and extract irrigation response information;

[0075] 5. Quantification of irrigation disaster reduction effects and drought risk assessment.

[0076] The specific contents are as follows:

[0077] 1. Obtain three-dimensional spatial index grid data within the monitoring time

[0078] (1) Data standardization

[0079] Since the meteorological elements have different sources and different measurement units, in order to conveniently incorporate different meteorological elements into the same calculation process, it is necessary to standardize the acquired meteorological data and unify its numerical range to [0,1] or [-1,1], so that the meteorological data becomes dimensionless data.

[0080] In this embodiment, meteorological data of different three-dimensional spatial grid scales within the monitoring time range of the target area are obtained, and drought index data are calculated based on these meteorological data.

[0081] The obtained meteorological data is normalized, and the formula is as follows:

[0082] ;

[0083] ;

[0084] ;

[0085] in, Represents data after normalization or standardization. Represents the original data, represents the mean, represents variance; Indicates the maximum and minimum original data.

[0086] (2) Calculation of drought index SAPEI

[0087] Drought index data is calculated based on the standardized meteorological data. The drought index here is SAPEI (i.e., the precipitation evapotranspiration index before standardization). Then, a three-dimensional sliding window is used to smooth the data in the longitude, latitude, and time dimensions to remove outliers. In addition, the smoothed data is interpolated and filled to fill in the missing values ​​in the data. Finally, the preprocessed three-dimensional grid SAPEI drought index data is obtained.

[0088] Specifically, the calculation formula of SAPEI is as follows:

[0089] ;

[0090] In this embodiment, the constant term 、 、 、 、 、 ;

[0091] Intermediate variables The calculation formula is as follows:

[0092] ;

[0093] Where, is the cumulative probability density, as follows:

[0094] ;

[0095] Where, is the probability distribution function, and the formula is as follows:

[0096] ;

[0097] Where, 、 、 are scale, shape and initial state parameters respectively. These three parameters can be obtained by linear moment method fitting. The specific formula is as follows:

[0098] ;

[0099] ;

[0100] ;

[0101] in, for Gamma function, , , yes Probability-weighted moments of sequences;

[0102] ;

[0103] Where, is the precipitation evapotranspiration series The ordinal number of the sequence in ascending order, ranging from 1, 2, ..., n; is the sample length of the time series. The precipitation evapotranspiration difference of a month is the sum of the precipitation evapotranspiration difference of the previous k-1 months and the current month as follows:

[0104] ;

[0105] ;

[0106] Where, For different time scales, is the monthly precipitation, is the monthly potential evaporation, is an intermediate variable.

[0107] 2. Dynamic detection of drought event clustering

[0108] The ST-KMeans clustering method is used to identify drought events. According to the SAPEI drought index classification, the drought index level and corresponding value of the selected core points are determined. In this embodiment, , and obtain a 3D point set containing all drought pixels in the format of [time index t, latitude lat, longitude lon, SAPEI value].

[0109] Furthermore, in order to construct a spatiotemporal feature sample set, each drought pixel is converted into a feature vector form: [x position, y position, time index t, drought intensity], and the feature vectors of all drought pixels are combined into a sample set, where x and y use latitude and longitude or grid index, and t is the day number.

[0110] In order to prevent the time dimension from suppressing the space, the time and space three-dimensional features of the sample set are normalized separately. The normalization formula is as follows:

[0111] ;

[0112] ;

[0113] Where, and Represent the original values ​​in space and time respectively, represents the mean of the original data, is the standard deviation of the data.

[0114] The standardized drought grids are then aggregated into several "event clusters," each representing a single drought event. The number of clusters is determined automatically by trying various preset K values ​​(e.g., K = 50, 80, 100) or by using methods such as the Silhouette coefficient and the Elbow method.

[0115] Then, ST-KMeans is used to add temporal continuity or spatial connectivity constraints. The distance function design for spatiotemporal clustering is determined by three-dimensional Euclidean distance. At the same time, the weight of the time dimension is adjusted according to the research focus (for example, increasing the influence of time in the distance function). Finally, several drought event clusters are obtained, namely the three-dimensional drought event dataset. The specific process is as follows: Figure 2 shown.

[0116] The three-dimensional Euclidean distance is calculated as follows:

[0117] ;

[0118] Where, and represent the values ​​in space and time respectively, Represents the weight of the time dimension.

[0119] In particular, in this embodiment, the SAPEI drought index is divided into the following levels:

[0120] Level 4: Extremely wet, ;

[0121] Level 3: Severely wet, ;

[0122] Level 2: Medium moist, ;

[0123] Level 1: slightly moist, ;

[0124] Level 0: Normal, ;

[0125] Level-1: Mild drought, ;

[0126] Level-2: Moderate drought, ;

[0127] Level-3: severe drought, ;

[0128] Level -4: Extreme drought, .

[0129] 3. Risk loss data docking and indicator construction

[0130] (1) Screening and merging drought events

[0131] The resulting 3D drought event dataset was screened and merged in time and space. Drought event screening followed the same criteria as in the technical proposal. A drought patch was considered invalid if it met any of these criteria. Patch merging followed the same criteria as above.

[0132] Drought events were screened according to the following guidelines:

[0133] Criterion 1: Determine whether the minimum duration of an event (e.g., t ≥ 15 days) meets the requirements to filter out short-term or transient drought signals. The minimum duration of drought can be set based on actual conditions to better meet requirements.

[0134] Criterion 2: Determine whether the drought patch spatial coverage threshold is greater than a set patch area threshold (e.g., ≥100 square kilometers). If so, the drought patch is considered invalid and the localized small-scale drought is excluded. Otherwise, it is retained. Similarly, the patch area threshold can be set according to actual conditions.

[0135] Criterion 3: Screen drought events that evolve continuously in time and space (such as the gradual migration or spread of the drought center) and exclude isolated patches.

[0136] To determine whether drought patches meet the merging requirements, it is necessary to merge drought patches based on the completion of patch screening. Based on the reference patch, other patches are traversed in temporal and spatial order to determine whether they meet the patch merging criteria. The merging criteria for drought events are as follows:

[0137] Criterion 1: If the time interval between two drought patches (e.g., the difference between the end and start times) is less than a preset threshold (e.g., 5 days), they can be merged. This threshold needs to be adjusted based on the climate characteristics of the target area and the typical duration of drought events.

[0138] Criterion 2: If drought patches partially overlap in time, it indicates that the drought events are continuous and can be directly merged.

[0139] Criterion 3: If the spatial overlap of drought patches exceeds a certain proportion (e.g., 30%), even if the time interval is short, it may be a continuation of the same event.

[0140] Criterion 4: If the changes in drought intensity (such as SAPEI values) between drought patches show continuous evolution (such as gradual worsening or alleviation) rather than sudden changes, patch merging is supported.

[0141] (2) Risk loss data connection

[0142] Historical drought disaster data include the time, location (county) and specific direct economic losses caused by each drought event. The events and locations of drought disasters are required here.

[0143] The filtered and merged drought events were calibrated based on actual drought disasters recorded in the historical drought data from the disaster risk survey. The spatiotemporal extent of each drought event recorded in the historical drought data from the disaster risk survey was converted to a raster format consistent with the SAPEI data. Affected areas were labeled (1 for affected, 0 for unaffected). The spatial overlap (IoU) between the clustered drought areas and the historical disaster areas was calculated. In this example, the target IoU was ≥ 0.7. The temporal matching ratio between the two data sets was also calculated.

[0144] (3) Construction of drought indicators

[0145] The retained drought events were numbered and recorded in temporal and spatial order. Characteristic variables were then calculated using the Intensity-Area-Duration (IAD) method, following the order of the numbers. The calculated indicators were then saved according to their corresponding numbers to generate a dynamic dataset of drought event characteristics, which was used to verify and quantify the effectiveness of irrigation as an adaptive measure for disaster prevention and mitigation.

[0146] The indicators mainly include: drought duration, drought severity, drought intensity and drought area, as follows:

[0147] Feature 1: Drought duration , refers to the duration of drought, that is, the number of time steps between the start and end of the drought.

[0148] Feature 2: Drought severity , refers to the absolute value of the accumulated SAPEI values ​​of all drought grids during the drought event, and the formula is as follows:

[0149] ;

[0150] Where, represents the duration of a drought event, Indicates the SAPEI value at the i-th time step (e.g. day).

[0151] Feature 3: Drought intensity , refers to the average SAPEI value of all grids during the drought period, which is the severity of drought and drought duration The ratio is as follows:

[0152] ;

[0153] Feature 4: Drought Area It refers to the maximum scope involved in a drought event, that is, the maximum area that was in a drought state during the drought event, and the sum of the areas of all grids when the coverage area is the largest.

[0154] In this embodiment, the sample data results and the sum of economic losses identified and retained through the above steps show basically the same trend in spatial distribution, that is, generally speaking, the greater the drought intensity, the higher the economic losses. The relationship between drought intensity and economic losses is as follows: Figure 3 As shown. Figure 3 As can be seen, as the absolute value of the drought index increases, economic losses show a clear upward trend. The verified coefficient of determination reaches 0.54, indirectly verifying the effectiveness and reliability of the drought event identification algorithm used in this example. This conclusion provides an important scientific basis for the subsequent optimization of drought monitoring technology and the development of disaster response strategies.

[0155] 4. Irrigation response information extraction

[0156] Combined with land cover and land use data and irrigated farmland area data, the resulting dynamic dataset of drought event characteristics was used to extract information on the response of irrigation as a drought adaptation measure. The spatiotemporal extent of the annual land cover and land use data and irrigated farmland area data needed to be pre-processed to a 0.25° grid format consistent with the drought event data.

[0157] First, we screened the land use types that may cause direct economic losses due to drought. The selected land use types mainly included cultivated land (such as paddy fields and dry land), economic forest land (such as nurseries, orchards, tea gardens and other types of gardens) and grasslands with different coverage levels (such as natural grasslands and improved grasslands).

[0158] Because droughts have the most direct and severe impact on agriculture, this example uses a 1km radius of land use types, primarily including cultivated land, economic forestland, and grasslands of varying degrees of coverage. Furthermore, spatial interpolation is required to convert the temporal and spatial extent of annual land cover and land use data into a raster format consistent with the drought event data.

[0159] On this basis, different land use types are assigned different weight coefficients according to economic output value and vulnerability, and the land use weight matrix (first weight matrix) is obtained, in which the weight coefficient values ​​can be changed according to actual conditions.

[0160] Secondly, based on spatial correspondence, it is necessary to count the high-resolution irrigation area data of each year into the corresponding 0.25° grid to obtain the irrigation weight matrix (second weight matrix). The first weight matrix and the second weight matrix are multiplied correspondingly at the time and space grid points to obtain the third weight matrix.

[0161] Then, the drought intensity and other characteristics of all drought events in the drought event characteristic dynamic dataset are extracted to obtain the disaster factor intensity matrix (first intensity matrix). The first intensity matrix and the third weight matrix are multiplied correspondingly to obtain the disaster factor intensity matrix weakened by disaster prevention and mitigation capabilities (second intensity matrix). Finally, the second intensity matrix is ​​summed within the year to extract the effectiveness information of irrigation on drought mitigation, such as Figure 4 shown.

[0162] 5. Risk assessment and quantification of mitigation effects

[0163] Combining historical drought data from the disaster risk survey with county boundary vector data, we quantify the effectiveness of irrigation disaster mitigation capacity and assess its risk based on the intensity matrix of hazard factors weakened by disaster prevention and mitigation capabilities, as obtained in step 1. As mentioned previously, historical drought data includes the time, location (county), and specific direct economic losses caused by each drought event. Here, we only need to consider the total annual economic losses at the county level caused by drought.

[0164] First, the second intensity matrix is ​​re-screened and distributed based on the administrative boundaries of each county. The weight of the intensity value of the second intensity matrix at each grid point in the county is calculated separately to the sum of the intensities of all grid points in the county, and the final redistribution weight ratio of the total economic loss grid point in the county is obtained (the fourth weight matrix).

[0165] Then, the fourth weight matrix is ​​multiplied by the county's annual total economic losses in time and space to obtain the annual drought economic loss data in raster format, and the data are verified at the spatial and provincial and municipal scales.

[0166] Finally, the grid-based drought economic losses were fitted with a logistic function to obtain the drought economic vulnerability curve (the first vulnerability curve) that takes into account irrigation adaptation measures. Based on this vulnerability curve, the risk of different drought intensities under the condition of considering irrigation adaptation measures was assessed.

[0167] At the same time, the irrigation weight is removed from the second intensity matrix to obtain the original hazard intensity matrix (the third intensity matrix). This means that the disaster reduction effect of irrigation is not considered when calculating the redistribution weights. Subsequently, as described above, the weight of the intensity value of the third intensity matrix at each grid point within the county is calculated by adding the sum of the intensity values ​​of all grid points within the county to obtain the redistribution weight ratio at the economic loss grid point within the county (the fifth weight matrix).

[0168] Then, the fifth weight matrix is ​​multiplied by the county's annual total economic losses to obtain the annual drought economic loss data in raster format, which is also verified at the spatial and provincial and municipal scales.

[0169] Next, we fitted the grid-based drought economic losses with a spatially dynamic explicit model to generate a drought economic vulnerability curve (the second vulnerability curve) that does not consider irrigation adaptation measures. Based on this vulnerability curve, we assessed the risk of different drought intensities without considering adaptation measures.

[0170] Finally, under the condition of the same intensity of the hazard factors, the loss value of the second vulnerability curve minus the loss value of the first vulnerability curve can be used to obtain the quantitative value of the disaster reduction effectiveness of irrigation as a drought adaptation measure. Figure 5 shown.

[0171] Among them, the spatial dynamic explicit model has the following functional form:

[0172] ;

[0173] Where, express economic losses incurred; , ) represents the location of the i-th drought core area; represents the drought intensity of the i-th drought core area; represents spatial weight, reflecting regional economic density or vulnerability; Represents the impact radius of the drought event.

[0174] In summary, the quantitative assessment method of drought disaster risk that integrates the irrigation disaster reduction effect of the present invention can integrate irrigation information on the basis of drought event identification, and quantitatively evaluate the regulatory impact of irrigation on drought disaster prevention and reduction at different time and spatial scales, thereby providing scientific and technological support for precise drought prevention and the assessment and reduction of the adverse effects of socioeconomic drought. It is of great significance to the formulation of relevant response policies, drought resistance and disaster reduction, and ecological and environmental protection under future climate change.

[0175] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations that fall within the meaning and range of equivalents of the claims be included within the present invention. Any reference numerals in the claims should not be construed as limiting the claims to which they relate. Thus, the present invention is intended to encompass such modifications and variations as fall within the scope of the claims and their equivalents.

Claims

1. A quantitative assessment method for drought disaster risk integrating irrigation disaster reduction effects, characterized by: The steps include: Step 1: Obtain drought index data: Obtain different meteorological data at three-dimensional spatial grid scales within the target area during the monitoring period, perform standardization, and calculate the drought index; Step 2: Spatiotemporal clustering of drought events: Use the ST-KMeans clustering method to perform dynamic detection of drought event clusters and identify drought events; Step 3: Risk loss data integration and drought event index construction: Drought events are screened and merged. The spatial overlap between clustered drought areas and historical disaster areas is calculated based on historical drought disaster data to integrate risk loss data. The retained drought events are numbered and recorded in temporal and spatial order. The characteristic variables are calculated using the intensity-area-duration method to generate a dynamic dataset of drought event characteristics and measure drought events. The risk loss data docking includes the following sub-steps: Step 3.1, screening drought events: If the drought patch meets any drought event screening criteria, the patch is deemed invalid; Criterion 311: Determine whether the minimum duration of a drought event meets the set time threshold requirements and filter out short-term or transient drought signals; Criterion 312: Determine whether the spatial coverage area of ​​the drought patch is greater than the set patch area threshold. If so, the drought patch is considered invalid and the local small-scale drought is excluded. Otherwise, the drought patch is retained. Guideline 313: Screen for drought events that evolve continuously in time and space, including: gradual migration or diffusion of drought centers and exclusion of isolated patches; Step 3.2, merging drought events: Based on the patch screening results, drought patches are merged. Based on the reference patch, other patches are traversed in temporal and spatial order to determine whether they meet the patch merging criteria: Criterion 321: If the time interval between two drought patches is less than a preset interval threshold, merge them; the interval threshold is adjusted based on the climate characteristics of the target area and the typical duration of drought events; Guideline 322: If drought patches partially overlap in time, indicating that the drought events are continuous, they should be directly merged; Rule 323: If the spatial overlap of drought patches exceeds a preset ratio, they are considered to be a continuation of the same event and should be merged; Guideline 324: If the drought intensity changes between drought patches show a continuous evolution, gradually increasing or easing, they should be merged; Step 3.3, Risk Loss Data Interfacing: Convert the spatiotemporal extent of each drought event recorded in the historical drought disaster data into a 0.25° grid format consistent with the SAPEI data. Label the affected areas, with 1 representing affected and 0 representing non-affected. Calculate the spatial overlap and temporal matching rate between the clustered drought areas and the historical drought disaster areas. Drought indicators are constructed based on the following drought characteristics to measure drought events: Feature 1. Drought duration : refers to the number of time steps between the start time of drought and the end time of drought, which is used to indicate the duration of drought; Feature 2: Drought severity : refers to the absolute value of the accumulated SAPEI values ​​of all drought grids during the drought event. The formula is as follows: ； Where, represents the SAPEI value of the i-th time step; Feature 3: Drought intensity : refers to the average SAPEI value of all grids during the drought period, which is the severity of drought and drought duration The ratio is as follows: ； Feature 4: Drought Area : refers to the maximum range covered by a drought event, which is the maximum area that was in a drought state during the drought event. It is the sum of the areas of all grids when the coverage area is the largest. Step 4: Extracting irrigation response information: Based on the obtained drought event characteristic dynamic dataset, combined with the target area land cover and utilization data and farmland irrigation area data, extract irrigation response information as a drought adaptation measure; Step 5: Drought risk assessment and quantification of irrigation disaster reduction effects: Based on the historical drought disaster data from the disaster risk survey and the county boundary vector data, the effectiveness of irrigation disaster reduction capacity and risk assessment are quantified based on the irrigation response information.

2. The method for quantitatively assessing drought disaster risk integrating irrigation disaster reduction effects according to claim 1 is characterized in that: In step 1, the normalization process unifies the numerical range of the acquired meteorological data to [0, 1] or [-1, 1], making the meteorological data dimensionless data. The normalization formula is as follows: ； ； ； in, Represents data after normalization or standardization. Represents the original data, represents the mean, represents variance; Indicates the maximum and minimum original data.

3. The method for quantitatively assessing drought disaster risk integrating irrigation disaster reduction effects according to claim 1 is characterized in that: In step 1, the drought index is based on the standardized meteorological data and is smoothed in the three dimensions of longitude, latitude, and time using a three-dimensional sliding window to remove outliers. The smoothed data is then interpolated to fill in missing values, obtaining the preprocessed three-dimensional grid SAPEI drought index data. The calculation formula is as follows: ； Where, 、 、 、 、 、 is a constant term; Intermediate variables The calculation formula is as follows: ； Where, is the cumulative probability density, expressed as follows: ； Where, is the probability distribution function, and the formula is as follows: ； Where, 、 、 are the scale, shape and initial state parameters respectively, which are obtained by linear moment method fitting. The formula is as follows: ； ； ； in, for Gamma function, , , yes Probability-weighted moments of sequences; ； Where, is the precipitation evapotranspiration series The ordinal number of the sequence in ascending order, ranging from 1, 2, ..., n; is the sample length of the time series. The precipitation evapotranspiration difference of a month is the sum of the precipitation evapotranspiration difference of the previous k-1 months and the current month as follows: ； ； Where, For different time scales, is the monthly precipitation, is the monthly potential evaporation, is an intermediate variable.

4. The method for quantitatively assessing drought disaster risk integrating irrigation disaster reduction effects according to claim 3 is characterized in that: In step 2, the ST-KMeans clustering method is used to identify drought events, which includes the following sub-steps: Step 2.

1. Determine the drought index level and corresponding SAPEI value of the selected core points based on the SAPEI drought index classification, and obtain a three-dimensional point set containing all drought pixels in the format of [time index t, latitude lat, longitude lon, SAPEI value]; Step 2.2: Construct a spatiotemporal feature sample set. Convert each drought pixel into a feature vector: [x position, y position, time index t, drought intensity]. The feature vectors of all drought pixels form a sample set, where x and y are latitude and longitude or grid index, and t is the day number. Step 2.3, normalize the temporal and spatial three-dimensional features of the spatiotemporal feature sample set respectively; Step 2.4: Aggregate the standardized drought grids into several event clusters, each cluster representing a drought event; Step 2.5: Determine the number of event clusters. Use ST-KMeans, add temporal continuity or spatial connectivity constraints, and use three-dimensional Euclidean distance as the distance function for spatiotemporal clustering. Adjust the weight of the time dimension according to the research focus to obtain several drought event clusters and form a three-dimensional drought event dataset. The three-dimensional Euclidean distance is calculated as follows: ； Where, and represent the values ​​in space and time respectively, Represents the weight of the time dimension.

5. The method for quantitatively assessing drought disaster risk integrating irrigation disaster reduction effects according to claim 4 is characterized in that: In step 2.1, the SAPEI drought index is classified into the following categories: Level 4: Extremely wet, ; Level 3: Severely wet, ; Level 2: Medium moist, ; Level 1: slightly moist, ; Level 0: Normal, ; Level-1: Mild drought, ; Level-2: Moderate drought, ; Level-3: severe drought, ; Level -4: Extreme drought, .

6. The method for quantitatively assessing drought disaster risk integrating irrigation disaster reduction effects according to claim 1, characterized in that: In step 4, irrigation response information extraction includes the following sub-steps: Step 4.1: Process the spatiotemporal range of annual land cover and land use data and farmland irrigation area data into a 0.25° grid format consistent with the drought event data; Step 4.2: Screen land use types that may cause direct economic losses due to drought, including cultivated land, economic forest land, and grasslands of varying degrees of coverage. Different land use types are assigned different weight coefficients based on economic output and vulnerability to obtain the first land use weight matrix. Step 4.3: Based on the spatial correspondence, the high-resolution annual irrigation area data of the region is statistically mapped to the corresponding 0.25° grid to obtain the second weight matrix of irrigation area. The first weight matrix and the second weight matrix are multiplied at the corresponding time and space grid points to obtain the third weight matrix. Step 4.4: Extract the drought intensity and other characteristics of all drought events in the drought event feature dynamic dataset to obtain a first hazard factor intensity matrix. Multiply the first hazard factor intensity matrix and the third weight matrix accordingly to obtain a second hazard factor intensity matrix weakened by disaster prevention and mitigation capabilities. Step 4.5: Sum the second hazard factor intensity matrix within the year to obtain the response information of irrigation drought disaster reduction.

7. The method for quantitatively assessing drought disaster risk integrating irrigation disaster reduction effects according to claim 6, characterized in that: In step 5, the drought risk assessment and irrigation disaster reduction effect quantification include the following sub-steps: Step 5.1: Reselect and redistribute the second hazard factor intensity matrix based on the administrative boundaries of each county. Calculate the weight of the intensity value of the second hazard factor intensity matrix at each grid point within the county relative to the sum of the intensities of all grid points within the county. This yields the fourth weight matrix for the redistribution of the total economic loss within the county. Step 5.2: Multiply the fourth weight matrix by the county's total annual economic losses in time and space to obtain drought economic loss data in raster format, and perform data verification at the spatial and provincial / municipal scales; Step 5.3: Fit the grid-based drought economic losses with a logistic function to obtain the first drought economic vulnerability curve considering irrigation adaptability measures, and assess the risks of different drought intensities under the condition of considering irrigation adaptability measures; Step 5.4: Remove the irrigation weight from the obtained second hazard intensity matrix to obtain the original third hazard intensity matrix. The disaster reduction effect of irrigation is not considered when calculating the redistribution weights. Step 5.5: Calculate the weight of the intensity value of the third hazard factor intensity matrix at each grid point in the county relative to the sum of the intensities of all grid points in the county, and obtain the fifth weight matrix for redistribution of economic loss grid points in the county; Step 5.6: Multiply the fifth weight matrix by the county's annual total economic losses to obtain drought economic loss data in raster format for data validation at the spatial and provincial / municipal scales. Step 5.7: Fit the grid-based drought economic losses by constructing a spatial dynamic explicit model to obtain the second drought economic vulnerability curve without considering irrigation adaptation measures, and assess the risk of different drought intensities without considering adaptation measures; Step 5.8: Under the condition of the same intensity of the hazard factors, the loss value of the second vulnerability curve is subtracted from the loss value of the first vulnerability curve to obtain the quantitative value of the disaster reduction effect of irrigation as a drought adaptation measure.

8. The method for quantitatively assessing drought disaster risk integrating irrigation disaster reduction effects according to claim 7 is characterized in that: In step 5.7, the spatial dynamic explicit model has the following functional form: ； Where, express economic losses incurred; , ) represents the location of the i-th drought core area; represents the drought intensity of the i-th drought core area; represents the spatial weight, reflecting the regional economic density or vulnerability; Represents the impact radius of the drought event.

Images