Mountain torrent disaster risk evaluation method based on multi-source data

Abstract

The invention provides a mountain torrent disaster risk evaluation method based on multi-source data, and relates to the technical field of natural disaster risk evaluation. Comprising the following steps: based on mountain torrent disaster risk evaluation indexes, collecting multi-source index data of a to-be-evaluated object to construct a characteristic value matrix; then, determining a specific point value of each index through a membership calculation formula by utilizing the data and the standard interval; thirdly, calculating relative membership degrees between the point values and the characteristic values so as to evaluate the risk relative degree of each index; and then, in combination with the relative risk membership degree and the weight of each index, the overall risk membership degree of the object is obtained through comprehensive calculation. According to the comprehensive risk membership degree, further calculating to obtain a risk level characteristic value; and finally, according to a variable fuzzy judgment criterion and the calculated risk level characteristic value, judging the mountain torrent disaster risk level of the to-be-evaluated object. According to the method, fuzzy and clear quantitative description in the mountain torrent disaster risk can be realized, and the mountain torrent disaster risk can be accurately evaluated.

Description

Technical Field

[0001] The present application relates to the technical field of natural disaster risk assessment, and in particular to a flash flood disaster risk assessment method based on multi-source data. Background Art

[0002] Flash flood disasters are common in mountainous areas and are caused by local heavy rainfall. Due to their suddenness and high destructiveness, they have attracted much attention in mountainous countries. In order to scientifically and accurately assess the risk of flash flood disasters and provide strong support for disaster prevention and mitigation, flash flood forecasting is necessary. However, due to the poor predictability of flash flood disasters, complex causes, wide coverage, and difficulty in forecasting, flash flood disasters are still a difficult point in disaster prevention and mitigation work in mountainous areas.

[0003] Flash flood disaster risk assessment is an effective non-engineering measure for flood prevention and disaster reduction that changes passive disaster resistance into active disaster avoidance. Flash flood disaster risk involves many factors (or indicators), and each factor has a cause-effect relationship with each other, forming a very complex system. Some indicators in the system are clear, but the influence between indicators in the system is ambiguous and uncertain. Therefore, flash flood disaster risk assessment can be regarded as an effective description problem of "fuzzy" and "clear".

[0004] At present, flash flood disaster risk analysis is mostly qualitative evaluation or semi-qualitative and semi-quantitative evaluation, and the basic data is limited, and the effective regional spatial information is not fully mined. At present, it is still challenging to achieve effective quantitative description of "fuzzy" and "clear" under multi-source spatial data. Summary of the invention

[0005] The purpose of this application is to provide a flash flood disaster risk assessment method based on multi-source data, which can achieve the quantitative characterization of "fuzziness" and "clarity" in flash flood disaster risks and solve the problem of uncertainty assessment of flash flood disaster risks.

[0006] This application is implemented as follows:

[0007] The present application provides a flash flood disaster risk assessment method based on multi-source data, comprising the following steps:

[0008] According to the determined flash flood disaster risk assessment indicators, the multi-source indicator data information corresponding to the object to be evaluated is obtained to construct the indicator characteristic value data matrix; according to the flash flood disaster risk level and the standard interval matrix of the indicators of the object to be evaluated, the indicator point value matrix of the object to be evaluated is calculated through the membership calculation formula; the relative relationship between the indicator characteristic value data matrix and the indicator point value matrix of the object to be evaluated is calculated through the relative membership calculation formula to obtain the relative risk membership of each indicator of the object to be evaluated; according to the relative risk membership of each indicator of the object to be evaluated and the corresponding weight value, the comprehensive risk membership of the object to be evaluated is obtained through the comprehensive membership calculation formula; according to the comprehensive risk membership of the object to be evaluated, the level characteristic value calculation formula is used to calculate and obtain the risk level characteristic value of the object to be evaluated; according to the variable fuzzy judgment criteria and the risk level characteristic value, the flash flood disaster risk level of the object to be evaluated is determined.

[0009] Compared with the prior art, the present invention has at least the following advantages or beneficial effects:

[0010] This application proposes a flash flood disaster risk assessment method based on multi-source data. Through the integration of multi-source data, membership calculation, relative membership analysis, and calculation of comprehensive membership and risk level characteristic values, a scientific, accurate, and feasible risk quantitative assessment system is constructed.

[0011] First, it obtains multi-source indicator data information based on the determined flash flood disaster risk assessment indicators. This step ensures the comprehensiveness and diversity of the basic data for evaluation, and overcomes the problem of one-sided evaluation or large errors that may be caused by a single data source in the past. The integration of multi-source data makes the risk assessment closer to the actual situation and improves the accuracy and reliability of the assessment results.

[0012] Then, through the membership calculation formula, the indicator data of the object to be evaluated is converted into an indicator point value matrix. This process realizes the effective quantification and standardization of complex data. The introduction of membership enables indicators of different dimensions and properties to be unified under the same evaluation framework for comparison, greatly enhancing the objectivity and operability of the evaluation.

[0013] Then, using the relative membership calculation formula, we further analyzed the relative relationship between the characteristic value data matrix of the indicator to be evaluated and the indicator point value matrix, and obtained the relative risk membership of each indicator. This step not only considers the absolute level of the indicator, but also takes into account the relative position relationship between the indicators, making the risk assessment more practical and able to more accurately reflect the contribution of each indicator to the overall risk.

[0014] Next, through the comprehensive membership calculation formula, the relative risk membership of each indicator is combined with the corresponding weight value to obtain the comprehensive risk membership of the object to be evaluated. This step realizes the comprehensive integration from a single indicator to the overall risk, providing a solid foundation for the final risk level determination. Subsequently, the risk level characteristic value calculated based on the comprehensive risk membership provides a scientific basis for the quantitative representation of the risk level, making the division of risk levels clearer and more reasonable.

[0015] In addition, this application adopts a variable fuzzy evaluation criterion, which makes the determination of risk level flexible and adaptable. When facing flash flood disaster risks in different regions and under different climatic conditions, the evaluation criteria and parameter settings can be adjusted according to the actual situation, so as to obtain risk assessment results that are more in line with the local actual situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 This is a flow chart of an embodiment of a flash flood disaster risk assessment method based on multi-source data of the present application;

[0018] Figure 2 It is a characteristic value data layer of the average annual rainfall in one embodiment of the present application;

[0019] Figure 3 It is a characteristic value data layer of the maximum 60-minute rainfall extreme value in one embodiment of the present application;

[0020] Figure 4 It is a characteristic value data layer of terrain relief in one embodiment of the present application;

[0021] Figure 5 It is a characteristic value data layer of river network density in one embodiment of the present application;

[0022] Figure 6 It is a characteristic value data layer of population density in one embodiment of the present application;

[0023] Figure 7 This is a characteristic value data layer of average GDP per capita in one embodiment of the present application;

[0024] Figure 8 It is a characteristic value data layer of the maximum 6-hour rainfall extreme value in an embodiment of the present application;

[0025] Fig. 9 It is a characteristic value data layer of the NDVI vegetation index in one embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0027] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0028] Flash flood disasters pose a major threat to mountain safety due to their suddenness and high destructiveness. Existing flash flood disaster risk assessment methods mostly rely on qualitative or semi-quantitative methods, and are limited by data and ambiguity, making it difficult to achieve scientific and accurate assessments. The inventors deeply analyzed the shortcomings of the existing technology and found its limitations in dealing with complex systems, multi-source data and ambiguity, which gave rise to the idea of ​​developing a new evaluation method.

[0029] In response to the above problems, this application proposes an innovative flash flood disaster risk assessment method, the core of which is to integrate multi-source spatial data, and achieve an effective quantitative description of "fuzzy" and "clear" by constructing an indicator eigenvalue data matrix, calculating the indicator point value matrix and relative risk membership, comprehensive risk membership and risk level eigenvalues. This method not only overcomes the shortcomings of existing technologies in data utilization and fuzzy processing, but also improves the accuracy and scientificity of flash flood disaster risk assessment, providing more reliable decision-making support for disaster prevention and mitigation work.

[0030] See also Figure 1 The flash flood disaster risk assessment method based on multi-source data includes the following steps:

[0031] Step S101: According to the determined flash flood disaster risk assessment index, multi-source index data information corresponding to the object to be assessed is obtained to construct an index characteristic value data matrix;

[0032] In step S101, first, according to the predetermined flash flood disaster risk assessment indicators (usually including annual average rainfall, annual maximum 6-hour rainfall, annual maximum 60-minute rainfall, river network density, vegetation index, terrain relief, population density, GDP per capita, etc.), the multi-source indicator data information corresponding to the object to be evaluated (such as a mountainous basin) is collected. This information can come from multiple channels such as meteorological observations, remote sensing images, geological surveys, etc. Subsequently, these data are organized into a matrix form, namely, an indicator eigenvalue data matrix, in which each row represents an evaluation indicator and each column represents an object to be evaluated or data at different time points.

[0033] Exemplarily, multi-disciplinary cross-technical means such as GIS technology (geographic information system) and mean point algorithm can be used to analyze and process multi-source indicator data to obtain the characteristic value data of each indicator of the object to be evaluated, based on which the indicator characteristic value data matrix can be further constructed. Among them, GIS technical means mainly include geographic spatial statistical analysis, spatial interpolation and spatial distribution. Spatial statistical analysis refers to the use of GIS spatial analysis functions, such as buffer analysis, overlay analysis, etc., to reveal spatial patterns and relationships. Spatial interpolation refers to the use of interpolation algorithms in GIS (such as Kriging interpolation, inverse distance weighted interpolation, etc.) to estimate the characteristic values ​​of the entire area for point source data (such as data from meteorological observation stations). Spatial distribution is a technical means used in GIS to display the spatial distribution characteristics of geographical phenomena. Through spatial distribution, geographical data can be intuitively displayed in the form of maps, charts, etc., to help users better understand the spatial distribution laws and characteristics of geographical phenomena. Spatial distribution is not limited to simple data visualization, but also includes spatial analysis and mining of data to reveal deep information hidden behind the data.

[0034] Examples of the data matrix of the characteristic values ​​of each indicator are as follows:

[0035]

[0036] Among them, a ij is the indicator characteristic value data, i = 1, 2, ..., n; n represents the total number of indicators. j is the object to be evaluated with number j.

[0037] Step S102: according to the flash flood disaster risk level and the standard interval matrix of the index of the object to be evaluated, the index point value matrix of the object to be evaluated is calculated by the membership calculation formula;

[0038] In step S102, according to the known flash flood disaster risk level and the standard interval matrix of the indicator to be evaluated, the eigenvalue of each indicator is converted into the corresponding membership value using the membership calculation formula. These membership values ​​reflect the degree of proximity of each indicator to the risk level standard, thereby forming an indicator point value matrix. The membership calculation formula is usually based on fuzzy mathematics theory and can handle the fuzzy relationship between the indicator value and the standard value. The establishment of the indicator point value matrix realizes the conversion of raw data into comparable membership values, which provides a basis for the subsequent calculation of relative risk membership. At the same time, the membership calculation formula can handle fuzzy relationships and improve the accuracy of the evaluation.

[0039] For example, the flash flood disaster risk level is usually set at five levels, namely Level I, Level II, Level III, Level IV, and Level V, corresponding to very low risk, lower risk, medium risk, higher risk, and very high risk respectively.

[0040] Step S103: Calculate the relative relationship between the indicator characteristic value data matrix and the indicator point value matrix of the object to be evaluated by using the relative membership calculation formula to obtain the relative risk membership of each indicator of the object to be evaluated;

[0041] In step S103, the relative relationship between the indicator characteristic value data matrix and the indicator point value matrix is ​​analyzed by the relative membership calculation formula. This step aims to reveal the degree of association between each indicator and the risk level in the overall risk assessment. The relative membership calculation formula may involve mathematical methods such as matrix operations and weighted summation to obtain the relative risk membership of each indicator of the object to be evaluated. That is, the calculation of the relative risk membership further refines the contribution of each indicator in the risk assessment, making the evaluation results more detailed and accurate. At the same time, this step provides an important basis for the subsequent calculation of the comprehensive risk membership.

[0042] Step S104: According to the relative risk membership of each indicator of the object to be evaluated and the corresponding weight value, a calculation is performed through a comprehensive membership calculation formula to obtain a comprehensive risk membership of the object to be evaluated;

[0043] In step S104, according to the relative risk membership of each indicator of the object to be evaluated and the corresponding weight value (these weight values ​​reflect the relative importance of each indicator in the overall risk assessment), a weighted sum operation is performed through the comprehensive membership calculation formula to obtain the comprehensive risk membership of the object to be evaluated. The comprehensive membership calculation formula is usually a linear or nonlinear function that can comprehensively consider the influence of each indicator. Among them, the calculation of the comprehensive risk membership realizes the integration of the risk membership of multiple indicators into a comprehensive value, which is convenient for comprehensive risk assessment of the object to be evaluated. At the same time, the introduction of weight values ​​makes the evaluation results more in line with the actual situation and improves the rationality and scientificity of the evaluation.

[0044] Step S105: According to the comprehensive risk membership of the object to be evaluated, the level characteristic value calculation formula is used to calculate and obtain the risk level characteristic value of the object to be evaluated;

[0045] In step S105, the risk level characteristic value is obtained by calculating the comprehensive risk membership of the object to be evaluated through the level characteristic value calculation formula. The level characteristic value calculation formula usually maps the comprehensive risk membership to a specific risk level based on a certain grading standard or functional relationship. This step realizes the conversion from numerical value to level. That is, the evaluation result is more intuitive and easy to understand through the calculation of the risk level characteristic value. Decision makers can quickly judge the risk status of the object to be evaluated based on the risk level characteristic value, and formulate corresponding disaster prevention and mitigation measures accordingly.

[0046] Step S106: Determine the flash flood disaster risk level of the object to be evaluated according to the variable fuzzy evaluation criteria and the risk level characteristic value.

[0047] In step S106, the flash flood disaster risk level of the object to be evaluated is determined according to the variable fuzzy evaluation criteria and the risk level characteristic value. The variable fuzzy evaluation criteria is a judgment method based on fuzzy mathematics theory, which can handle the uncertainty and ambiguity in the evaluation process. By comparing and matching the risk level characteristic value with the judgment criteria, the risk level of the object to be evaluated is finally determined. That is, through the application of the variable fuzzy evaluation criteria, the uncertainty and ambiguity in the evaluation process can be fully considered, making the evaluation results more accurate and reliable. At the same time, the determination of the risk level helps decision makers to formulate more accurate and effective disaster prevention and mitigation strategies.

[0048] In summary, this application constructs a scientific, accurate and practical risk assessment system through the integration of multi-source data, membership calculation, relative membership analysis, and calculation of comprehensive membership and risk level characteristic values.

[0049] First, it obtains multi-source indicator data information based on the determined flash flood disaster risk assessment indicators. This step ensures the comprehensiveness and diversity of the basic data for evaluation, and overcomes the problem of one-sided evaluation or large errors that may be caused by a single data source in the past. The integration of multi-source data makes the risk assessment closer to the actual situation and improves the accuracy and reliability of the assessment results.

[0050] Then, through the membership calculation formula, the indicator data of the object to be evaluated is converted into an indicator point value matrix. This process realizes the effective quantification and standardization of complex data. The introduction of membership enables indicators of different dimensions and properties to be unified under the same evaluation framework for comparison, greatly enhancing the objectivity and operability of the evaluation.

[0051] Then, using the relative membership calculation formula, we further analyzed the relative relationship between the characteristic value data matrix of the indicator to be evaluated and the indicator point value matrix, and obtained the relative risk membership of each indicator. This step not only considers the absolute level of the indicator, but also takes into account the relative position relationship between the indicators, making the risk assessment more practical and able to more accurately reflect the contribution of each indicator to the overall risk.

[0052] Next, through the comprehensive membership calculation formula, the relative risk membership of each indicator is combined with the corresponding weight value to obtain the comprehensive risk membership of the object to be evaluated. This step realizes the comprehensive integration from a single indicator to the overall risk, providing a solid foundation for the final risk level determination. Subsequently, the risk level characteristic value calculated based on the comprehensive risk membership provides a scientific basis for the quantitative representation of the risk level, making the division of risk levels clearer and more reasonable.

[0053] In addition, this application adopts a variable fuzzy evaluation criterion, which makes the determination of risk level flexible and adaptable. When facing flash flood disaster risks in different regions and under different climatic conditions, the evaluation criteria and parameter settings can be adjusted according to the actual situation, so as to obtain risk assessment results that are more in line with the local actual situation.

[0054] Based on the aforementioned scheme, in some implementation methods of the present application, the multi-source indicator data information corresponding to the object to be evaluated is obtained to construct an indicator characteristic value data matrix, including: obtaining and aligning, vectorizing, and converting the obtained rainfall indicator data of the object to be evaluated into a point value layer; using the point layer to perform spatial interpolation through the thin plate tension spline method to obtain the characteristic value data of each rainfall indicator; and constructing a corresponding indicator characteristic value data matrix based on the characteristic value data of each rainfall indicator.

[0055] In the above implementation, in order to construct the indicator characteristic value data matrix for subsequent evaluation, the following key links are mainly included:

[0056] (1) Data acquisition and preprocessing: First, collect rainfall index data about the object to be evaluated from multiple data sources (such as meteorological observation stations, satellite remote sensing, radar, etc.). Since different data sources may be inconsistent in time, space or format, it is necessary to align these data to ensure their consistency in time, space and format, so as to facilitate subsequent data processing and analysis. Then, convert the rainfall index data from the original format (such as text, image, etc.) into a vector format, and further convert it into a point value map. The point value map can intuitively show the distribution of rainfall indicators in different geographical locations, providing a basis for subsequent spatial interpolation.

[0057] (2) Spatial interpolation: Based on the converted point value map, a layer containing rainfall index point data is generated. At the same time, the thin plate tension spline method (TPS method) is used for spatial interpolation to infer the characteristic value of rainfall index in unknown areas or locations based on the known point layer data. The TPS method has the advantages of maintaining the spatial continuity and smoothness of the data and can better reflect the spatial variation trend of rainfall indicators.

[0058] (3) Construct the indicator characteristic value data matrix: The characteristic value data of each rainfall indicator obtained by the TPS method are sorted to ensure the accuracy and completeness of the data. Then, according to the sorted characteristic value data, the indicator characteristic value data matrix is ​​constructed according to certain rules (such as geographical location, time, etc.).

[0059] In summary, in the above implementation, the accuracy and consistency of rainfall index data are improved through preprocessing steps such as data registration, vectorization and point value map conversion, providing a reliable data basis for subsequent risk assessment. The TPS method is used for spatial interpolation, which can accurately infer the characteristic values ​​of rainfall indicators in unknown areas or locations, enhance the spatial analysis ability of data, and help reveal the spatial distribution patterns and characteristics of rainfall indicators. At the same time, the constructed indicator characteristic value data matrix is ​​structured and standardized, which is convenient for subsequent risk assessment, analysis or prediction. This matrix provides intuitive and comprehensive data support for decision makers, helping them to better understand the risk of flash flood disasters and formulate corresponding response measures.

[0060] Based on the aforementioned scheme, in some implementation methods of the present application, the multi-source indicator data information corresponding to the object to be evaluated is obtained to construct an indicator characteristic value data matrix, including: obtaining the terrain data of the object to be evaluated; determining the regional terrain undulation and river network density series values ​​at different grid unit scales according to the terrain data of the object to be evaluated; using the mean change point algorithm to analyze the regional terrain undulation and river network density series values ​​to determine the optimal grid unit analysis scale threshold of the terrain undulation and river network density of the object to be evaluated; recalculating or extracting the indicator characteristic value data of the terrain undulation and river network density of the object to be evaluated according to the optimal grid unit analysis scale threshold of the terrain undulation and river network density; and constructing the corresponding indicator characteristic value data matrix according to the indicator characteristic value data.

[0061] In the above implementation, terrain data and river network density are introduced as important multi-source indicators to more comprehensively assess the flash flood disaster risk of the object to be evaluated (such as a region, river basin or city). Specifically, the following steps are included:

[0062] (1) Obtain terrain data:

[0063] Topographic data is one of the essential basic data for assessing flash flood disaster risks. It usually includes elevation models (such as DEM data), slope, aspect and other information. These data can be obtained through remote sensing technology, ground measurement or digital maps.

[0064] (2) Determine the regional terrain relief and river network density:

[0065] Based on the terrain data, the terrain relief at different grid unit scales can be calculated. The terrain relief reflects the complexity of the surface morphology and has an important impact on the occurrence and development of flash floods. In addition, the river network density refers to the total length of rivers per unit area, which reflects the development degree and drainage capacity of the water system.

[0066] (3) Determine the optimal grid cell analysis scale threshold:

[0067] The mean change point algorithm is used to analyze the series of values ​​of regional terrain relief and river network density. The mean change point algorithm is a method used to detect mutation points in data series. It can help us find the points where terrain relief and river network density change most significantly at different grid unit scales, that is, the optimal grid unit analysis scale threshold.

[0068] The significance of determining the optimal grid unit analysis scale threshold is that it can help us find the optimal grid scale that can reflect the details of terrain and river network characteristics while maintaining data stability and reliability, thereby improving the accuracy and effectiveness of subsequent risk assessment.

[0069] (4) Recalculate or extract indicator characteristic value data:

[0070] According to the determined optimal grid unit analysis scale threshold, the indicator characteristic value data of the terrain relief and river network density of the object to be evaluated are recalculated or extracted. This step ensures that the data used in subsequent analysis is obtained at the optimal grid scale, thereby improving the representativeness and reliability of the data.

[0071] (5) Construct the indicator eigenvalue data matrix:

[0072] The recalculated or extracted terrain relief and river network density index characteristic value data, as well as the characteristic value data of other related rainfall indicators (such as rainfall, rainfall intensity, etc.), are constructed into an indicator characteristic value data matrix according to certain rules (such as geographical location, time, etc.). This matrix will serve as the basic data for subsequent risk assessment, analysis or prediction.

[0073] In summary, by determining the optimal grid unit analysis scale threshold, the data accuracy of key indicators such as terrain relief and river network density is ensured, thereby improving the accuracy of the entire risk assessment process. Recalculating or extracting indicator characteristic value data at the optimal grid scale can better reflect the spatial distribution characteristics of terrain and river network, and enhance the spatial analysis ability of data. In addition, by incorporating key indicators such as terrain relief and river network density into the risk assessment system, the assessment results are more comprehensive and objective, which helps decision makers to more accurately grasp the risk of flash flood disasters.

[0074] Based on the above scheme, in some implementations of the present application, the mean change point algorithm includes: dividing the sample into two segments: X1, X2, ..., X i-1 and X i , X i+1 , ...X N Calculate the arithmetic mean of each sample segment according to the following formula and And the statistic S i , S, and the expected value E(SS i ):

[0075]

[0076] Among them, S i is the difference between the sum of squares of deviations of the two samples; S is the total sum of squares of deviations; is the arithmetic mean of the total samples; N is the total number of samples; the expected value E(SS i ) is the point with the largest value as the optimal grid unit analysis scale threshold.

[0077] Based on the above scheme, in some implementations of the present application, the step of obtaining multi-source indicator data information corresponding to the object to be evaluated to construct an indicator characteristic value data matrix includes: obtaining the population density and GDP per capita data of the object to be evaluated. According to the population density and GDP per capita data of the object to be evaluated, the distribution density function b(x, y) of the population density and GDP per capita is determined, and b(x, y) is integrated in two-dimensional space to obtain the two-dimensional spatial distribution data of the indicator; wherein the integral operation is calculated as: A is the two-dimensional space within the region; b(x,y) is the population density or the distribution density function of GDP per capita at a certain point in the two-dimensional space within the statistical region; (x,y) is the geographic spatial coordinates of a certain point in the two-dimensional space; B is the characteristic value data of the population density or GDP per capita. According to the characteristic value data of the population density and GDP per capita, the corresponding indicator characteristic value data matrix is ​​constructed.

[0078] In the above implementation, spatial statistics and integral operations are combined to quantify the distribution of population density and GDP per capita in two-dimensional space, and a data matrix is ​​constructed accordingly. The implementation includes the following steps:

[0079] (1) Data acquisition: Obtain the population density and GDP per capita of the object to be evaluated (such as a city, region, etc.) from relevant data sources (such as statistical departments, remote sensing data, geographic information systems, etc.). These data usually exist in the form of grid, point or surface data, representing the population and economic activity intensity in different geographical locations.

[0080] (2) Determine the distribution density function: Based on the obtained population density and GDP per capita data, use spatial statistical methods (such as kernel density estimation, Kriging interpolation, etc.) to determine the distribution density function b(x, y) of these indicators in two-dimensional space. This function describes the estimated value of population density or GDP per capita at any point (x, y) in the statistical area.

[0081] (3) Integration operation: Integrate the distribution density function b(x, y) in two-dimensional space to calculate the total amount or some form of cumulative value of the population density or GDP per capita in the entire region. The calculation formula of the integration operation is usually expressed as the integration of the function b(x, y) in region A (i.e., the entire statistical area), where (x, y) is the geographic spatial coordinate in the two-dimensional space. The result of the integration B is the characteristic value data of the population density or GDP per capita indicator, which reflects the overall level or distribution of the indicator in the entire region.

[0082] (4) Constructing a data matrix: Based on the eigenvalue data of population density and GDP per square kilometer (i.e., the result of integration operation), as well as other possible related indicator data, construct the corresponding indicator eigenvalue data matrix. Each row or column of this matrix represents a specific indicator (such as population density, GDP per square kilometer, etc.), and the elements in the matrix are the eigenvalues ​​of the indicator in different evaluation units (such as grids, regions, etc.).

[0083] Based on the above scheme, in some implementations of the present application, the membership calculation formula includes a point value calculation formula with a membership of 1: Among them, m ih is the standard interval of the indicator of the object to be evaluated [x ih ,y ih ] is the point value with a membership degree of 1, i is the indicator number, h is the flash flood disaster risk level, h = 1, 2,…, c.

[0084] In the above implementation, for each indicator of the object to be evaluated, the standard intervals under different risk levels are defined, and the point values ​​with a membership degree of 1 in these intervals (i.e., the values ​​that best meet the risk level standards) are calculated to achieve a quantitative assessment of the flash flood disaster risk. For each evaluation indicator (such as rainfall, terrain slope, soil type, etc.), the standard intervals [x ih ,y ih ]. These intervals reflect the reasonable value range of this indicator under different risk levels.

[0085] Based on the above scheme, in some implementations of the present application, the relative membership calculation formula includes:

[0086]

[0087] Among them, m ih is the standard interval of the indicator of the object to be evaluated [x ih ,y ih ] The point value with a membership degree of 1, i is the index number, i = 1, 2, ..., n, n is the total number of indicators; h is the flash flood disaster risk level, h = 1, 2, ..., c; O j is the object to be evaluated with number j, and μ is the relative risk membership.

[0088] Based on the above scheme, in some implementations of the present application, the comprehensive membership calculation formula includes:

[0089]

[0090] Among them, v(O j ) h is the object to be evaluated O j For the comprehensive membership of risk level h; α is the optimization criterion parameter, which takes a value of 1 or 2; p is the distance parameter, which takes a value of 1 or 2.

[0091] Since for the comprehensive membership calculation formula, α and p are both 1 or 2, there are four combinations of α and p:

[0092] (1) When α = 1, p = 1, the comprehensive membership degree calculation formula is:

[0093]

[0094] (2) When α = 1, p = 2, the comprehensive membership degree calculation formula is:

[0095]

[0096] (3) When α = 2, p = 1, the comprehensive membership degree calculation formula is:

[0097]

[0098] (4) When α = 2, p = 2, the comprehensive membership degree calculation formula is:

[0099]

[0100] Based on the above solution, in some implementations of the present application, the level characteristic value calculation formula includes: Among them, v k 0 (O j ) h It is v k (O j ) h Normalized comprehensive membership; H k is the level characteristic value of the object to be evaluated, k=1,2,3,4.

[0101] Based on the above solution, in some implementations of the present application, the level characteristic value calculation formula includes: is the average value of level eigenvalue;

[0102] Variable fuzzy evaluation criteria include:

[0103] In order to enable those skilled in the art to understand the present application more intuitively, a specific example will be used here to illustrate.

[0104] Taking Guizhou Province as an example, we collected multi-source indicator data information, used GIS technology, mean change point algorithm and other multidisciplinary cross-disciplinary technical means to analyze and process the multi-source indicator data, and obtained the characteristic value data layer of each indicator in Guizhou Province, such as Figure 2 to Figure 9 shown.

[0105] Accordingly, based on the characteristic value data layer of the flash flood disaster risk assessment index in Guizhou Province, the layer attribute statistics function of the GIS software is used to extract the attribute information of each indicator layer, and further construct the indicator characteristic value data matrix as follows:

[0106]

[0107] Among them, the total number of indicators n is 8; the object to be evaluated O1 represents Tongren City, and O2 represents Qianxinan Buyi and Miao Autonomous Prefecture.

[0108] For example, the standard interval matrix of flash flood disaster risk indicators for Tongren City and Qianxinan Buyi and Miao Autonomous Prefecture is:

[0109]

[0110] Therefore, according to the standard interval matrix of flash flood disaster risk indicators of Tongren City and Qianxinan Buyi and Miao Autonomous Prefecture, and the point value calculation formula with a membership degree of 1, the indicator point value matrix M of the object to be evaluated is calculated as follows:

[0111]

[0112] The relative risk membership of Tongren City is:

[0113]

[0114] The relative risk affiliation of Qianxinan Buyi and Miao Autonomous Prefecture is:

[0115]

[0116] Among them, the weight values ​​of eight indicators including annual average rainfall, annual maximum 6-hour rainfall, annual maximum 60-minute rainfall, river network density, vegetation index, terrain undulation, population density, and per capita GDP are: W = {0.0826, 0.0776, 0.3215, 0.1276, 0.0164, 0.0064, 0.2397, 0.1283}.

[0117] According to the comprehensive membership calculation formula, the comprehensive risk membership of Tongren City is determined as follows:

[0118]

[0119] The comprehensive risk affiliation degree of Qianxinan Buyi and Miao Autonomous Prefecture is:

[0120]

[0121] According to the calculation formula of the average value of the level characteristic value, the calculation results of the flash flood disaster risk level characteristic value and its average value are obtained, as shown in Table 1.

[0122] Table 1 Calculation results of characteristic values ​​of flash flood disaster risk level

[0123] Object to be evaluated <![CDATA[H1]]> <![CDATA[H2]]> <![CDATA[H3]]> <![CDATA[H4]]> H <![CDATA[O1]]> 2.417 2.345 2.697 2.529 2.497 <![CDATA[O2]]> 2.417 2.315 1.949 1.980 2.165

[0124] According to the variable fuzzy evaluation criteria, it can be judged that the flash flood disaster risk level of Tongren City and Qianxinan Buyi and Miao Autonomous Prefecture is both level II, which is a low risk and medium risk. The risk level of Tongren City is higher than that of Qianxinan Buyi and Miao Autonomous Prefecture. The level characteristic values ​​of the two evaluation objects are both high, especially Tongren City, which is close to medium risk.

[0125] According to the results of the mountain torrent disaster investigation in the jurisdiction from 2013 to 2015, the risk in Tongren City is higher than that in Qianxinan Buyi and Miao Autonomous Prefecture (the former has 3,126 dangerous areas, while the latter has 1,956). It can be seen that the calculation results of this example are basically consistent with the actual situation.

[0126] That is, the evaluation method of the present application can realize the quantitative evaluation from multi-source geographic spatial data to the flash flood disaster risk level, and provides a practical implementation idea and calculation scheme that coordinates GIS technology with variable set knowledge for quantitatively describing the coexistence of "fuzzy" and "clear" characteristics in the flash flood disaster risk system. It is of direct help to the early warning and prevention of flash flood disasters, and has important application prospects in regional land use planning and other aspects.

[0127] It will be apparent to those skilled in the art that the present application is not limited to the details of the exemplary embodiments described above, and that the present application can be implemented in other specific forms without departing from the spirit or essential features of the present application. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the present application is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims be included in the present application. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A flash flood disaster risk assessment method based on multi-source data, characterized in that: The following steps are involved: According to the determined flash flood disaster risk assessment indicators, the multi-source indicator data information corresponding to the object to be evaluated is obtained to construct an indicator characteristic value data matrix; According to the flash flood disaster risk level and the standard interval matrix of the indicators of the object to be evaluated, the indicator point value matrix of the object to be evaluated is calculated by the membership calculation formula; By using the relative membership calculation formula, the relative relationship between the indicator characteristic value data matrix and the indicator point value matrix of the object to be evaluated is calculated to obtain the relative risk membership of each indicator of the object to be evaluated; According to the relative risk membership of each indicator of the object to be evaluated and the corresponding weight value, the comprehensive risk membership of the object to be evaluated is calculated through the comprehensive membership calculation formula to obtain the comprehensive risk membership of the object to be evaluated; According to the comprehensive risk membership of the object to be evaluated, the risk level characteristic value of the object to be evaluated is calculated through the level characteristic value calculation formula; According to the variable fuzzy evaluation criteria and the risk level characteristic value, the flash flood disaster risk level of the object to be evaluated is determined.

2. The method according to claim 1, characterized in that The step of obtaining multi-source indicator data information corresponding to the object to be evaluated to construct an indicator characteristic value data matrix includes: Acquire and register, vectorize, and convert the acquired rainfall index data of the object to be evaluated into a point value layer; Using the point map, the thin plate tension spline method is used for spatial interpolation to obtain the characteristic value data of each rainfall index; According to the eigenvalue data of each rainfall indicator, the corresponding indicator eigenvalue data matrix is ​​constructed.

3. The method according to claim 1, characterized in that The step of obtaining multi-source indicator data information corresponding to the object to be evaluated to construct an indicator characteristic value data matrix includes: Obtaining terrain data of the object to be evaluated; According to the terrain data of the object to be evaluated, determine the regional terrain relief and river network density series values ​​at different grid unit scales; The mean change point algorithm is used to analyze the regional terrain relief and river network density series values ​​to determine the optimal grid unit analysis scale threshold for the terrain relief and river network density of the object to be evaluated; Recalculate or extract the index characteristic value data of the terrain relief and river network density of the object to be evaluated according to the optimal grid unit analysis scale threshold of the terrain relief and river network density; According to the indicator eigenvalue data, the corresponding indicator eigenvalue data matrix is ​​constructed.

4. The method according to claim 3, characterized in that The mean change point algorithm includes: Divide the sample into two segments: X1, X2, ..., X i-1 and X i , X i+1 , ...X N ; Calculate the arithmetic mean of each sample segment according to the following formula: and And the statistic S i , S, and the expected value E(SS i ): Among them, S i is the difference between the sum of squares of deviations of the two samples; S is the total sum of squares of deviations; is the arithmetic mean of the total samples; N is the total number of samples; the expected value E(SS i ) is the point with the largest value as the optimal grid unit analysis scale threshold.

5. The method according to claim 1, characterized in that The step of obtaining multi-source indicator data information corresponding to the object to be evaluated to construct an indicator characteristic value data matrix includes: Obtain the population density and per capita GDP data of the object to be evaluated; According to the population density and GDP per capita data of the object to be evaluated, the distribution density function b(x, y) of population density and GDP per capita is determined, and b(x, y) is integrated in two-dimensional space to obtain the two-dimensional spatial distribution data of the indicator; the calculation formula for the integration operation is: A is the two-dimensional space within the region; b(x,y) is the population density or the distribution density function of GDP per capita at a certain point in the two-dimensional space within the statistical region; (x,y) is the geographic spatial coordinates of a certain point in the two-dimensional space; B is the characteristic value data of the population density or GDP per capita indicator; According to the eigenvalue data of population density and per capita GDP indicators, the corresponding indicator eigenvalue data matrix is ​​constructed.

6. The method according to claim 1, characterized in that The membership calculation formula includes a point value calculation formula with a membership of 1: Among them, m ih is the standard interval of the indicator of the object to be evaluated [x ih ,y ih ] is the point value with a membership degree of 1, i is the indicator number, h is the flash flood disaster risk level, h = 1, 2,…, c.

7. The method according to claim 1, characterized in that The relative membership calculation formula includes: Among them, m ih is the standard interval of the indicator of the object to be evaluated [x ih ,y ih ] The point value with a membership degree of 1, i is the index number, i = 1, 2, ..., n, n is the total number of indicators; h is the flash flood disaster risk level, h = 1, 2, ..., c; O j is the object to be evaluated with number j, and μ is the relative risk membership.

8. The method according to claim 1, characterized in that The comprehensive membership calculation formula includes: Among them, v(O j ) h is the object to be evaluated O j For the comprehensive membership of risk level h; α is the optimization criterion parameter, which takes a value of 1 or 2; p is the distance parameter, which takes a value of 1 or 2.

9. The method according to claim 1, characterized in that: The level characteristic value calculation formula includes: Among them, v k 0 (O j ) h It is v k (O j ) h Normalized comprehensive membership; H k is the level characteristic value of the object to be evaluated, k=1,2,3,4.

10. The method according to claim 9, characterized in that The level characteristic value calculation formula includes: is the average value of level eigenvalue; Variable fuzzy evaluation criteria include:

Images