A method for exploring urban waterlogging risk influencing factors based on landscape analysis

Abstract

The application discloses a kind of urban waterlogging risk influence factor exploration method based on landscape analysis, first, the soil type data needed to simulate urban waterlogging surface runoff, land use data and rainfall data are collected, and based on the obtained urban drainage planning, further sub-catchment is divided, so as to obtain the water depth and water range of each sub-catchment in combination with elevation DEM data;Then, the landscape of the study area is analyzed, the macro pattern of urban green space is analyzed using the MSPA method, and the micro form of urban green space is analyzed using landscape pattern index, to obtain the macro and micro urban green space landscape analysis results;Finally, using geographic detector software, coupling analysis waterlogging water depth and macro pattern, micro landscape form, to explore the most influential factor on waterlogging water depth under the condition of landscape pattern spatial differentiation.The application establishes a method system for urban landscape analysis, which can provide a landscape influence factor exploration method for urban areas affected by waterlogging.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of urban waterlogging prevention and treatment, and more particularly to a method for exploring urban waterlogging risk influencing factors based on landscape analysis. BACKGROUND

[0002] With the advancement of urbanization and the influence of climate change, urban flooding events occur frequently, causing great damage to human health and safety and property safety. At present, in order to effectively utilize rainwater resources, China has also proposed a water resource management method of "sponge city", and corresponding low-impact development measures are also applied in the design of mitigating rainstorm waterlogging. Low-impact development measures are mainly micro regional catchment unit construction based on source control, which is an important part of the "micro catchment unit-mesoscopic regional drainage system-macroscopic level water ecological system" waterlogging prevention and treatment system under the concept of building a sponge city, and plays a role in dredging flood from the source. Urban green land itself has a sponge effect and can play a role in absorbing and storing rainwater, thereby reducing flood disasters from the source. Therefore, by analyzing the current urban green land pattern, the influence of the macro pattern and micro form on the urban waterlogging water depth is explored, and the most influential influencing factor is determined, which can provide improvement suggestions for urban waterlogging prevention and treatment. SUMMARY

[0003] The present application aims to overcome the shortcomings of the prior art and provide a method for exploring urban waterlogging risk influencing factors based on landscape analysis, which improves the current urban landscape analysis system, constructs a landscape analysis system from the macro pattern and micro form, and uses geographic detector software to explore the influence of urban landscape on waterlogging water depth risk, thereby providing a targeted improvement direction for regional landscape improvement.

[0004] The technical purpose of the present application is achieved by the following technical scheme.

[0005] A method for exploring urban waterlogging risk influencing factors based on landscape analysis, according to the following steps:

[0006] Step 1, collect the data required for constructing the SCS-CN model, including soil type data, land use data and rainfall data, and then simulate in ArcGIS to obtain the surface runoff value in the study area under different rainfall return periods.

[0007] In step 1, data is collected and unified into 30m x 30m precision; soil type data comes from the Harmonized World Soil Database (v1.2) dataset provided by the Food and Agriculture Organization of the United Nations (http: / / www.fao.org / soils-portal / en / ); land use data comes from the Landsat8 OLI_TIRS remote sensing dataset provided by the Geospatial Data Cloud (http: / / www.gscloud.cn / ), which is interpreted by ENVI remote sensing analysis software to obtain the land use situation of the study area; rainfall data is determined according to the rainfall intensity formula in the "Technical Guidelines for the Construction of Tianjin Sponge City", which is based on the division standard of Tianjin storm intensity zoning, and the form is as follows:

[0008]

[0009] In the formula: q is the design storm intensity, unit is L / (s·hm 2 ); t is the rainfall duration, unit is min; P1 is the rainfall return period, unit is year. In this study, the rainfall duration is set to 120 min, and the rain peak position coefficient is set to 0.333.

[0010] In step 1, the SCS-CN model is obtained based on the relationship between rainfall, runoff and water storage in the soil, and the form of the rainfall runoff formula is as follows:

[0011]

[0012] In the formula: S is the maximum water storage value of soil when different soil types and land use methods jointly act, unit is mm; Q is the actual surface runoff after experiencing rainfall, unit is mm; P is the design or actual rainfall, unit is mm.

[0013] The calculation formula of S is as follows:

[0014]

[0015] In the formula: CN is an empirical value obtained according to different soil types, land use methods and soil moisture conditions (AMC).

[0016] Step 2, according to the urban drainage planning and design in the study area, the sub-catchment area is divided, combined with DEM elevation data, following the principle of "low-lying submergence", the total surface runoff in the sub-catchment area is distributed, and the water depth and water range of different sub-catchment areas are obtained;

[0017] In step 2, the terrain data is the slope vector data converted from DEM elevation data, the economic and population density data is the text data obtained from the statistical data of the yearbook of the study area, and the landscape data is the vector data obtained by analyzing the remote sensing image.

[0018] In step 2, the sub-catchment division uses the rainwater pump station data points in the obtained urban drainage planning data, and uses the neighborhood analysis tool in ArcGIS to create Thiessen polygons to obtain sub-catchment zoning; the calculation of water depth and water range follows the principle of passive flooding, considering that the rainfall in the region is uniform, and all points in low-lying areas are likely to be flooded; the surface runoff volume is taken as the total volume of the flooded water body, and the following formula is calculated:

[0019]

[0020] In the formula: Q ij is the yield of the i-th row and j-th column grid cell after grid division, unit is m 3 ; Qtotal is the total yield in the catchment area, unit is m 3 .

[0021] In step 2, the surface runoff of different catchments under different rainfall return periods is extracted in ArcGIS, and combined with DEM data, the flooding depth of each pixel is calculated according to the principle of "low-lying point flooding", and the flooding range under different rainfall return periods is calculated. For easy analysis, the urban waterlogging water depth map under different rainfall return periods is resampled to a 250m×250m grid.

[0022] Step 3, based on the land use data obtained in step 1, extract the data of land use types of forest and grassland, merge them into urban green space, and use Guidos software to perform morphological spatial pattern analysis (MSPA) to obtain urban green space macro-pattern analysis map;

[0023] In step 3, the macro spatial pattern analysis is performed by MSPA, and the city green space is divided into different forms by calculating the shape and size of the city green space layout structure; the city green space is taken as the foreground data, and the rest of the land use types are taken as the background data, and finally it is divided into 7 green space forms, namely core area, island, pore, edge area, ring, bridge area and branch. For easy analysis, the obtained urban green space macro-pattern analysis map is resampled to a 250m×250m grid.

[0024] Step 4, based on the land use data obtained in step 1, extract the data of land use types of forest and grassland, merge them into urban green space, and use Fragstats 4 software to calculate the landscape pattern index at the type level to obtain the urban green space micro-form analysis map.

[0025] In step 4, micro green space morphology analysis is carried out by using landscape pattern index, in order to explore the present distribution pattern of green space, reflect the type level of green space landscape, CA (patch area), PD (patch density), NP (patch number), LPI (the proportion of the largest patch area), AI (patch aggregation degree), COHESION (patch cohesion) and PLAND (the proportion of patch area in landscape) are selected as landscape pattern index research; in order to facilitate analysis, the obtained micro morphology analysis diagram of urban green space is reclassified by using natural breakpoint method, each landscape factor data diagram is divided into 5 categories, and the reclassified data diagram is resampled to 250m*250m grid.

[0026] In step 5, the waterlogging water depth value obtained in step 2 and the macro pattern of urban green space obtained in step 3 and the micro morphology analysis diagram obtained in step 4 are matched and input into the geographic detector software, single factor detection and factor interaction detection are carried out, the influence of macro pattern and micro morphology on waterlogging water depth under different rainfall return periods is obtained, and the influence of micro morphology interaction combination factor on waterlogging water depth is further analyzed.

[0027] In step 5, the waterlogging water depth value is kept as the dependent variable, the macro landscape pattern analysis result, CA (patch area), PD (patch density), NP (patch number), LPI (the proportion of the largest patch area), AI (patch aggregation degree), COHESION (patch cohesion) and PLAND (the proportion of patch area in landscape) are reclassified in ArcGIS and converted into type quantity as the independent variable, the influence of macro pattern and micro morphology on waterlogging water depth under single factor, the influence of micro morphology interaction combination factor on waterlogging water depth are analyzed.

[0028] Firstly, the soil type data, land use data and rainfall data required for simulating urban waterlogging surface runoff are collected, and based on the obtained urban drainage planning, the subcatchment is further divided, so that the waterlogging depth and waterlogging range of each subcatchment are obtained in combination with the elevation DEM data; then, the landscape of the research range is analyzed, the macro pattern of urban green space is analyzed by using the MSPA method, and the micro morphology of urban green space is analyzed by using the landscape pattern index, the macro and micro urban green space landscape analysis results are obtained; finally, the geographic detector software is used to couple and analyze the waterlogging water depth and the macro pattern and micro landscape morphology, and the factor with the greatest influence on the waterlogging water depth under the spatial differentiation of the landscape pattern is obtained. The method system of urban landscape analysis is established, and the exploration method of the landscape influence factor of the urban area suffering from waterlogging is provided. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 is the flow chart of the urban waterlogging risk influencing factor exploration method based on landscape analysis of the present application.

[0030] Figure 2 is the water depth value map obtained by simulating waterlogging water depth in different rainfall return periods in the present application, wherein the left side from top to bottom is 10a, 5a and 2a, and the right side from top to bottom is 100a, 50a and 20a.

[0031] Figure 3 is the landscape factor map obtained by analyzing the macroscopic pattern of urban green space in the present application.

[0032] Figure 4 is the map of each landscape factor (patch area, patch density, patch number, proportion of area occupied by the largest patch, patch aggregation degree, patch cohesion degree and proportion of landscape area occupied by the patch) obtained by analyzing the microscopic form of urban green space in the present application. DETAILED DESCRIPTION

[0033] The technical solutions of the present application will be further described below through specific examples.

[0034] An urban waterlogging risk influencing factor exploration method based on landscape analysis is carried out according to the following steps:

[0035] Step 1, collect the data required for constructing the SCS-CN model, including soil type data, land use data and rainfall data, and then simulate in ArcGIS to obtain the surface runoff value in the study area under different rainfall return periods.

[0036] Step 11, collect data and unify the data into 30m x 30m accuracy: the soil type data comes from the Harmonized World Soil Database (v1.2) data set provided by the United Nations Food and Agriculture Organization (http: / / www.fao.org / soils-portal / en / ); the land use data comes from the Landsat8 OLI_TIRS remote sensing data set provided by the geographic spatial data cloud (http: / / www.gscloud.cn / ), and the land use situation of the study area is obtained by interpreting the ENVI remote sensing analysis software; the rainfall data is determined according to the rainfall intensity formula in the “Tianjin Sponge City Construction Technical Guidelines” according to the division standard of Tianjin storm intensity zoning, which is as follows:

[0037]

[0038] In the formula, q is the design storm intensity, the unit is L / (s·hm2 ); t is the rainfall duration, in min; P1 is the rainfall return period, in years. In this study, the rainfall duration is set to 120 min, and the rain peak position coefficient is set to 0.333. The design rainfall return periods P are set to 2a, 5a, 10a, 20a, 50a and 100a, respectively, to simulate six different rainfall scenarios.

[0039] Step 12, construction of the SCS-CN model: the SCS-CN model is obtained according to the relationship between rainfall, runoff and water storage in the soil, and the rainfall runoff formula is as follows:

[0040]

[0041] In the formula: S is the maximum water storage value of the soil when different soil types and land use modes jointly act, in mm; Q is the actual surface runoff after experiencing rainfall, in mm; P is the designed or actual rainfall, in mm.

[0042] Step 121, the calculation formula of S is as follows:

[0043]

[0044] In the formula: CN is an empirical value obtained according to different soil types, land use modes and soil moisture conditions (AMC). The soil moisture condition is determined according to the soil moisture condition in the 5 days before rainfall, and the present application only considers the soil condition under normal circumstances. Combined with the obtained soil type data and land use mode data, the CN value of this study is as follows:

[0045] Table 1 CN value table

[0046]

[0047] Step 2, according to the urban drainage planning and design in the research scope, the sub-catchment area is divided, combined with the DEM elevation data, and the principle of "low-lying submergence" is followed, the total surface runoff in the sub-catchment area is distributed, and the water depth and water range of different sub-catchment areas are obtained.

[0048] Step 21, the sub-catchment area is divided by using the rainwater pump station data points in the obtained urban drainage planning data, and the Thiessen polygon is created by using the neighborhood analysis tool in ArcGIS to obtain the sub-catchment area partition.

[0049] Step 22, the calculation of water depth and water range follows the principle of passive submergence, and considers that the regional uniform rainfall can submerge all points in low-lying places.

[0050] Step 221, the total amount of surface runoff is calculated as the total volume of the flooded water body as follows:

[0051]

[0052] Qtotal = ååååQij ij Qij is the yield of the i-th row and j-th column grid unit after grid division, with the unit of m 3 ; Qtotal is the total yield of the catchment area, with the unit of m 3 .

[0053] Step 222, in ArcGIS, the surface runoff of different catchment areas under different rainfall return periods is extracted, and combined with DEM data, the inundation depth of each pixel is calculated according to the principle of "low-lying point inundation", and the inundation range under different rainfall return periods is calculated. For easy analysis, the urban waterlogging water depth map under different rainfall return periods is resampled to a grid of 250m x 250m ( Figure 2 ).

[0054] Step 3, based on the land use data obtained in step 1, the data of land use types of forest land and grassland are extracted, which are combined as urban green space, and morphological spatial pattern analysis (MSPA) is carried out by using Guidos software to obtain the macroscopic pattern analysis map of urban green space.

[0055] Step 31, taking the urban green space as the foreground data and the rest of the land use types as the background data, it is finally divided into 7 green space forms, namely core area, island, pore, edge area, ring, bridge area and branch.

[0056] Step 32, for easy analysis, the obtained macroscopic pattern analysis map of urban green space is resampled to a grid of 250m x 250m ( Figure 3 ).

[0057] Step 4, based on the land use data obtained in step 1, the data of land use types of forest land and grassland are extracted, which are combined as urban green space, and the landscape pattern index on the type level is calculated by using Fragstats 4 software to obtain the microscopic form analysis map of urban green space.

[0058] Step 41, based on land use data, it is divided into 2 categories in ArcGIS - green area and non-green area, the area of non-green area is assigned a background value 0, and the urban green area is assigned a value 1, to obtain the processed urban green map, and the size of the study area is determined to determine the size of the fishing net grid, the processed urban green map is divided by using the fishing net grid, and then imported into Fragstats 4 to calculate the landscape pattern index, including but not limited to CA (patch area), PD (patch density), NP (patch number), LPI (the proportion of the largest patch area), AI (patch aggregation degree), COHESION (patch cohesion) and PLAND (patch area proportion in the landscape).

[0059] Step 42, for the convenience of analysis, the obtained urban green micro-morphology analysis map is reclassified by using the natural breakpoint method, each landscape factor data map is divided into 5 categories, and the reclassified data map is resampled to a grid of 250m*250m. Figure 4

[0060] Step 5, the waterlogging depth value obtained in step 2 and the urban green macro pattern obtained in step 3 and the urban green micro-morphology analysis map obtained in step 4 are matched and input into the geographic detector software, single factor detection and factor interaction detection are carried out, the influence of the macro pattern and the micro-morphology under different rainfall return periods on the waterlogging depth is obtained, and the influence of the micro-morphology interaction combination factor on the waterlogging depth is further analyzed.

[0061] Step 51, differentiation and factor detection: mainly used to explore the spatial differentiation of the dependent variable, and to detect the influence of the independent variable on the dependent variable. The q value of the single factor is expressed as follows:

[0062]

[0063] In the formula: h=1,...,L is the stratification of the dependent variable or the independent variable; N h And N are the number of units in h layer and the whole study area; and are the variances of the dependent variable in h layer and the whole study area.

[0064] The single factor detection results of the macro pattern and the micro-morphology are shown in Table 2 and Table 3.

[0065] Table 2 Macro pattern factor detection q value under different rainfall return periods

[0066]

[0067] Table 3 Micro-morphology factor detection q value under different rainfall return periods

[0068] ​

[0069] Step 52, interaction detection: used to explore the strength of the influence of two different independent variables on the dependent variable through interaction. The interaction judgment basis table is as follows.

[0070] Table 4 judgment basis table of interaction type of two independent variables on dependent variable

[0071]

[0072]

[0073] Note: Assuming that the q values of two independent variables are q(X1) and q(X2), the q value after interaction is q(X1∩X2)

[0074] The micro-morphology interaction factor detection results of the present application are shown in Table 5, and Table 5 only shows part of the data to show the feasibility of the present application.

[0075] Table 5 interaction detection results q value at 5a

[0076] A B A∩B Results A B A∩B Results X1 X2 0.64％ ↑↑ X3 X6 1.96％ ↑↑ X1 X3 0.85％ ↑↑ X3 X7 0.48％ ↑ X1 X4 0.77％ ↑↑ X4 X5 0.67％ ↑↑ X1 X5 1.21％ ↑↑ X4 X6 2.40％ ↑↑ X1 X6 1.63％ ↑↑ X4 X7 0.30％ ↑ X1 X7 0.83％ ↑↑ X5 X6 3.20％ ↑↑ X2 X3 0.51％ ↑ X5 X7 0.97％ ↑↑ X3 X5 0.97％ ↑↑ X6 X7 2.29％ ↑↑

[0077] Note: The "↑" in the table indicates that the two independent variables are enhanced, and "↑↑" indicates that the two independent variables are nonlinearly enhanced.

[0078] The landscape factor exploration of the present application found that when exploring the landscape space pattern, it was found that the explanation degree of the differentiation phenomenon of the urban green space pattern on the water depth pattern was weak; when exploring the driving force of the single factor of the landscape factor, it was found that the dominant type of PD was the strongest, and with the increase of the rainfall return period, the change of the driving factor was different: the influence of AI, CA, COHESION, LPI and PLAND factors gradually increased; the influence of PD gradually weakened; the change of NP was not obvious. Combined with NP, PD and COHESION landscape factors, it was found that the fragmentation of urban green space landscape and the connectivity of landscape had the greatest interactive influence on the urban water depth.

[0079] The above has made an exemplary description of the present application, and it should be explained that without departing from the core of the present application, any simple modification, modification or other equivalent replacement which can not cost creative labor of those skilled in the art falls within the protection scope of the present application.

Claims

1. A method for exploring urban waterlogging risk influencing factors based on landscape analysis, comprising the following steps: Step 1: Collect data required for constructing the SCS-CN model, including soil type data, land use data, and rainfall data, then simulate in ArcGIS to obtain the surface runoff value in the study area under different rainfall return periods; Step 2: Divide the sub-catchment according to the urban drainage planning design in the study area, combine with DEM elevation data, follow the principle of "low-lying submergence", and distribute the total surface runoff in the sub-catchment to obtain the waterlogging depth and waterlogging range of different sub-catchments; Step 3: Based on the land use data obtained in Step 1, extract the data of land use types of forest and grassland, merge them into urban green space, and use Guidos software for morphological spatial pattern analysis (MSPA) to obtain the macro-pattern analysis graph of urban green space; Step 4: Based on the land use data obtained in Step 1, extract the data of land use types of forest and grassland, merge them into urban green space, and use Fragstats 4 software to calculate the landscape pattern index at the type level to obtain the micro-morphology analysis graph of urban green space; Step 5: Match the waterlogging depth value obtained in Step 2 with the macro-pattern of urban green space obtained in Step 3 and the micro-morphology analysis graph of urban green space obtained in Step 4, and input them into the geographic detector software for single factor detection and factor interaction detection to obtain the influence of macro-pattern and micro-morphology single factors on waterlogging depth under different rainfall return periods, and further analyze the influence of micro-morphology interaction combination factors on waterlogging depth; Keep the waterlogging depth value as the dependent variable, input the macro-landscape pattern analysis results, CA patch area, PD patch density, NP patch number, LPI maximum patch area ratio, AI patch aggregation degree, COHESION patch cohesion, and PLAND patch area ratio in ArcGIS for recategorization operation to convert them into type quantity as the independent variable, analyze the influence of macro-pattern and micro-morphology on waterlogging depth under single factor, and the influence of micro-morphology interaction combination factors on waterlogging depth.

2. The method for exploring urban flooding risk influencing factors based on landscape analysis according to claim 1, characterized in that, In Step 1, the data is collected and unified to 30m x 30m precision.

3. The method of claim 1, wherein, In Step 1, the soil type data comes from the Harmonized World Soil Database v1.2 dataset provided by the United Nations Food and Agriculture Organization; the land use data comes from the Landsat8 OLI_TIRS remote sensing dataset provided by the Geographic Spatial Data Cloud, which is interpreted by ENVI remote sensing analysis software to obtain the land use situation of the study area; the rainfall data is determined according to the rainfall intensity zoning standard in the "Tianjin Sponge City Construction Technical Guidelines", and the formula is as follows: In the formula, q is the design storm intensity, L / (s·hm 2 ); t is the rainfall duration, min; and P1 is the rainfall return period, year.

4. The method of claim 1, wherein, In step 1, the SCS-CN model is obtained according to the relationship between rainfall, runoff and water storage in the soil, and the rainfall runoff formula is as follows: In the formula: S is the maximum water storage value of the soil when different soil types and land use methods jointly act, with unit of mm; Q is the actual surface runoff after experiencing rainfall, with unit of mm; P is the designed or actual rainfall, with unit of mm; The calculation formula of S is as follows: In the formula: CN is an empirical value obtained according to different soil types, land use methods and soil moisture conditions (AMC).

5. The method of claim 1, wherein, In step 2, the sub-catchment area is divided by using the rainwater pumping station data points in the obtained urban drainage planning data, and the Thiessen polygon is created by using the neighborhood analysis tool in ArcGIS to obtain the sub-catchment area division; the calculation of water depth and water range follows the principle of passive flooding, and it is considered that the uniform rainfall in the region may be flooded; the surface flow accumulation is taken as the total volume of the flooded water body, and the following formula is calculated: In the formula, Q ij is the yield of the i-th row and j-th column grid unit after grid division, with the unit of m 3 ; Qtotal is the total yield in the catchment area, with the unit of m 3 ; the surface runoff of different catchment areas under different rainfall return periods is extracted in ArcGIS, and the submerged depth of each pixel is calculated by following the principle of "low-lying point submergence" in combination with DEM data, the submerged range under different rainfall return periods is calculated, and the urban waterlogging depth map under different rainfall return periods is resampled to a 250m×250m grid.

6. The method of claim 1, wherein, In step 3, the MSPA is used for macro spatial pattern analysis, and the urban green space is divided into different forms by calculating the shape and size of the urban green space layout structure; the urban green space is taken as the foreground data, and the remaining land use types are taken as the background data, and finally it is divided into 7 green space forms, namely core area, island, pore, edge area, ring, bridge area and branch, and the obtained macro pattern analysis diagram of urban green space is resampled to 250m×250m grid.

7. The method of claim 1, wherein, In step 4, the landscape pattern index is used for micro green space form analysis, in order to explore the present distribution pattern of green space and reflect the type level of green space landscape pattern, CA patch area, PD patch density, NP patch number, LPI maximum patch area ratio, AI patch aggregation degree, COHESION patch cohesion and PLAND patch area ratio in landscape are selected as landscape pattern index research; the obtained micro form analysis diagram of urban green space is reclassified by using the natural breakpoint method, and each landscape factor data diagram is divided into 5 categories, and the reclassified data diagram is resampled to 250m×250m grid.

Images