Main-subgrid flood inundation area high-precision reconstruction method and computer equipment

Abstract

The application discloses a main-subgrid flood inundation area high-precision reconstruction method and computer equipment, and the method comprises the following steps: constructing a main grid system, equally dividing each main grid cell in the main grid system into four subgrid cells to form a subgrid system; taking the main grid as a calculation unit, constructing a two-dimensional water dynamic model, adopting an alternating implicit ADI algorithm for numerical solution, optimizing the model by calibrating the roughness parameter, and obtaining water level field and / or flow velocity field calculation results; realizing water level field and / or flow velocity field reconstruction based on a bilinear interpolation algorithm; setting a dynamic threshold to judge the subgrid cell inundation condition; and adopting four grid traversal sequences, i.e., forward scanning, reverse scanning, column-priority forward scanning and column-priority reverse scanning, to generate an inundation area set through multiple scanning. The application inserts low-resolution calculation data into a high-resolution grid, reconstructs a flood inundation area with high precision, and accurately presents the flood inundation range and related states.

Description

Technical Field

[0001] This invention relates to the field of flood simulation and calculation technology, and more specifically to a high-precision reconstruction method and computer equipment for main-subgrid flood inundation areas for hydrodynamic models. Background Technology

[0002] As a frequent type of natural disaster that poses a serious threat to human security, accurate simulation of the inundation process of floods is crucial for improving the timeliness of flood warnings, optimizing risk assessment systems, and improving flood prevention and disaster reduction decision-making. With the iterative development of computing technology, hydrodynamic models have become the core means of flood evolution simulation. However, high-precision flood simulation often requires high-resolution computing grids, which places extremely high demands on computing resources.

[0003] In traditional flood simulation methods, directly using high-resolution grids for global calculations leads to an exponential increase in computational load, making real-time or efficient computation difficult under current computing power conditions. While using low-resolution grids can improve computational efficiency, it cannot accurately capture the detailed features of the flood-inundated area, such as the evolution of water flow in complex terrain and the localized inundation around buildings, resulting in insufficient accuracy in the simulation results.

[0004] To balance computational efficiency and simulation accuracy, existing technologies often employ local mesh refinement strategies. However, these methods are prone to numerical oscillations in complex terrain areas or where the mesh scale changes abruptly, and the refinement strategies are difficult to adapt to the universal requirements of different geomorphic units. Furthermore, traditional interpolation algorithms, when mapping low-resolution data to high-resolution meshes, often fail to fully consider topographic gradients and water flow direction constraints, leading to a loss of spatial continuity in the data. This results in discrete noise points or discontinuous regions in the identification of inundation areas, affecting the reliability of the simulation results.

[0005] In terms of inundation state determination, existing models mostly employ fixed threshold judgment mechanisms, which cannot be dynamically adjusted based on local terrain features or the inundation state of neighboring areas. This results in insufficient adaptability to different geomorphic conditions, such as mountainous watersheds and plain river networks. Furthermore, different grid traversal orders may cause the inundated area to grow excessively in a certain direction, affecting the accuracy of the inundation range simulation.

[0006] Therefore, how to achieve high-precision mapping from low-resolution computational data to high-resolution grids under limited computing power, while ensuring the spatial continuity and physical rationality of the flooded area simulation, is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0007] The present invention provides a high-precision reconstruction method and computer equipment for a main-subgrid flood-inundated area that effectively balances computational efficiency and computational accuracy, which can at least solve one of the above-mentioned technical problems.

[0008] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0009] A high-precision reconstruction method for flood-inundated areas using a master-subgrid approach includes the following steps:

[0010] S1. Construct the main grid system. Based on the quadtree recursive splitting strategy, divide each main grid cell in the main grid system into four sub-grid units at equal intervals to form a sub-grid system.

[0011] S2. Using the main grid as the computational unit, a two-dimensional hydrodynamic model is constructed based on the two-dimensional shallow water equation. The alternating implicit ADI algorithm is used for numerical solution, and the model is optimized by calibrating the roughness parameter to obtain the calculation results of the water level field and / or flow velocity field, providing basic data for subgrid interpolation.

[0012] S3. Reconstruct the water level field and / or flow velocity field based on the bilinear interpolation algorithm;

[0013] S4. Set a dynamic threshold to determine the flooding status of subgrid cells, and suppress isolated flooding points caused by interpolation noise through parameter optimization;

[0014] S5. Four grid traversal orders are used: forward scan, reverse scan, column-first forward scan, and column-first reverse scan. Multiple scans are performed to generate a set of flooded areas. The frequency of each subgrid cell being marked as "flooded" is counted, and cells with a frequency ≥ 50% are retained to avoid the flooded area from growing in a certain direction due to different scanning methods.

[0015] Furthermore, S1 further includes:

[0016] S11. Define the hierarchical structure of the main grid and subgrids. Based on the quadtree recursive splitting strategy, divide each main grid cell into four subgrid units at equal intervals along the horizontal and vertical axes to form a grid system with parent-child mapping relationship.

[0017] S12. Generate the main grid fishing net vector file, set the starting coordinates and row and column spacing, and generate a dataset containing grid center point features;

[0018] S13. Convert the main grid fishing net into point features, extract the elevation values ​​of each grid point based on the corresponding precision digital elevation model, and store them in the attribute table;

[0019] S14. Perform quadtree partitioning on each main grid cell, generate four subgrids by calculating the midpoint of the vertex coordinates, assign IDs to the subgrids in row priority order, and make them inherit the ID prefix of the main grid to establish spatial association.

[0020] S15. Convert the subgrid fishing net into point features, extract the elevation values ​​of each subgrid point based on a higher-precision digital elevation model, and map the main grid ID attribute through spatial association tools to ensure the consistency of parent and child grid topology.

[0021] S16. Link the elevation attributes of the main grid and subgrid point features to the corresponding fishing net vector file to generate digital elevation models of different precision, and verify the grid topology integrity and the consistency between the elevation data and the measured terrain.

[0022] Furthermore, S2 further includes:

[0023] S21. Perform registration, cropping and resampling operations on land use type data to make the data coordinate system consistent with the cell size and digital terrain model, and create model partition roughness files according to land use type.

[0024] S22. Based on the master grid system and model partition roughness file, a two-dimensional hydrodynamic model is developed using the alternating implicit ADI algorithm and OpenMP parallel computing technology.

[0025] S23. Based on the measured flow rate and water level boundary conditions, optimize the roughness value of the model, reenact the flood inundation situation in a typical year, and complete the parameter calibration of the hydrodynamic model by comparing the measured inundation range data.

[0026] Furthermore, the development process of the two-dimensional hydrodynamic model in S22 is as follows:

[0027] S221. The two-dimensional Saint-Venant equations are adopted as the fundamental equations for hydrodynamic calculations. In Cartesian coordinates, the vector form of these equations is expressed as follows:

[0028]

[0029]

[0030]

[0031]

[0032]

[0033] Of these five styles, The horizontal coordinate is the coordinate of the space. The vertical coordinate is the coordinate of the space. For time, For the source term vector, A variable vector, including water level and unit width flow , Both F and G are flux vectors, including water depth h. and Directions and The flow rate on the surface, , is the Xie Cai coefficient. This is the Manning coefficient;

[0034] S222. The mass and momentum equations are discretized using the finite difference method with alternating implicit and explicit methods, and the equation matrices generated in each direction are solved using the chasing method.

[0035] Furthermore, S3 further includes:

[0036] S31. Normalize the interpolation point coordinates to obtain the relative position parameters of the interpolation point coordinates in the horizontal and vertical directions within the main grid cell. This process is expressed as follows:

[0037]

[0038]

[0039] In these two equations, (x0, y0) are the coordinates of the center point of the lower left main grid cell, (x1, y0) are the coordinates of the center point of the lower right main grid cell, (x0, y1) are the coordinates of the center point of the upper left main grid cell, (x1, y1) are the coordinates of the center point of the upper right main grid cell, and (x, y) are the coordinates of the point to be interpolated. and The value range of is [0, 1], which represents the relative position of the interpolation point in the horizontal and vertical directions, respectively;

[0040] S32. Perform two linear interpolations in the horizontal direction. The specific process is as follows:

[0041] At the base of y=y0, with respect to Z 00 and Z 10 Perform linear interpolation to obtain the intermediate value Z. 底 :

[0042]

[0043] At the top edge of y=y1, for Z 01 and Z 11 Perform linear interpolation to obtain the intermediate value Z. 顶 :

[0044]

[0045] In these two equations, Z 00 Given the known water level value at the center point (x0, y0) of the lower left main grid cell, Z 10 Z represents the known water level value at the center point (x1, y0) of the lower right main grid cell.01 Given the known water level value at the center point (x0, y1) of the top-left main grid cell, Z 11 The center point of the main grid cell in the upper right corner ( , The known water level value;

[0046] S33. In the vertical direction, for Z 底 and Z 顶 Perform linear interpolation to obtain the final interpolation result Z(x, y):

[0047] ;

[0048] S34. Combine the expressions in S32 and S33 to obtain a unified expression for calculating the water level of the target interpolation point:

[0049] .

[0050] Furthermore, S4 further includes:

[0051] S41. Taking each sub-grid cell as the center, extract the 8-neighborhood range, that is, the 8 adjacent cells in the up, down, left, right and four diagonal directions, and construct the neighborhood analysis window.

[0052] S42. Count the number of units marked as flooded in the 8-neighborhood and calculate the proportion of the number of flooded units to the total number of units in the 8-neighborhood.

[0053] S43. For each subgrid cell, compare its calculated water level with the terrain elevation plus a dynamic threshold. If the condition is met, mark it as "submerged".

[0054] When the flooding rate is ≥30%, it is determined that there is a trend of large-scale flooding, and a low threshold is adopted. At this point, the water level needs to be higher than the terrain elevation + Only then is it marked as submerged;

[0055] When the flooding rate is less than 30%, it is considered to have an isolated point risk, and a high threshold is applied. , At this point, the water level needs to be higher than the terrain elevation + Only then is it marked as submerged;

[0056] S44, Calibration based on historical flood measurement data and The specific value of the threshold difference is chosen to ensure that it can suppress interpolation noise while preserving the true flooding boundary.

[0057] Furthermore, S5 further includes:

[0058] S51. Determine the four grid traversal orders. The specific process is as follows:

[0059] Forward scanning, row priority: Access cells in ascending order of row and column from top left to bottom right. The traversal path is (i,j) → (i,j+1) → (i+1,0), where i is the row number and j is the column number.

[0060] Reverse scan, row-first: Access cells in descending order of row and column from bottom right to top left. The traversal path is (i,j) → (i,j-1) → (i-1,M-1), where... The total number of columns;

[0061] Column-first forward scan: Access cells in ascending order of column and row from top left to bottom right. The traversal path is (i,j) → (i+1,j) → (0,j+1).

[0062] Column-first reverse scan: Access cells in descending column and row order from bottom right to top left. The traversal path is (i,j) → (i-1,j) → (N-1,j-1), where... Total number of rows;

[0063] S52. Perform multiple scans and frequency statistics. The specific process is as follows:

[0064] The grid is scanned multiple times using the four sequences described above, with each scan generating a flooding marker matrix F. t (i,j), F t (i,j)=1 indicates submersion, F t (i,j)=0 indicates no flooding. The flooding frequency of each sub-grid cell is calculated using the following formula:

[0065]

[0066] Where F(i,j) is the frequency matrix, reflecting the frequency at which the cell (i,j) is marked as submerged in multiple scans;

[0067] S53. Select and retain submerged cells. The specific process is as follows:

[0068] A robust screening of flooded cells is implemented using a decision rule with a frequency threshold of ≥50%. A four-directional scanning mechanism (row, column, forward, and reverse) is used to prevent the flooded area from growing excessively in a single direction, thus ensuring the spatial consistency of the extracted flooded area results.

[0069] A computer device, including a memory and a processor, wherein the memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the above-described high-precision reconstruction method for the main-subgrid flood-inundated area.

[0070] The beneficial effects of this invention are reflected in:

[0071] 1. A master-subgrid system was constructed using a quadtree recursive splitting strategy, ensuring the consistency of parent and child grid topologies. The reliability of the grid system was guaranteed by verifying the consistency between elevation data and measured terrain, laying a solid foundation for subsequent hydrodynamic calculations and flood zone reconstruction.

[0072] 2. By constructing a hierarchical structure of the main grid and subgrids and using a bilinear interpolation algorithm, low-resolution computational data can be interpolated to a high-resolution grid, which can accurately represent the flood inundation range, water level field and other conditions. Compared with traditional low-resolution models, this significantly improves the accuracy of flood area reconstruction.

[0073] 3. By setting a threshold that is dynamically adjusted based on the local neighborhood status and terrain features, and by optimizing the neighborhood inundation ratio parameter, isolated inundation points caused by interpolation noise are effectively suppressed, the formation of large inundation areas is encouraged, and the inundation judgment is more in line with the actual flood evolution law, thus avoiding misjudgment caused by noise interference.

[0074] 4. The flooded area set is generated by scanning multiple times using four different grid traversal orders. By statistically analyzing the flooding frequency and retaining high-frequency units, the growth of the flooded area in a single direction caused by different scanning methods is avoided, ensuring the objectivity and accuracy of the flooded area boundary. Compared with a single scanning method, the results are more reliable.

[0075] 5. The alternating implicit ADI algorithm and OpenMP parallel computing technology are adopted in the hydrodynamic model solution. While ensuring the calculation accuracy, the solution efficiency is improved, which is suitable for scenarios with limited computing power. Compared with traditional algorithms, the calculation time is reduced and the practicality of the model is improved. Attached Figure Description

[0076] The accompanying drawings, which are provided to further illustrate this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application.

[0077] Figure 1 This is a schematic diagram of the overall process of the method according to an embodiment of the present invention.

[0078] Figure 2 This is a schematic diagram illustrating the specific execution flow of the method according to an embodiment of the present invention.

[0079] Figure 3 This is a schematic diagram of geographic information data (GIS) of the research area according to a specific embodiment of the present invention.

[0080] Figure 4 This is a schematic diagram of the main grid and subgrid system topology according to a specific embodiment of the present invention.

[0081] Figure 5 This is a schematic diagram of land use in the study area according to a specific embodiment of the present invention.

[0082] Figure 6 This is a schematic diagram of the calculation process of a two-dimensional hydrodynamic model according to a specific embodiment of the present invention.

[0083] Figure 7 This is a comparison diagram of the water level at the outlet gate station, constructed using a two-dimensional hydrodynamic model according to a specific embodiment of the present invention.

[0084] Figure 8 This is a schematic diagram comparing the water level and inundation range 17 hours into the flood process, according to a specific embodiment of the present invention.

[0085] Figure 9 This is a schematic diagram comparing the water level and inundation range 39 hours into the flood process, according to a specific embodiment of the present invention.

[0086] Figure 10 This is a schematic diagram comparing the overall rendering effects of the subgrid (20m) and the main grid (40m) at water depths of 10h, 30h, 50h and 80h during flooding, according to a specific embodiment of the present invention.

[0087] Figure 11 This is a rendering detail of the calculation results of a 40m master grid according to a specific embodiment of the present invention.

[0088] Figure 12 This is a rendering detail of the 20m sub-mesh processing result according to a specific embodiment of the present invention.

[0089] Figure 13 This is a rendering detail of the 20m grid calculation results according to a specific embodiment of the present invention.

[0090] Figure 14 This is a rendering detail of the calculation results of the 40m main grid under the flooding state of 58 hours in a specific embodiment of the present invention.

[0091] Figure 15 This is a rendering detail of the calculation results of a 20m subgrid under a flood inundation state of 58 hours in a specific embodiment of the present invention.

[0092] Figure 16 This is a rendering detail of the calculation results of a 20m grid under a flood inundation state of 58 hours in a specific embodiment of the present invention.

[0093] Figure 17 This is a structural block diagram of a computer device according to an embodiment of the present invention. Detailed Implementation

[0094] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0095] It should be noted that the meaning of "and / or" throughout the text includes three parallel solutions. Taking "A and / or B" as an example, it includes solution A, solution B, or a solution that simultaneously satisfies A and B. Furthermore, "multiple" refers to two or more. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0096] See Figures 1-2 This invention provides a high-precision reconstruction method for a main-subgrid flood-inundated area, comprising the following steps:

[0097] S1. Construct the main grid system. Based on the quadtree recursive splitting strategy, divide each main grid cell in the main grid system into four sub-grid units at equal intervals to form a sub-grid system.

[0098] S2. Using the main grid as the computational unit, a two-dimensional hydrodynamic model is constructed based on the two-dimensional shallow water equation. The alternating implicit ADI algorithm is used for numerical solution, and the model is optimized by calibrating parameters such as roughness to obtain the calculation results of water level field and / or flow velocity field, providing basic data for subgrid interpolation.

[0099] S3. Reconstruct the water level field and / or flow velocity field based on the bilinear interpolation algorithm;

[0100] S4. Set a dynamic threshold to determine the flooding status of subgrid cells, and suppress isolated flooding points caused by interpolation noise through parameter optimization;

[0101] S5. Four grid traversal orders are used: forward scan, reverse scan, column-first forward scan, and column-first reverse scan. Multiple scans are performed to generate a set of flooded areas. The frequency of each subgrid cell being marked as "flooded" is counted, and cells with a frequency ≥ 50% are retained to avoid the flooded area from growing in a certain direction due to different scanning methods.

[0102] In this embodiment, S1 further includes:

[0103] S11. Define the hierarchical structure of the main grid and subgrids. Based on the quadtree recursive splitting strategy, divide each main grid cell into four subgrid units at equal intervals along the horizontal and vertical axes to form a grid system with parent-child mapping relationship.

[0104] S12. Generate the main grid fishing net vector file, set the starting coordinates and row and column spacing, and generate a dataset containing grid center point features;

[0105] S13. Convert the main grid fishing net into point features, extract the elevation values ​​of each grid point based on the corresponding precision digital elevation model, and store them in the attribute table;

[0106] S14. Perform quadtree partitioning on each main grid cell, generate four subgrids by calculating the midpoint of the vertex coordinates, assign IDs to the subgrids in row priority order, and make them inherit the ID prefix of the main grid to establish spatial association.

[0107] S15. Convert the subgrid fishing net into point features, extract the elevation values ​​of each subgrid point based on a higher-precision digital elevation model, and map the main grid ID attribute through spatial association tools to ensure the consistency of parent and child grid topology.

[0108] S16. Link the elevation attributes of the main grid and subgrid point features to the corresponding fishing net vector file to generate digital elevation models of different precision, and verify the grid topology integrity and the consistency between the elevation data and the measured terrain.

[0109] In this embodiment, S2 further includes:

[0110] S21. Perform registration, cropping and resampling operations on land use type data to make the data coordinate system consistent with the cell size and digital terrain model, and create model partition roughness files according to land use type.

[0111] S22. Based on the master grid system and model partition roughness file, a two-dimensional hydrodynamic model is developed using the alternating implicit ADI algorithm and OpenMP parallel computing technology.

[0112] S23. Based on the measured flow rate and water level boundary conditions, optimize the roughness value of the model, reenact the flood inundation situation in a typical year, and complete the parameter calibration of the hydrodynamic model by comparing the measured inundation range data.

[0113] In this embodiment, the development process of the two-dimensional hydrodynamic model in S22 is as follows:

[0114] S221. The two-dimensional Saint-Venant equations are adopted as the fundamental equations for hydrodynamic calculations. In Cartesian coordinates, the vector form of these equations is expressed as follows:

[0115]

[0116]

[0117]

[0118]

[0119]

[0120] Of these five styles, The horizontal coordinate is the coordinate of the space. The vertical coordinate is the coordinate of the space. For time, For the source term vector, A variable vector, including water level and unit width flow , Both F and G are flux vectors, including water depth h. and Directions and The flow rate on the surface, , is the Xie Cai coefficient. This is the Manning coefficient;

[0121] S222. The mass and momentum equations are discretized using the finite difference method with alternating implicit and explicit methods, and the equation matrices generated in each direction are solved using the chasing method.

[0122] In this embodiment, S3 further includes:

[0123] S31. Normalize the interpolation point coordinates to obtain the relative position parameters of the interpolation point coordinates in the horizontal and vertical directions within the main grid cell. This process is expressed as follows:

[0124]

[0125]

[0126] In these two equations, (x0, y0) are the coordinates of the center point of the lower left main grid cell, (x1, y0) are the coordinates of the center point of the lower right main grid cell, (x0, y1) are the coordinates of the center point of the upper left main grid cell, (x1, y1) are the coordinates of the center point of the upper right main grid cell, and (x, y) are the coordinates of the point to be interpolated. and The value range of is [0, 1], which represents the relative position of the interpolation point in the horizontal and vertical directions, respectively;

[0127] S32. Perform two linear interpolations in the horizontal direction. The specific process is as follows:

[0128] At the base of y=y0, with respect to Z 00 and Z 10 Perform linear interpolation to obtain the intermediate value Z. 底 :

[0129]

[0130] At the top edge of y=y1, for Z 01 and Z 11 Perform linear interpolation to obtain the intermediate value Z. 顶 :

[0131]

[0132] In these two equations, Z 00 Given the known water level value at the center point (x0, y0) of the lower left main grid cell, Z 10 Z represents the known water level value at the center point (x1, y0) of the lower right main grid cell. 01 Given the known water level value at the center point (x0, y1) of the top-left main grid cell, Z 11 The known water level value is the center point (x1, y1) of the main grid cell in the upper right corner;

[0133] S33. In the vertical direction, for Z 底 and Z 顶 Perform linear interpolation to obtain the final interpolation result Z(x, y):

[0134] ;

[0135] S34. Combine the expressions in S32 and S33 to obtain a unified expression for calculating the water level of the target interpolation point:

[0136] .

[0137] In this embodiment, step S4 further includes:

[0138] S41. Taking each sub-grid cell as the center, extract the 8-neighborhood range, that is, the 8 adjacent cells in the up, down, left, right and four diagonal directions, and construct the neighborhood analysis window.

[0139] S42. Count the number of units marked as flooded in the 8-neighborhood and calculate the proportion of the number of flooded units to the total number of units in the 8-neighborhood.

[0140] S43. For each subgrid cell, compare its calculated water level with the terrain elevation plus a dynamic threshold. If the condition is met, mark it as "submerged".

[0141] When the flooding rate is ≥30%, it is determined that there is a trend of large-scale flooding, and a low threshold is adopted. At this point, the water level needs to be higher than the terrain elevation + Only then is it marked as submerged;

[0142] When the flooding rate is less than 30%, it is considered to have an isolated point risk, and a high threshold is applied. , At this point, the water level needs to be higher than the terrain elevation + Only then is it marked as submerged;

[0143] S44, Calibration based on historical flood measurement data and The specific value of the threshold difference is chosen to ensure that it can suppress interpolation noise while preserving the true flooding boundary.

[0144] In this embodiment, S5 further includes:

[0145] S51. Determine the four grid traversal orders. The specific process is as follows:

[0146] Forward scanning, row priority: Access cells in ascending order of row and column from top left to bottom right. The traversal path is (i,j) → (i,j+1) → (i+1,0), where i is the row number and j is the column number.

[0147] Reverse scan, row-first: Access cells in descending order of row and column from bottom right to top left. The traversal path is (i,j) → (i,j-1) → (i-1,M-1), where... The total number of columns;

[0148] Column-first forward scan: Access cells in ascending order of column and row from top left to bottom right. The traversal path is (i,j) → (i+1,j) → (0,j+1).

[0149] Column-first reverse scan: Access cells in descending column and row order from bottom right to top left. The traversal path is (i,j) → (i-1,j) → (N-1,j-1), where... Total number of rows;

[0150] S52. Perform multiple scans and frequency statistics. The specific process is as follows:

[0151] The grid is scanned multiple times using the four sequences described above, with each scan generating a flooding marker matrix F. t (i,j), F t (i,j)=1 indicates submersion, F t (i,j)=0 indicates no flooding. The flooding frequency of each sub-grid cell is calculated using the following formula:

[0152]

[0153] Where F(i,j) is the frequency matrix, reflecting the frequency at which the cell (i,j) is marked as submerged in multiple scans;

[0154] S53. Select and retain submerged cells. The specific process is as follows:

[0155] A robust screening of flooded cells is implemented using a decision rule with a frequency threshold of ≥50%. A four-directional scanning mechanism (row, column, forward, and reverse) is used to prevent the flooded area from growing excessively in a single direction, thus ensuring the spatial consistency of the extracted flooded area results.

[0156] To further verify the feasibility and superiority of the main-subgrid high-precision reconstruction method for flood-inundated areas, the present invention provides the following specific example for illustration:

[0157] Figure 3 The diagram shown is a GIS map of the study area in this example. Taking area H as an example, area H is an important flood storage area. Its terrain is characterized by a "bowl-shaped" shape, with higher elevations on all sides and a low-lying, flat center. Low hills surround the northwest and southeast, forming natural barriers, while the central area is a vast plain depression crisscrossed by rivers and ditches. The overall terrain is predominantly plain depressions, with some hilly features. This unique topographical feature gives it typical inundation characteristics during flood evolution.

[0158] Figure 4 The diagram shows the topology of the main grid and subgrid system in region H. Figure 4 In the diagram, thick dark lines delineate the main grid cells, forming the main grid system and serving as the macroscopic spatial computation framework. Based on a quadtree recursive splitting strategy, each main grid cell is equidistantly divided into four sub-grid cells by thin light-colored lines, forming a more detailed sub-grid system. This achieves spatial stratification from macro to micro, laying a clear "main-sub" hierarchical spatial foundation for subsequent steps such as two-dimensional hydrodynamic model calculations, interpolation, and inundation analysis. The layout of the sub-grid system is shown in Table 1 below.

[0159] Table 1. Layout of Subgrid System

[0160]

[0161] Figure 5 The diagram shows the land use of the study area. Different colors distinguish various land use types, such as paddy fields, dry land, and ditches, with each type corresponding to a different roughness value. When creating the distributed roughness file, the roughness parameters for each land use type were first defined. Then, based on the spatial distribution of this diagram, the roughness values ​​were assigned to the main and subgrid systems. This provides crucial input for establishing the two-dimensional hydrodynamic model and lays a solid foundation for subsequent parameter calibration and optimization, helping to accurately reproduce the region's hydrodynamic characteristics. The results of the distributed roughness configuration are shown in Table 2 below.

[0162] Table 2. Layout of Distributed Roughness

[0163]

[0164] Figure 6 This is a schematic diagram of the calculation process for the two-dimensional hydrodynamic model. The specific modeling process is as follows: A new model task is created, and the necessary variables such as concentration and water level are initialized. Boundary conditions, including watershed topography, flow rate, and water level, are read. The discrete solution of the hydrodynamic module uses the ADI method in the finite difference method. One time step in the numerical calculation process is divided into two half-steps. Through implicit and explicit alternating directional differences, a tridiagonal linear equation system is obtained. The tridiagonal sparse matrix is ​​then solved quickly using the OpenMP reconstruction chasing method (TDMA), ensuring both stability and high efficiency. The two-dimensional flow field, the vertical and horizontal velocity components of the cells, and water level information are calculated.

[0165] Figure 7 A two-dimensional hydrodynamic model was constructed to simulate the water level comparison results at the outlet gate station. Figure 7 In the comparison of the simulated water level by the self-developed model, the measured water level, and the simulated water level by MIKE21, the simulated water level trend of the self-developed model is highly consistent with the measured data. Although there are slight deviations in some areas, the overall model can accurately reflect the dynamic evolution of the flood. This verification result provides strong support for the calibration process of optimizing parameters such as roughness when constructing a two-dimensional hydrodynamic model based on two-dimensional shallow water equations as described above in S2. It shows that the parameter settings are scientific and reasonable, and can ensure that the basic data such as the water level field and velocity field calculated by the model can provide reliable support for subgrid interpolation.

[0166] Figures 8-9 This diagram illustrates the comparison of water levels and inundation extent during the flood evolution process. By comparing the inundation extent data and visualization results of the models constructed at different flood inflow times with those of MIKE21, it can be seen that the inundation boundary morphology and regional distribution of the two are highly consistent, with a high proportion of shared inundation grids. This verifies that, under the same grid configuration (grid size 20×20m, total number of grids 1,976,800), after parameter calibration, this model can accurately reproduce the hydrodynamic inundation process and has equivalent simulation accuracy to mature commercial models. This demonstrates that, under the synergistic effect of spatial discretization (grid size-number) and parameterization scheme, the model meets the accuracy requirements of hydrodynamic motion simulation and exhibits reliable computational performance and simulation accuracy. The inundation extent data and visualization results of the models constructed at different flood inflow times and those of MIKE21 are shown in Table 3 below.

[0167] Table 3. Data and visualization results of the inundation range of the model constructed at different flood entry times and MIKE21.

[0168]

[0169] Figure 10 The illustration shows the water depth rendering effects of the subgrid (20m) and main grid (40m) at 10h, 30h, 50h and 80h of flood inundation. Due to resolution limitations, the main grid presents relatively coarse water depth details in local complex areas (such as water flow branches and topographic undulations). The subgrid, with its high-resolution cells formed by quadtree recursive splitting, has a water depth distribution and inundation boundary that better match the actual water flow evolution characteristics, and there are no obvious isolated inundation points. This demonstrates the effective suppression of interpolation noise by the dynamic threshold, achieving a balance between computational efficiency and characterization accuracy in flood inundation simulation, and providing strong support for applications such as high-precision flood inundation range assessment and flood control planning.

[0170] Figures 11-16 This study compares the detailed rendering effects of the main grid (40m), sub-grid (20m), and computational grid (20m) for water depth inundation (58 hours after flooding). The overall large image serves as a technical verification platform for the flood inundation simulation of region H during the 58-hour flooding period using the main-sub-grid and full-resolution grids. Figures 11-13 and Figures 14-16 The group comparison system confirms the innovative logic of "the main grid undertaking basic calculations and the sub-grids achieving fine reconstruction." The comparison of computation time for different grid configurations further verifies the core advantages of this invention. Without GPU computing power, with the same parameter settings and simulation time, using a uniform 3.5GHz CPU processor with 12 cores and 24 threads, the comparison results of different grid configurations, grid resolutions, and computation time are shown in Table 4 below.

[0171] Table 4. Comparison of computation time for different grid configurations, grid resolutions, and computation times.

[0172]

[0173] in, Figure 11 and Figure 14 The 40m grid results, combined with the computation time of the 40m global grid, and relying on the two-dimensional hydrodynamic model constructed based on the two-dimensional Saint-Venant equations in S2 mentioned above, were solved by alternating implicit ADI algorithm discretization and OpenMP parallelization optimization, with a computation time of only 8 minutes and 1 second. This demonstrates the natural efficiency advantage of low-resolution global computation—achieving fast solutions based on a simplified grid size, outputting the macroscopic flood inundation trend, and fully reflecting the computational efficiency advantage of the main grid under computational constraints. However, limited by the low resolution, the edges of the inundation zone are "stepped," and there are isolated points caused by interpolation noise in some areas. Micro-topographical details such as embankments are blurred, exposing the limitations of single low-resolution computation in terms of accuracy. 20m global grid (corresponding to...) Figure 13 and Figure 16While it can capture details of flooding at high resolution (such as smooth boundaries and clear micro-topography), it takes as long as 37 minutes and 55 seconds, significantly increasing computational costs. This reflects the technical contradiction of traditional high-resolution global computing: "high precision inevitably comes at the cost of efficiency." Figure 12 and Figure 15 The 20m subgrid result, taking only 11 minutes and 7 seconds, is only about 3 minutes longer than the 40m full-domain calculation, yet achieves a significant improvement in accuracy. Based on the quadtree recursive splitting strategy described in S1, a high-resolution subgrid system is generated by subdividing the 40m main grid. The bilinear interpolation algorithm described in S3 achieves accurate mapping of the main grid's water level field to the subgrid. Combined with the dynamic threshold mechanism based on the neighborhood inundation ratio described in S4 to suppress isolated points, and through multiple scans and statistical inundation frequencies using four grid traversal sequences described in S5, the final inundation zone boundary is continuous and smooth, with clearly discernible micro-topographic details such as embankments and scattered depressions. Figure 13 and Figure 16 The calculation results from the 20m full-resolution grid show a high degree of accuracy, accurately reproducing the true form of flood evolution under complex terrain. This technical comparison not only intuitively demonstrates that the 20m subgrid, while inheriting the computational efficiency of the 40m main grid, almost reaches the accuracy level of the 20m full-resolution grid through a hierarchical grid architecture, adaptive interpolation algorithm, and multi-strategy optimization mechanism, but also reflects that this invention provides basic data support through the efficient computation of the main grid and achieves a leap in detail accuracy through the fine reconstruction of the subgrid, constructing a synergistic balance system of "macroeconomic efficiency and microeconomic accuracy." In scenarios with limited computing power, it breaks through the technical contradiction of "efficiency and accuracy being mutually exclusive" in traditional models.

[0174] This invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the above-described high-precision reconstruction method for the main-subgrid flood-inundated area.

[0175] See Figure 17 The present invention also provides a computer device, including a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the high-precision reconstruction method for the main-subgrid flood inundation area described above.

[0176] This invention also provides a computer program product containing instructions that, when run on a computer, cause the computer to perform the steps of the high-precision reconstruction method for the main-subgrid flood-inundated area described above.

[0177] It is understood that the systems, devices, and storage media provided in the embodiments of the present invention correspond to the methods provided in the embodiments of the present invention, and the explanations, examples, and beneficial effects of the relevant content can be referred to the corresponding parts of the above-described high-precision reconstruction method for main-subgrid flood-inundated areas.

[0178] It should be noted that those skilled in the art will understand that all or part of the steps implemented in the embodiments of the present invention can be implemented entirely or partially by software, hardware, firmware, or any combination thereof. When implemented in hardware, it can be implemented entirely or partially by purchasing standard parts or modifications. When implemented in software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid state disks (SSDs)).

[0179] In summary, this invention addresses the technical challenge of achieving high-precision mapping from low-resolution computational data to high-resolution grids under limited computing power, while ensuring the spatial continuity and physical rationality of flooded area simulations. It proposes a high-precision reconstruction method for flooded areas using a master-subgrid approach for hydrodynamic models. Through hierarchical grid construction, adaptive interpolation algorithms, and multi-strategy optimization mechanisms, this method resolves the contradiction between computational efficiency and accuracy in traditional simulation methods, thus achieving high-precision reconstruction of flooded areas.

[0180] It should be understood that the examples and embodiments described herein are for illustrative purposes only and are not intended to limit the invention. Those skilled in the art can make various modifications or changes based on them. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the invention should be included within the protection scope of the invention.

Claims

1. A high-precision reconstruction method for flood-inundated areas using a main-subgrid architecture, characterized in that: Includes the following steps: S1. Construct the main grid system. Based on the quadtree recursive splitting strategy, divide each main grid cell in the main grid system into four sub-grid units at equal intervals to form a sub-grid system. S2. Using the main grid as the computational unit, a two-dimensional hydrodynamic model is constructed based on the two-dimensional shallow water equation. The alternating implicit ADI algorithm is used for numerical solution, and the model is optimized by calibrating the roughness parameter to obtain the calculation results of the water level field and / or flow velocity field, providing basic data for subgrid interpolation. S3. Reconstruct the water level field and / or flow velocity field based on the bilinear interpolation algorithm; S4. Set a dynamic threshold to determine the flooding status of subgrid cells, and suppress isolated flooding points caused by interpolation noise through parameter optimization; S5. Four grid traversal orders are adopted: forward scan, reverse scan, column-first forward scan, and column-first reverse scan. Multiple scans are used to generate a flooded area set. The frequency of each subgrid cell being marked as "flooded" is counted. Cells with a frequency ≥ 50% are retained to avoid the flooded area from growing in a certain direction due to different scanning methods. S4 further includes: S41. Taking each sub-grid cell as the center, extract the 8-neighborhood range, that is, the 8 adjacent cells in the up, down, left, right and four diagonal directions, and construct the neighborhood analysis window. S42. Count the number of units marked as flooded in the 8-neighborhood and calculate the proportion of the number of flooded units to the total number of units in the 8-neighborhood. S43. For each subgrid cell, compare its calculated water level with the terrain elevation plus a dynamic threshold. If the condition is met, mark it as "submerged": When the flooding rate is ≥30%, it is determined that there is a trend of large-scale flooding, and a low threshold is adopted. At this point, the water level needs to be higher than the terrain elevation + Only then is it marked as submerged; When the flooding rate is less than 30%, it is considered to have an isolated point risk, and a high threshold is applied. , At this point, the water level needs to be higher than the terrain elevation + Only then is it marked as submerged; S44, Calibration based on historical flood measurement data and The specific value of the threshold difference is chosen to ensure that it can suppress interpolation noise while preserving the true flooding boundary.

2. The high-precision reconstruction method for flood-inundated areas using a main-subgrid system as described in claim 1, characterized in that, S1 further includes: S11. Define the hierarchical structure of the main grid and subgrids. Based on the quadtree recursive splitting strategy, divide each main grid cell into four subgrid units at equal intervals along the horizontal and vertical axes to form a grid system with parent-child mapping relationship. S12. Generate the main grid fishing net vector file, set the starting coordinates and row and column spacing, and generate a dataset containing grid center point features; S13. Convert the main grid fishing net into point features, extract the elevation values ​​of each grid point based on the corresponding precision digital elevation model, and store them in the attribute table; S14. Perform quadtree partitioning on each main grid cell, generate four subgrids by calculating the midpoint of the vertex coordinates, assign IDs to the subgrids in row priority order, and make them inherit the ID prefix of the main grid to establish spatial association. S15. Convert the subgrid fishing net into point features, extract the elevation values ​​of each subgrid point based on a higher-precision digital elevation model, and map the main grid ID attribute through spatial association tools to ensure the consistency of parent and child grid topology. S16. Link the elevation attributes of the main grid and subgrid point features to the corresponding fishing net vector file to generate digital elevation models of different precision, and verify the grid topology integrity and the consistency between the elevation data and the measured terrain.

3. The high-precision reconstruction method for flood-inundated areas using a main-subgrid system as described in claim 1, characterized in that, S2 further includes: S21. Perform registration, cropping and resampling operations on land use type data to make the data coordinate system consistent with the cell size and digital terrain model, and create model partition roughness files according to land use type. S22. Based on the master grid system and model partition roughness file, a two-dimensional hydrodynamic model is developed using the alternating implicit ADI algorithm and OpenMP parallel computing technology. S23. Based on the measured flow rate and water level boundary conditions, optimize the roughness value of the model, reenact the flood inundation situation in a typical year, and complete the parameter calibration of the hydrodynamic model by comparing the measured inundation range data.

4. The high-precision reconstruction method for flood-inundated areas using a main-subgrid structure as described in claim 3, characterized in that, The development process of the two-dimensional hydrodynamic model in S22 is as follows: S221. The two-dimensional Saint-Venant equations are adopted as the fundamental equations for hydrodynamic calculations. In Cartesian coordinates, the vector form of these equations is expressed as follows: Of these five styles, The horizontal coordinate is the coordinate of the space. The vertical coordinate is the coordinate of the space. For time, For the source term vector, A variable vector, including water level and unit width flow , Both F and G are flux vectors, including water depth h. and Directions and The flow rate on the surface, , is the Xie Cai coefficient. This is the Manning coefficient; S222. The mass and momentum equations are discretized using the finite difference method with alternating implicit and explicit methods, and the equation matrices generated in each direction are solved using the chasing method.

5. The high-precision reconstruction method for flood-inundated areas using a main-subgrid system as described in claim 1, characterized in that, S3 further includes: S31. Normalize the interpolation point coordinates to obtain the relative position parameters of the interpolation point coordinates in the horizontal and vertical directions within the main grid cell. The formulaic expression is as follows: In these two equations, (x0, y0) are the coordinates of the center point of the lower left main grid cell, (x1, y0) are the coordinates of the center point of the lower right main grid cell, (x0, y1) are the coordinates of the center point of the upper left main grid cell, (x1, y1) are the coordinates of the center point of the upper right main grid cell, and (x, y) are the coordinates of the point to be interpolated. and The value range of is [0, 1], which represents the relative position of the interpolation point in the horizontal and vertical directions, respectively; S32. Perform two linear interpolations in the horizontal direction. The specific process is as follows: At the base of y=y0, with respect to Z 00 and Z 10 Perform linear interpolation to obtain the intermediate value Z. 底 : At the top edge of y=y1, for Z 01 and Z 11 Perform linear interpolation to obtain the intermediate value Z. 顶 : In these two equations, Z 00 Given the known water level value at the center point (x0, y0) of the lower left main grid cell, Z 10 Z represents the known water level value at the center point (x1, y0) of the lower right main grid cell. 01 Given the known water level value at the center point (x0, y1) of the top-left main grid cell, Z 11 The known water level value is the center point (x1, y1) of the main grid cell in the upper right corner; S33. In the vertical direction, for Z 底 and Z 顶 Perform linear interpolation to obtain the final interpolation result Z(x, y): ； S34. Combine the expressions in S32 and S33 to obtain a unified expression for calculating the water level of the target interpolation point: 。 6. The high-precision reconstruction method for flood-inundated areas using a main-subgrid system as described in claim 1, characterized in that, S5 further includes: S51. Determine the four grid traversal orders. The specific process is as follows: Forward scanning, row priority: Access cells in ascending order of row and column from top left to bottom right. The traversal path is (i,j) → (i,j+1) → (i+1,0), where i is the row number and j is the column number. Reverse scan, row-first: Access cells in descending order of row and column from bottom right to top left. The traversal path is (i,j) → (i,j-1) → (i-1,M-1), where... The total number of columns; Column-first forward scan: Access cells in ascending order of column and row from the top left to the bottom right. The traversal path is (i,j) → (i+1,j) → (0,j+1). Column-first reverse scan: Access cells in descending column and row order from bottom right to top left. The traversal path is (i,j) → (i-1,j) → (N-1,j-1), where... Total number of rows; S52. Perform multiple scans and frequency statistics. The specific process is as follows: The grid is scanned multiple times using the four sequences described above, with each scan generating a flooding marker matrix F. t (i,j), F t (i,j)=1 indicates submersion, F t (i,j)=0 indicates no flooding. The flooding frequency of each sub-grid cell is calculated using the following formula: Where F(i,j) is the frequency matrix, reflecting the frequency at which the cell (i,j) is marked as submerged in multiple scans; S53. Select and retain submerged cells. The specific process is as follows: A robust screening of flooded cells is implemented using a decision rule with a frequency threshold of ≥50%. A four-directional scanning mechanism (row, column, forward, and reverse) is used to prevent the flooded area from growing excessively in a single direction, thus ensuring the spatial consistency of the extracted flooded area results.

7. A computer device, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the high-precision reconstruction method for the main-subgrid flood-inundated area as described in any one of claims 1-6.

Images