Digital twin river basin river network high-precision extraction method

Abstract

The application relates to the technical field of hydrology, and discloses a digital twin river basin river network high-precision extraction method, which comprises the following steps: obtaining DEM data and carrying out splicing and cutting processing, judging whether to carry out depression filling processing, so as to generate DEM data after the river basin is filled with depressions; through unit grid division, the water flow direction of each unit grid is calculated, flow direction data is generated, and the cumulative flow data of each unit grid is calculated; a catchment threshold is set, whether the cumulative flow data of each unit grid exceeds the catchment threshold is judged, a first water body unit is generated, river linking, river network grading and vectorization processing are carried out, vectorized river network data is generated, an improved river network density index is calculated, an optimal catchment threshold is calculated, a digital river network corresponding to the optimal catchment threshold is generated, and the river network density is calculated, so that a joint evaluation index based on the superposition difference and the river network density is established, and the precision of the digital river network is evaluated; the method improves the digital river network extraction precision and the accuracy of the extraction result evaluation.

Description

Technical Field

[0001] This invention relates to the field of hydrology technology, specifically to a method for high-precision extraction of digital twin river networks in a watershed. Background Technology

[0002] River networks play a crucial role in hydrology and water resources management, providing key data for multiple interdisciplinary fields such as natural disaster monitoring and early warning (e.g., flood, drought, and landslide monitoring), and urban planning and management. Currently, the mainstream methods for extracting river networks include GIS and DEM-based methods and remote sensing imagery-based methods. Remote sensing imagery-based methods are easily affected by factors such as mountain shadows and cloud cover, leading to inaccurate river information and difficulty in extracting small rivers. Therefore, most current river network extraction research relies heavily on GIS and DEM data. However, the accuracy of river networks extracted based on GIS and DEM data is affected by various factors, among which the water catchment threshold is a key parameter affecting accuracy. The water catchment threshold determines the accumulated water flow, thus affecting river identification and classification. An excessively large water catchment threshold can easily create false channels, while an excessively small threshold may lead to the loss of small rivers. Therefore, determining the optimal water catchment threshold is crucial for river network extraction.

[0003] Among these methods, the river network density method is currently the mainstream approach for determining the optimal catchment threshold. Although it is widely used in various watershed experiments, there is still a certain gap between the river network extracted by the river network density method and the actual river network. This gap is caused by the inaccurate determination of the optimal catchment threshold. In addition, the important reference indicator for evaluating the accuracy of river network extraction is the overlay error. Although many researchers use the size of the overlay error to judge the quality of the extracted river network after extracting the digital river network, the overlay error, due to its calculation principle, is the area of ​​the fragmented region formed by the extracted river network and the actual river network lines that occupies the entire watershed. The proportion of area cannot accurately evaluate the accuracy of river network extraction. The reason is that when the fitting difference is used as the sole indicator of river network accuracy, as the river length decreases, the river network density decreases, the number of intersections between the extracted river network and the real river network decreases, the area of ​​the resulting fragmented polygons decreases, and the fitting difference decreases, which is considered to improve accuracy. Thus, it can be concluded that the lower the river network density, the higher the river network extraction accuracy. However, if the river network density is too low, the river network will be sparse, and many small rivers will not be extracted, resulting in low river network accuracy. This contradicts the conclusion that the lower the river network density, the higher the river network extraction accuracy. Summary of the Invention

[0004] To address the aforementioned shortcomings in existing technologies, this invention provides a high-precision extraction method for digital twin watershed river networks. This method solves the problems of insufficient accuracy of water collection thresholds and inaccurate extraction of river networks when using existing river network density methods to extract water collection thresholds. Furthermore, it considers river network density evaluation indicators based on existing overlay difference indicators to address the problem of inaccurate evaluation of river network extraction accuracy.

[0005] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows:

[0006] A method for high-precision extraction of river networks in a digital twin watershed includes the following steps:

[0007] S1. Obtain DEM data of the study area and perform splicing and cropping to generate DEM data of the watershed. At the same time, determine whether there are depressions in the DEM data of the watershed. If so, fill the depressions to obtain the DEM data of the watershed after filling. Otherwise, do not process and directly use the DEM data of the watershed as the basic topographic data for river network extraction.

[0008] S2. Divide the DEM data after filling depressions in the watershed or the basic topographic data extracted from the river network into unit grids, calculate the water flow direction of each unit grid, generate flow direction data, and calculate the cumulative flow data of each unit grid based on the flow direction data.

[0009] S3. By setting a water collection threshold, determine whether the cumulative flow data of each unit grid exceeds the water collection threshold. If so, identify each unit grid that exceeds the water collection threshold as the first water body unit; otherwise, do not identify it as the first water body unit.

[0010] S4. Perform river linking, river network classification, and vectorization on all first-level water body units to generate vectorized river network data;

[0011] S5. Based on vectorized river network data, obtain the length of the final-level rivers and the total length of the rivers, and calculate the improved river network density index.

[0012] S6. Establish an improved fitting function between the river network density index and the water collection threshold and perform second-order differentiation. Take the point where the fitting function approaches zero as the optimal water collection threshold point, and take the water collection threshold corresponding to the optimal water collection threshold point as the optimal water collection threshold. Then, determine whether the cumulative flow data of each cell exceeds the optimal water collection threshold. If so, identify each cell that exceeds the optimal water collection threshold as the second water body cell. Otherwise, do not identify it as the second water body cell.

[0013] S7. Link all second water body units through rivers, classify river networks, and vectorize them to generate a digital river network corresponding to the optimal water catchment threshold and obtain the river network density of the digital river network. Establish a joint evaluation index based on the overlap difference and river network density to evaluate the accuracy of the digital river network.

[0014] The present invention has the following beneficial effects:

[0015] The present invention proposes a high-precision extraction method for digital twin watershed river networks, which uses the density of the final-level rivers as the main reference object. Compared with the traditional river network density method, which uses the density of the entire river network as the reference object, it can determine a more accurate optimal water catchment threshold and thus extract more accurate river network results. Moreover, compared with the traditional river network extraction result accuracy evaluation index, the joint evaluation index proposed in this invention considers the influence of overlay difference and river network density on the river network extraction accuracy evaluation, making the evaluation of the river network extraction effect more accurate. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a high-precision extraction method for digital twin river networks proposed in this invention.

[0017] Figure 2 This is a schematic diagram of the depression-filling process in the embodiment;

[0018] Figure 3 This is a schematic diagram of the water flow direction process using the D8 algorithm in the embodiment;

[0019] Figure 4 This is a schematic diagram illustrating the cumulative flow calculation in the embodiment;

[0020] Figure 5 This is a schematic diagram of the river connection in the embodiment;

[0021] Figure 6 This is a schematic diagram of the river network classification using the Strahler classification method and the Shreve classification method in the embodiment;

[0022] Figure 7 This is a schematic diagram of the river network density curve fitting at a resolution of 30m in the embodiment;

[0023] Figure 8 This is a schematic diagram of the river network density curve fitting at a resolution of 90m in the example;

[0024] Figure 9 This is a schematic diagram of the river network density curve fitting at a resolution of 450m in the example;

[0025] Figure 10 This is a schematic diagram of the river network extraction effect at a resolution of 30m in the example;

[0026] Figure 11This is a schematic diagram of the river network extraction effect at a resolution of 90m in the example;

[0027] Figure 12 This is a schematic diagram of the river network extraction effect at a resolution of 450m in the example;

[0028] Figure 13 This is a schematic diagram showing a partial comparison of the river network extraction results at a resolution of 30m in the example. Detailed Implementation

[0029] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0030] like Figure 1 As shown, a method for high-precision extraction of river networks in a digital twin watershed includes the following steps S1-S7:

[0031] S1. Obtain DEM data of the study area and perform splicing and cropping to generate DEM data of the watershed. At the same time, determine whether there are depressions in the DEM data of the watershed. If so, perform depression filling to obtain DEM data of the watershed after depression filling. Otherwise, do not process and directly use the DEM data of the watershed as the basic topographic data for river network extraction.

[0032] In this embodiment, the ArcGIS hydrological analysis module is used to perform operations such as depression filling, flow direction calculation, runoff accumulation calculation, and river network extraction on DEM data. This allows for the acquisition of hydrological information such as river length, river network structure, and flow accumulation, enabling digital river network extraction based on DEM data. However, due to uncertainties in the accuracy of DEM data, "false depressions" may appear. The presence of these "false depressions" can cause interruptions or unreasonable diversions in subsequent river network extraction, forming "pseudo-channels" and thus affecting the extraction of the digital river network for the basin. Therefore, before further processing, the DEM data needs to be corrected, i.e., depression filling. The principle is to modify the elevation of depression points to the elevation of runoff points, specifically as follows: Figure 2 As shown, the conditions for the existence of depression points and sloping points are that within a unit confluence region, the elevation of the central cell is lower than the elevations of all surrounding boundary cells. In this case, a depression point and a sloping point are considered to exist within the unit confluence region, with the central cell being the depression point and the cell with the lowest elevation among the surrounding cells being the sloping point. Furthermore, a unit confluence region as shown... Figure 3As shown, a 3x3 grid is formed by a cell and its surrounding cells. The elevation difference between the depression point and the pouring point is set to z. Therefore, the filling result of the confluence area of ​​the cell can be determined by setting the value of z, that is, the elevation difference threshold. If a cell confluence area has a depression point and the elevation difference between the depression point and the pouring point is less than the set elevation difference threshold, then the elevation of the depression point is changed to the elevation of the pouring point to achieve the filling process. If it is greater than the elevation difference threshold, then no processing is performed.

[0033] Specifically, the specific process of filling depressions in step S1 is as follows:

[0034] The DEM data of the watershed is divided into cell grids to generate several cell confluence regions; each cell confluence region is a region composed of a cell grid and its adjacent cell grids.

[0035] Determine whether there are depressions and sluices in each unit's confluence area, and whether the elevation difference between the depression and the sluice is less than a set elevation difference threshold. If so, adjust the elevation of the depression to the elevation of the sluice to obtain the DEM data of the watershed after filling depressions; otherwise, do not process it.

[0036] Specifically, the conditions for the existence of depression points and sluice points are as follows: in a unit confluence region, if the elevation of the center cell is less than the elevation of all surrounding boundary cells, then the unit confluence region is considered to have depression points and sluice points, and the center cell of the unit confluence region is the depression point, and the cell with the lowest elevation among all surrounding boundary cells is the sluice point.

[0037] S2. Divide the DEM data after filling depressions in the watershed or the basic topographic data extracted from the river network into unit grids, calculate the water flow direction of each unit grid, generate flow direction data, and calculate the cumulative flow data of each unit grid based on the flow direction data.

[0038] In this embodiment, the D8 algorithm is used to calculate the river flow direction in ArcGIS. This algorithm considers that water flow from each cell can flow in one of eight directions around it. The cell adjacent to the center cell with the steepest slope is selected as the next flow direction. Figure 3 As shown, the specific operation process is as follows: step S21. Furthermore, by inputting the flow direction file (flow direction data) into the flow hydrology analysis module of ArcGIS, the cumulative runoff can be calculated. The basic principle is to use a raster as the unit, with each point representing a unit of water volume. Following the natural law of water flow from high to low, the cumulative water volume flowing through each raster is calculated based on the flow direction data, thus obtaining the cumulative runoff of each unit's catchment area. When the runoff accumulates to a certain level, surface water flow occurs. All raster cells with runoff exceeding the critical value are potential water body cells. These water body cells connect to form a river network, specifically as follows... Figure 4As shown, if the watershed area is a 16×16 grid area, after the flow direction of each grid is obtained through the D8 algorithm, assuming that each grid has one unit of water volume, then the cumulative flow of each grid is the cumulative water volume of the entire watershed flowing through this grid. The operation process is as follows: steps S22-S23.

[0039] Specifically, step S2 includes S21-S23:

[0040] S21. Divide the DEM data after filling depressions in the watershed or the basic topographic data extracted from the river network into cell grids. Use the D8 algorithm to select the adjacent cell grid with the largest slope to the central cell grid as the next flow direction of the water flow. Calculate the flow direction of the water flow in each cell grid, i.e.:

[0041]

[0042] Among them, D ij The expression represents the direction of water flow in cell (i,j), argmin represents the independent variable used to find the minimum value, N(i,j) represents the eight neighboring cells of cell (i,j), and Z... ij Z represents the elevation of the cell (i,j). kl This represents the elevation of the cell grid (k,l).

[0043] S22. Generate flow direction data based on the water flow direction of each cell grid.

[0044] S23. Based on the flow direction data, calculate the cumulative flow data for each cell grid, i.e.:

[0045]

[0046] Among them, A ij Let U(i,j) represent the cumulative flow data of cell (i,j), and let U(i,j) represent the set of all cells flowing to cell (i,j). kl This represents the cumulative water flow from cell (k,l) to cell (i,j).

[0047] S3. By setting a water collection threshold, determine whether the cumulative flow data of each unit grid exceeds the water collection threshold. If so, identify each unit grid that exceeds the water collection threshold as the first water body unit; otherwise, do not identify it as the first water body unit.

[0048] In this embodiment, by setting a specified water collection threshold, the data results of the water body unit are obtained. If the cumulative flow exceeds the set water collection threshold, the area is identified as having a river.

[0049] S4. Perform river linking, river network classification, and vectorization on all first-level water body units to generate vectorized river network data.

[0050] In this embodiment, the individual rivers identified in the above steps (where rivers are obtained by connecting water body units) are linked together to form a river network, specifically as follows: Figure 5 As shown, Figure 5 The red dots represent the confluences of each river, and the black lines represent river connections. Simultaneously, the obtained river network is classified, assigning different levels to each river. Common classification methods include the Strahler classification and the Shreve classification. The Strahler classification defines the source river as Level 1, and the river network level only increases when rivers of the same level converge. Therefore, the intersection of a Level 1 connector and a Level 2 connector retains the Level 2 connection. The Strahler and Shreve classifications both establish a Class 1 river as the source, but the river's level is the sum of the levels of all preceding rivers. For example, two Class 1 rivers intersecting create a Class 2 river, a Class 1 river intersecting a Class 2 river creates a Class 3 river, and a Class 2 river intersecting a Class 3 river creates a Class 5 river. Figure 6 As shown; furthermore, this invention uses the Strahler classification method as the river network classification method for subsequent operations. Finally, the obtained river network classification raster data is converted into a vector format river network through vectorization processing, thereby completing the conversion from raster to vector river network data; then, the river lengths of the vector data can be calculated to obtain the lengths of the first-level rivers (i.e., the last-level rivers) and the sum of the lengths of all rivers (the total river length).

[0051] Specifically, step S4 includes S41-S43:

[0052] S41. Link all first-level water body units together to form a river network.

[0053] S42. The Strahler classification method is used to classify the river network. By assigning different levels to each river in the river network, the classified river network raster data is obtained, specifically as follows:

[0054] The source rivers in the river network are defined as Level 1. The river level in the river network is upgraded only when rivers of the same level meet. That is, when a Level 1 river meets another Level 1 river, a Level 2 river is generated. When a Level 1 river meets another Level 2 river, the Level 2 river attribute is retained. When a Level 2 river meets another Level 2 river, a Level 3 river is generated, and so on, until the main stream of the basin is generated, thus completing the classification of the river network and obtaining the classified river network raster data.

[0055] S43. The hierarchical river network raster data is vectorized to generate vectorized river network data.

[0056] S5. Based on vectorized river network data, obtain the length of the final-level rivers and the total length of the rivers, and calculate the improved river network density index.

[0057] In this embodiment, vector data (vectorized river network data) is used to calculate the length of the first-order rivers (i.e., the final-order rivers), the sum of the lengths of all rivers, and an improved river network density index. The improved river network density index is based on the principle of the river network density method. Specifically, as the catchment threshold increases, the total river length decreases. This change in total river length is mainly caused by the decrease in the length of the final-order rivers, while the lengths of the main channels and other major river channels do not change significantly. Therefore, the total river length L is divided into the length of the final-order rivers and the lengths of other rivers. A coefficient weight is assigned to the latter to adjust the relationship between the two, reducing the proportion of other river lengths. This makes the curve change more representative of the changing trend of the final-order river length, thereby finding a more accurate inflection point as the optimal catchment threshold. The formula for calculating the improved river network density index used in the improved river network density method is: I = (L... a +B*(LL a )) / S.

[0058] Specifically, step S5 includes S51-S52:

[0059] S51. Based on vectorized river network data, obtain the length of the final-level rivers and the total length of the rivers; where the length of the final-level rivers is the length of the source rivers of the river network.

[0060] In this embodiment, the total river length also refers to the length of all rivers at all levels, or the sum of the lengths of all rivers. In the above steps, a vector file of the river network can be obtained. The attribute of this vector file is a line file. The total river length can be obtained by calculating the length of all lines in this line file. The calculation principle is as follows: the size of the cell is determined by the resolution of the initial DEM data. A 30m resolution means a 30m×30m cell. After determining the position of each cell identified as a water body, the center points of the water body cells are connected to form the river network. The sum of the lengths of all connecting lines is the total river length. This can generally be calculated using geographic information software—ArcGIS software. At the same time, the length of each individual river can also be calculated, or the length of each level of river can be calculated according to the river level.

[0061] S52. Based on the length of the terminal rivers and the total length of the rivers, calculate the improved river network density index, namely:

[0062] I = (L a +B*(LL a )) / S

[0063] Where I represents the improved river network density index, and L a L represents the length of the final-level river, B represents the total length of the river, and S represents the first empirical coefficient.

[0064] Among them, the value range of the first empirical coefficient is 0 to 1, and it has been verified through experiments that the improvement effect is best when B is 0.

[0065] S6. Establish an improved fitting function between the river network density index and the water collection threshold and perform second-order differentiation. Take the point where the fitting function approaches zero as the optimal water collection threshold point, and take the water collection threshold corresponding to the optimal water collection threshold point as the optimal water collection threshold. Then, determine whether the cumulative flow data of each cell exceeds the optimal water collection threshold. If so, identify each cell that exceeds the optimal water collection threshold as the second water body cell. Otherwise, do not identify it as the second water body cell.

[0066] In this embodiment, the calculated improved river network density index and the water catchment threshold are fitted by a function. Generally, the two satisfy a power function relationship. Then, the second derivative of the fitted function is solved. The water catchment threshold corresponding to the point where the second derivative approaches 0 is the optimal water catchment threshold.

[0067] Specifically, step S6 includes S61-S62:

[0068] S61. Establish an improved fitting function between the river network density index and the water catchment threshold, and take the second derivative of the fitting function to obtain the point where the second derivative approaches zero. Take this point as the optimal water catchment threshold point, and take the water catchment threshold corresponding to the optimal water catchment threshold point as the optimal water catchment threshold.

[0069] In this embodiment, the formula for calculating the second derivative of the fitted function is as follows: Where f′(x) is the first derivative, f″(x) is the second derivative, f(x) is the function value at x, f(x+h) is the function value at (x+h), and h is a value close to 0.

[0070] S62. Based on the optimal water collection threshold, determine again whether the cumulative flow data of each cell exceeds the optimal water collection threshold. If so, identify each cell that exceeds the optimal water collection threshold as a second water body cell; otherwise, do not identify it as a second water body cell.

[0071] S7. Link all second water body units through rivers, classify river networks, and vectorize them to generate a digital river network corresponding to the optimal water catchment threshold and obtain the river network density of the digital river network. Establish a joint evaluation index based on the overlap difference and river network density to evaluate the accuracy of the digital river network.

[0072] In this embodiment, a joint evaluation index based on the overlap difference and river network density is established to evaluate the accuracy of the river network density extracted by the optimal water catchment threshold and the density of the actual river network. Specifically, the difference in river network density is used as an auxiliary verification index for accuracy evaluation. The smaller the difference, the better the extraction effect. In order to present this index digitally, the difference in river network density is given a certain weight to participate in accuracy verification. The specific calculation formula is: i=c*|Dd| / d+T.

[0073] Specifically, step S7 includes S71-S73:

[0074] S71. Link all second water body units through rivers, classify river networks, and vectorize them to generate a digital river network corresponding to the optimal water catchment threshold, and obtain the river network density of the digital river network.

[0075] S72. Based on the river network density of the digital river network, establish a joint evaluation index based on the overlap difference and the river network density, namely:

[0076] i = c * |Dd| / d + T

[0077] Where i represents the joint evaluation index based on the overlap difference and river network density, c represents the second empirical coefficient, D represents the river network density of the digital river network extracted using the optimal water catchment threshold, d represents the actual river network density, and T represents the overlap difference; where, since the overlap difference ranges from 0 to 0.05 and the river network density difference ranges from 0 to 0.5, which is an order of magnitude different, c can be set to 0.1.

[0078] In this embodiment, the formula for calculating the river network overlap difference is: Where T is the difference in river network overlap, and A i The area of ​​the fragmented region is km² 2 S represents the total drainage area in km². 2 Furthermore, in this embodiment, the total watershed area refers to the area of ​​the watershed from which the river network is to be extracted. That is, in the initial step, the DEM data needs to be stitched and cropped. The cropping process requires determining the boundary of the watershed under study, such as the Jinsha River watershed. Once the boundary is determined, the area of ​​the watershed calculated is the total watershed area.

[0079] S73. The accuracy of digital river networks is evaluated using a joint evaluation index based on fitting difference and river network density.

[0080] In this embodiment, to verify the effectiveness of the proposed high-precision extraction method for digital twin river networks in a watershed, the Jinsha River basin was selected as the research area for the experiment, as detailed below:

[0081] The DEM products are: (1) ASTER GDEMV3 (30m resolution); (2) SRTM DEM (90m resolution); (3) GEBCODEM (450m resolution); the verification data is: real river network and the map is from the National Basic Geographic Information Center of the Ministry of Natural Resources; the research plan is set as follows: in order to discuss the advantages of the improved river network density method, DEM data of different resolutions are selected for river network extraction, specifically: first, the optimal threshold is calculated and the digital river network is extracted using the three DEM data and the traditional river network density method; second, the optimal threshold and digital river network of the three DEM data are extracted using the improved river network density method; finally, the accuracy of the river network extracted by the two methods is evaluated and verified based on the joint evaluation index using the river network data published by the Ministry of Natural Resources. Based on the above data foundation and experimental procedure, the experimental results are as follows:

[0082] (1) River network extraction results

[0083] Three types of DEM data from three watersheds were processed and analyzed. Appropriate catchment thresholds were set for each watershed to extract the river network and calculate its density. A curve showing the relationship between river network density and catchment threshold was also plotted. The results are as follows: Figures 7-9 As shown, it can be observed that there is a power function relationship between the water catchment threshold and the river network density, and the fitting correlation coefficient R0 2 All values ​​were greater than 0.9, indicating a good fit. Furthermore, the second derivatives of the fitted curves obtained from different DEM resolutions were calculated to determine the inflection points, yielding optimal catchment thresholds of 2.4 km for each of the three DEM data sets. 2 (30m resolution), 2.5km 2 (90m resolution), 2.6km 2 (450m resolution); Furthermore, the DEM data underwent the same processing. When calculating the river network vector file, the improved river network density index was calculated using the improved river network density method, and the optimal catchment threshold was calculated by fitting the relationship curve with the catchment threshold. The river network extraction effect is as follows: Figures 10-12 As shown, the result obtained is 1.8km. 2 (30m resolution), 1.9km 2 (90m resolution), 2.1km 2 (450m resolution) is completely different from the optimal catchment threshold obtained by the river network density method, and is generally too small.

[0084] (2) Accuracy verification of river network extraction results

[0085] The accuracy of the extracted river network results was verified, and the results are shown in Table 1:

[0086] Table 1. Accuracy Evaluation of River Network Extraction Results

[0087]

[0088] Table 1 shows that the improvement in river network evaluation indicators in the Jinsha River Basin was 0.56% (an increase of 38,000 km in river length) for GDEMV3 data, 0.74% for SRTM data (an increase of 45,000 km in river length), and 0.79% for GEBCODEM data (an increase of 46,000 km in river length). Different DEM data showed varying degrees of accuracy improvement. Figure 13 The comparison of river network extraction results using 30m resolution data is shown in detail. Since the results for 90m and 450m resolution data are similar, only the 30m resolution data results are presented here. Figure 13 The river network of the Jinsha River Basin, from top to bottom, is represented by the traditional river network density method, the improved river network density method, and the actual river network distribution. A and B show the extraction results of two terminal rivers in the Jinsha River Basin, while A1, A2, B1, and B2 are magnified displays of key areas. Figure 4 It can be observed that the main channels extracted by both methods basically match the actual river network, with the main difference lying in the extraction of tributary-level channels. Specifically, in region A, although both methods can identify the tributary, the channel length extracted by the improved method proposed in this invention is closer to the actual situation. In region B, the traditional method failed to identify the tributary, while the improved method proposed in this invention successfully extracted it. Enlarged views of A1, A2, B1, and B2 confirm that real, small channels do exist in these areas, rather than false "pseudo-channels." The results show that the improved river network density method has stronger adaptability and can effectively handle data with different terrain features and DEM resolutions, with a significantly improved extraction accuracy compared to the traditional method.

[0089] In summary, the high-precision river network extraction method proposed in this invention firstly, compared to the traditional river network density method which uses the density of the entire river network as a reference, uses the density of the final-level rivers as the main reference, enabling the calculation of a more accurate optimal catchment threshold and obtaining more precise river network extraction results. Secondly, compared to traditional accuracy evaluation indicators, the joint evaluation indicator proposed in this invention considers both the overlap difference and the river network density on the accuracy of river network extraction, enabling a more accurate evaluation of the extraction effect. Therefore, the method proposed in this invention improves the accuracy of the river network extraction results.

[0090] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

[0091] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A method for high-precision extraction of river networks in a digital twin watershed, characterized in that, Includes the following steps: S1. Obtain DEM data of the study area and perform splicing and cropping to generate DEM data of the watershed. At the same time, determine whether there are depressions in the DEM data of the watershed. If so, fill the depressions to obtain the DEM data of the watershed after filling. Otherwise, do not process and directly use the DEM data of the watershed as the basic topographic data for river network extraction. S2. Divide the DEM data after filling depressions in the watershed or the basic topographic data extracted from the river network into unit grids, calculate the water flow direction of each unit grid, generate flow direction data, and calculate the cumulative flow data of each unit grid based on the flow direction data. S3. By setting a water collection threshold, determine whether the cumulative flow data of each unit grid exceeds the water collection threshold. If so, identify each unit grid that exceeds the water collection threshold as the first water body unit; otherwise, do not identify it as the first water body unit. S4. Perform river linking, river network classification, and vectorization on all first-level water body units to generate vectorized river network data; S5. Based on vectorized river network data, obtain the length of the final-level rivers and the total length of the rivers, and calculate the improved river network density index. S6. Establish an improved fitting function between the river network density index and the water collection threshold and perform second-order differentiation. Take the point where the fitting function approaches zero as the optimal water collection threshold point, and take the water collection threshold corresponding to the optimal water collection threshold point as the optimal water collection threshold. Then, determine whether the cumulative flow data of each cell exceeds the optimal water collection threshold. If so, identify each cell that exceeds the optimal water collection threshold as the second water body cell. Otherwise, do not identify it as the second water body cell. S7. Link all second water body units through rivers, classify river networks, and vectorize them to generate a digital river network corresponding to the optimal water catchment threshold and obtain the river network density of the digital river network. Establish a joint evaluation index based on the overlap difference and river network density to evaluate the accuracy of the digital river network.

2. The method for high-precision extraction of river networks from digital twin watersheds according to claim 1, characterized in that, The specific process of filling depressions in step S1 is as follows: The DEM data of the watershed is divided into cell grids to generate several cell confluence regions; each cell confluence region is a region composed of one cell grid and its adjacent cell grids. Determine whether there are depressions and sluices in each unit's confluence area, and whether the elevation difference between the depression and the sluice is less than a set elevation difference threshold. If so, adjust the elevation of the depression to the elevation of the sluice to obtain the DEM data of the watershed after filling depressions; otherwise, do not process it.

3. The method for high-precision extraction of river networks from digital twin watersheds according to claim 2, characterized in that, The conditions for the existence of depression points and sluice points are as follows: in a unit confluence region, if the elevation of the center cell is less than the elevation of all surrounding boundary cells, then the unit confluence region is considered to have depression points and sluice points. The center cell of the unit confluence region is the depression point, and the cell with the lowest elevation among all surrounding boundary cells is the sluice point.

4. The method for high-precision extraction of river networks from digital twin watersheds according to claim 3, characterized in that, Step S2 specifically includes: S21. Divide the DEM data after filling depressions in the watershed or the basic topographic data extracted from the river network into cell grids. Use the D8 algorithm to select the adjacent cell grid with the largest slope to the central cell grid as the next flow direction of the water flow. Calculate the flow direction of the water flow in each cell grid, i.e.: Among them, D ij The expression represents the direction of water flow in cell (i,j), argmin represents the independent variable used to find the minimum value, N(i,j) represents the eight neighboring cells of cell (i,j), and Z... ij Z represents the elevation of the cell (i,j). kl Represents the elevation of the cell grid (k,l); S22. Generate flow direction data based on the water flow direction of each cell grid; S23. Based on the flow direction data, calculate the cumulative flow data for each cell grid, i.e.: Among them, A ij Let U(i,j) represent the cumulative flow data of cell (i,j), and let U(i,j) represent the set of all cells flowing to cell (i,j). kl This represents the cumulative water flow from cell (k,l) to cell (i,j).

5. The method for high-precision extraction of river networks from digital twin watersheds according to claim 4, characterized in that, Step S4 specifically includes: S41. Link all first-level water body units with rivers to generate a river network; S42. The Strahler classification method is used to classify the river network. By assigning different levels to each river in the river network, the classified river network raster data is obtained, specifically as follows: The source rivers in the river network are defined as Level 1. The river level in the river network is upgraded only when rivers of the same level meet. That is, when a Level 1 river meets another Level 1 river, a Level 2 river is generated. When a Level 1 river meets another Level 2 river, the Level 2 river attribute is retained. When a Level 2 river meets another Level 2 river, a Level 3 river is generated. This process continues until the main stream of the watershed is generated, thus completing the classification of the river network and obtaining the classified river network raster data. S43. The hierarchical river network raster data is vectorized to generate vectorized river network data.

6. The method for high-precision extraction of river networks from digital twin watersheds according to claim 5, characterized in that, Step S5 specifically includes: S51. Based on vectorized river network data, obtain the lengths of the final-level rivers and the total river length; where the length of the final-level rivers is the length of the source rivers of the river network; S52. Calculate the improved river network density index based on the length of the terminal rivers and the total length of the rivers.

7. The method for high-precision extraction of river networks in a digital twin watershed according to claim 6, characterized in that, The formula for calculating the improved river network density index in step S52 is as follows: I=(L a +B*(L-L a )) / S Where I represents the improved river network density index, and L a L represents the length of the final-level river, B represents the total length of the river, and S represents the first empirical coefficient.

8. The method for high-precision extraction of river networks in a digital twin watershed according to claim 7, characterized in that, Step S6 specifically includes: S61. Establish an improved fitting function between the river network density index and the water catchment threshold, and take the second derivative of the fitting function to obtain the point where the second derivative approaches zero. Take this point as the optimal water catchment threshold point, and take the water catchment threshold corresponding to the optimal water catchment threshold point as the optimal water catchment threshold. S62. Based on the optimal water collection threshold, determine again whether the cumulative flow data of each cell exceeds the optimal water collection threshold. If so, identify each cell that exceeds the optimal water collection threshold as a second water body cell; otherwise, do not identify it as a second water body cell.

9. The method for high-precision extraction of river networks from digital twin watersheds according to claim 8, characterized in that, Step S7 specifically includes: S71. Link all second water body units through rivers, classify river networks, and vectorize them to generate a digital river network corresponding to the optimal water catchment threshold, and obtain the river network density of the digital river network. S72. Based on the river network density of the digital river network, establish a joint evaluation index based on the overlap difference and the river network density; S73. The accuracy of digital river networks is evaluated using a joint evaluation index based on fitting difference and river network density.

10. The method for high-precision extraction of river networks in a digital twin watershed according to claim 9, characterized in that, The joint evaluation index based on the overlap difference and river network density in step S72 is: i = c * |Dd| / d + T Where i represents the joint evaluation index based on the overlap difference and river network density, c represents the second empirical coefficient, D represents the river network density of the digital river network extracted using the optimal water catchment threshold, d represents the actual river network density, and T represents the overlap difference.

Images