Large-scale watershed hydrological modeling method based on variable grids

Abstract

The invention provides a large-scale watershed hydrological modeling method based on variable grids, and belongs to the field of large-scale distributed hydrological modeling. The large-scale watershed hydrological modeling method comprises the following steps: firstly, determining the size of a watershed basic grid according to the spatial resolution of available basic data and the calculation time consumption of a distributed hydrological model under different resolutions; secondly, determining a grid level according to a rainfall level and a mountain torrent prevention and control level of a flood event; and finally, according to the grid levels of the different partitions, the basic grids are aggregated into coarse grids of different levels, and basin variable grid hydrological modeling is completed. The method is suitable for a large-scale watershed flood forecasting system, and can meet the timeliness demand of flood control decision while guaranteeing the refined modeling of a key area.

Description

Technical Field

[0001] The present invention belongs to the field of large-scale distributed hydrological modeling, and relates to a large-scale basin distributed hydrological modeling method based on variable grids. Background Art

[0002] The basin hydrological model is the core of the flood forecasting system. With the increasing requirements for flood forecasting accuracy and spatio-temporal refinement, lumped hydrological models are gradually developing towards distributed hydrological models that can finely describe the spatio-temporal characteristics of rainfall and underlying surface. However, large-scale basins cover a wide range and the underlying surface conditions are complex and diverse. For traditional fixed-grid modeling to establish a distributed flood forecasting scheme based on 1-km fine grids for the entire basin, the number of grid divisions reaches hundreds of thousands or even millions, and the time-consuming for calculating a flood is too long to meet the timeliness requirements of flood forecasting. Therefore, distributed hydrological refined modeling mostly stays at the theoretical method level or in small and medium-sized basins with good data conditions.

[0003] A rainstorm flood often occurs in a local area of a large basin. The runoff generation and concentration process in the rainstorm center area is crucial for flood formation. Using fine grids can accurately describe the runoff generation and concentration process in the rainstorm center area, while the non-primary runoff generation areas that occupy a large proportion of the basin have less impact on flood formation. Using coarse grids will not reduce the overall flood forecasting accuracy of the large basin, but can significantly improve the calculation efficiency and meet the requirements of flood forecasting for calculation time-consuming. Existing studies have explored distributed hydrological refined modeling methods. For example, Chinese invention patent CN115130396A proposes a distributed hydrological refined modeling method for the reservoir area of a river-type reservoir. By dividing sub-basins and slope basins and constructing refined hydrological response units, the spatial heterogeneity of the underlying surface of slope units is fully considered. However, this method uses high-resolution grids throughout the entire basin, resulting in a significant increase in the overall calculation cost and reducing the operating efficiency of the model. Chinese invention patent CN105160121A proposes a distributed hydrological modeling method controlled by finite elements. This method divides finite element units based on the distribution of rain gauges and the characteristics of the underlying surface, and can improve the modeling efficiency and simulation accuracy to a certain extent. However, the refinement degree of this method in the rainstorm center area is insufficient, and it is difficult to accurately depict the local hydrological response process caused by heavy rainfall, thus limiting the application performance of the model in extreme event forecasting. The above methods have deficiencies in terms of accuracy in the rainstorm center area and efficiency in non-primary runoff generation areas.

[0004] Therefore, in order to improve the modeling refinement degree of key flood control areas in large basins while taking into account the overall calculation time-consuming of the model, a large-scale basin distributed hydrological modeling method based on variable grids is proposed. For key flood control areas such as rainstorm centers and densely populated areas, fine-grid modeling is used, and for the remaining non-key areas, coarse-grid modeling is used. Summary of the Invention

[0005] Aiming at the problems existing in the prior art, the present invention provides a large-scale basin distributed hydrological modeling method based on variable grids, which realizes fine-grid modeling for key areas within a large basin and coarse-grid modeling for non-key areas, ensuring the model accuracy while meeting the requirements of computational timeliness.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A large-scale basin distributed hydrological modeling method based on variable grids. The research idea of the large-scale basin distributed hydrological modeling method is to adopt the idea of grid "aggregation", aggregating fine grids into coarse grids, mainly including three parts: determining the size of the basin basic grid, determining the grid levels of different areas in the basin, and aggregating the basin basic grid into coarse grids of different levels. First, based on the spatial resolution of the available basic data (underlying surface data and meteorological and hydrological data) and weighing the computational time-consuming of the distributed hydrological model at different resolutions, the size of the basin basic grid is determined. Secondly, on this basis, according to the rainfall magnitude of a flood event and the mountain flood prevention and control level, the grid level is determined. Finally, according to the grid levels of different partitions, the basic grid is aggregated into coarse grids of different levels to complete the variable-grid hydrological modeling of the basin. The specific steps are as follows:

[0008] Step 1, determine the size of the basin basic grid.

[0009] The basin basic grid is the smallest grid obtained by analyzing the collected basin basic data required for distributed hydrological modeling. The specific process is as follows:

[0010] Step 1.1, collect the basin basic data required for distributed hydrological modeling. The basin basic data includes meteorological and hydrological data and underlying surface data, where the meteorological and hydrological data includes rainfall, evaporation, temperature, wind speed data and their station locations and spatial distributions, and the underlying surface data includes remote sensing image data including topography, soil, vegetation, and land use.

[0011] Step 1.2, use the basin basic data collected in Step 1.1 as the input data of the distributed hydrological model, classify according to the influence degree of the above input data on distributed hydrological modeling, and respectively determine the requirements of various meteorological and hydrological data and underlying surface data for the grid size of distributed hydrological modeling. The specific requirements are as follows:

[0012] If the meteorological and hydrological data has an obvious influence on distributed hydrological modeling, then for this type of data, the size of the basin basic grid is not lower than the spatial resolution of the meteorological and hydrological data

[0013] (1)

[0014] Among them, Grid 降雨 , Grid 蒸发 , Grid 气温 , Grid 风速 are the spatial resolutions of rainfall, evaporation, air temperature, and wind speed data respectively.

[0015] Step 1.2.2: The underlying surface data is obtained through remote sensing images. The spatial resolution of some underlying surface remote sensing images is much higher than that of the meteorological and hydrological basic data (for example, the resolution of the terrain DEM image can be better than 30 m), which can provide more detailed surface cover information. According to the spatial resolution of the underlying surface data, the requirements for the grid size are as follows:

[0016] (2)

[0017] Among them, Grid 地形 , Grid 土壤 , Grid 植被 , Grid 土地利用 represent the spatial resolutions of terrain, soil, vegetation, and land use respectively.

[0018] However, when the underlying surface data is evenly distributed or consistent as a whole in space, even if a very fine grid is used for division, the impact on the hydrological modeling results is very small, but it will significantly increase the model calculation time. Therefore, although the underlying surface information can be obtained through remote sensing images with a spatial resolution better than 30 m, when constructing the basic grid of the basin, there is no need to use such ultra-high resolution data. It is necessary to balance the spatial distribution characteristics of the underlying surface and the calculation time of the hydrological model, and reasonably control Grid 下垫面 to control the calculation time while ensuring the simulation accuracy.

[0019] Step 1.3: Based on the requirements of the above various input data for the spatial resolution of distributed hydrological modeling, determine the reasonable size Grid of the basic basin grid 基础网格 , and the specific formula is:

[0020] (3)

[0021] In summary, determine the size of the basic basin grid.

[0022] Step 2: Determine the grid levels in different regions of the basin.

[0023] First, select a typical flood process in the basin. According to the spatial distribution of rainfall in this flood, count the maximum rainfall P max ,24h in consecutive 24 hours in different regions of the basin. Then, collect the spatial distribution and flood control levels of mountain flood prevention areas in the basin, perform a spatial overlay of the rainfall spatial distribution and the mountain flood prevention areas, and according to P max ,24hIdentify the key areas of concern within the basin for flood prevention and control of mountain torrents. The grid level division rules for different regions are as follows:

[0024] (1) For high-risk areas of mountain torrents:

[0025] (4)

[0026] Among them, Grid represents the spatial resolution of the i-th sub-region of the basin.

[0027] P max ,24h The area with a rainstorm center where P 基础网格 exceeds 100 mm belongs to the key flood control area and requires basic grid modeling Grid max ,24h ; for non-rainstorm center areas where P

[0028] (2) For medium- and low-risk areas of mountain torrents:

[0029] (5)

[0030] When P max ,24h is less than or equal to 50 mm, model using a grid size 10 times that of Grid 基础网格 ; when P max ,24h exceeds 50 mm, model using a grid size 5 times that of Grid 基础网格 .

[0031] In summary, the grid level division results for different sub-regions of the basin are obtained.

[0032] Step 3: Aggregate the basic grids of the basin into coarser grids of different levels.

[0033] According to the grid level division results of different sub-regions of the basin obtained in Step 2, perform variable grid aggregation to aggregate the basic grids of the basin into coarser grids of different levels. For high-risk areas of mountain torrents located at the rainstorm center, retain the basic grid size of the basin for modeling to finely describe the runoff generation and concentration processes within the area. For medium- and low-risk areas of mountain torrents or non-rainstorm center areas, aggregate the basic grids of the basin into coarser grids of different multiples. Variable grid aggregation includes two parts: grid runoff generation area calculation and grid confluence flow direction extraction. The specific content is as follows:

[0034] Step 3.1: Grid runoff generation area calculation. Variable grid aggregation causes the boundaries of the basic grids of the basin and the coarser grids to overlap, resulting in incorrect calculation of the runoff generation area of the coarser grids. Therefore, by establishing a mapping relationship between the coarser grids and the basic grids of the basin, count the number of basic grids of the basin contained in each coarser grid, and accumulate the runoff generation area A 基础网格,i of the basic grids contained in the coarser grid to obtain the runoff generation area A 粗网格 of the coarser grid.

[0035] (6)

[0036] Among them, n represents the number of basic grids included in the coarse grid.

[0037] Step 3.2, extraction of grid confluence flow direction. The variable grid aggregation changes the cell sizes of different grids, resulting in the inapplicability of the traditional fixed calculation method for the flow direction of a single type of grid confluence. Therefore, it is necessary to conduct multiple rounds of iterative search and correction among various grid sizes with incorrect flow directions to reconstruct the topological relationship of the variable grid river network. The specific approach is as follows: First, identify the grid cells with interrupted or abnormal flow directions, then search for target cells with lower terrain and reasonable connection relationships in their adjacent areas, and adjust their flow directions to point to the correct downstream cells. By continuously correcting the abnormal points through multiple iterations, finally, a complete and connected river network topological structure is reconstructed to ensure the smooth confluence of water flow throughout the region, especially maintaining the continuity and accuracy of the confluence path at the grid connection points.

[0038] Thus, the variable grid division work of the large basin is completed, and a variable grid distributed hydrological modeling scheme for the large basin is constructed.

[0039] The beneficial effects of the present invention are as follows:

[0040] The present invention provides a large-scale basin distributed hydrological modeling method based on variable grids, which can, while ensuring the modeling accuracy of key areas including the rainstorm center area, effectively reduce the model calculation time by adopting coarse grids in non-key areas, and achieve the unity of the accuracy and efficiency of the distributed hydrological model. This method has a flexible grid level division mechanism, can adapt to different basin characteristics and flood forecasting requirements, meet the timeliness requirements of flood forecasting, and provides an efficient and reliable modeling scheme for the large basin flood forecasting system. Description of the Drawings

[0041] Figure 1 is the flow chart of the present invention.

[0042] Figure 2 is the result diagram of the grid aggregation of the Jialing River variable grid distributed hydrological modeling.

[0043] Figure 3 is the result diagram of the river network flow direction calculation of the Jialing River variable grid distributed hydrological modeling.

[0044] Figure 4 is the distribution diagram of the stations in the Jialing River basin (the specific stations are A, B, C, D, E, F, G, H, I, J).

[0045] Figure 5 is the flood simulation result diagram of the hydrological model at Station A.

[0046] Figure 6 It is a flood simulation result diagram of the hydrological model for Site B.

[0047] Figure 7 It is a flood simulation result diagram of the hydrological model for Site C.

[0048] Figure 8 It is a flood simulation result diagram of the hydrological model for Site D.

[0049] Figure 9 It is a flood simulation result diagram of the hydrological model for Site E.

[0050] Figure 10 It is a flood simulation result diagram of the hydrological model for Site F.

[0051] Figure 11 It is a flood simulation result diagram of the hydrological model for Site G.

[0052] Figure 12 It is a flood simulation result diagram of the hydrological model for Site H.

[0053] Figure 13 It is a flood simulation result diagram of the hydrological model for Site I.

[0054] Figure 14 It is a flood simulation result diagram of the hydrological model for Site J. Detailed implementation manner

[0055] Based on the field of large-scale basin distributed hydrological modeling, the present invention proposes a large-scale basin distributed hydrological modeling method based on variable grids.

[0056] The following further illustrates the present invention through examples.

[0057] Select the typical large basin Jialing River as the research area. The drainage area of the Jialing River is 157,000 km 2 , located in the heavy rain area of the Daba Mountains, where heavy rains occur frequently within the basin. There are a large number of mountain flood prevention areas in the upper reaches of the Jialing River. In the face of local heavy rains, mountain flood disasters are extremely likely to occur in a short time. Therefore, in the face of the numerous mountain flood prevention areas with a wide distribution in the Jialing River, there are relatively high requirements for the accuracy and timeliness of hydrological modeling.

[0058] Step 1, determine the size of the basic grid of the basin.

[0059] The basin basic grid is the smallest grid obtained by analyzing the basin basic data required for distributed hydrological model modeling. The specific process is as follows: Step 1.1, collect the basin basic data required for distributed hydrological modeling of the Jialing River Basin. The basic data includes meteorological and hydrological data and underlying surface data of the Jialing River Basin. Among them, the meteorological and hydrological data includes rainfall, evaporation, temperature, wind speed data and their station locations and spatial distributions, and the underlying surface data includes remote sensing image data including topography, soil, vegetation, and land use.

[0060] Step 1.2, use the basin basic data collected in Step 1.1 as the input data of the distributed hydrological model, classify according to the influence degree of the above input data on distributed hydrological modeling, and respectively determine the requirements for the grid size of various meteorological and hydrological data and underlying surface data for distributed hydrological model modeling. The specific requirements are as follows:

[0061] Step 1.2.1, if the meteorological and hydrological data has an obvious impact on distributed hydrological model modeling, then for this type of data, the basin basic grid size is not lower than the spatial resolution of the meteorological and hydrological data. The density of rainfall stations in the Jialing River Basin is 442 km 2 / station, and the evaporation, temperature and wind speed data use monthly potential evapotranspiration data in China with a spatial resolution of 1 km. Therefore, determine Grid 水文气象 to be 1 km.

[0062] (1)

[0063] Among them, Grid 降雨 、Grid 蒸发 、Grid 气温 、Grid 风速 are the spatial resolutions of rainfall, evaporation, temperature and wind speed data respectively.

[0064] Step 1.2.2, the underlying surface data is obtained through remote sensing images. The spatial resolution of underlying surface data such as soil and vegetation is 1 km, and the spatial resolution of terrain data is 30 m. The spatial resolution of its terrain data is much higher than that of meteorological and hydrological basic data, and it can provide more detailed surface coverage information. According to the spatial resolution of the underlying surface data, the requirements for the grid size are:

[0065] = 30 m (2)

[0066] Among them, Grid 地形 、Grid 土壤 、Grid 植被 、Grid 土地利用 represent the spatial resolutions of terrain, soil, vegetation and land use respectively.

[0067] However, since the elevation change of DEM data with a precision of 30m has little impact on the results of hydrological modeling but will significantly increase the calculation time, under the trade-off between the spatial distribution characteristics of the underlying surface and the calculation time of the hydrological model, the Grid is reasonably controlled 下垫面 to be 1km to improve the calculation efficiency.

[0068] Step 1.3, based on the requirements of the above various types of input data for the spatial resolution of distributed hydrological modeling, determine the reasonable size of the basic basin grid Grid 基础网格 to be 1km, and the specific formula is:

[0069] = 1km (3)

[0070] In summary, determine the size of the basic basin grid Grid 基础网格 to be 1km, and the number of 1km basic grids in the Jialing River Basin is 152,243.

[0071] Step 2, determine the grid levels in different regions of the basin.

[0072] Select the typical flood process from August 9th to 29th, 2020 in the Jialing River Basin. First, divide the basin into 10 sub-regions according to the spatial distribution of the mountain flood prevention areas in the Jialing River Basin. Then, count the maximum rainfall P in different regions of the basin for 24 consecutive hours max ,24h . Finally, according to the grid level division rules, determine the grid sizes in different regions. The grid level division rules for different regions are as follows:

[0073] (1) For high-risk mountain flood areas:

[0074] (4)

[0075] Among them, Gridi represents the spatial resolution of the i-th sub-region of the basin.

[0076] P max ,24h The area of the rainstorm center where the rainfall exceeds 100mm belongs to the key flood prevention area and needs to use the basic grid for modeling Grid 基础网格 ; for non-rainstorm center areas where P max ,24h is less than 100mm, use coarse grids with different multiples of 3, 5, and 10 for modeling.

[0077] (2) For medium and low-risk mountain flood areas:

[0078] (5)

[0079] When P max ,24h is less than or equal to 50mm, use a grid with 10 times the size of Grid 基础网格 for modeling; when P max ,24h exceeds 50mm, use a grid with 5 times the size of Grid基础网格 Size modeling.

[0080] In summary, the grid level division results of different sub - regions of the basin are obtained, as shown in Table 1.

[0081] Table 1 Grid level division results of different sub - regions of the Jialing River Basin

[0082]

[0083] Step 3: Aggregate the basic grid of the basin into coarser grids of different levels.

[0084] According to the grid level division results of different sub - regions of the basin obtained in the second step, variable grid aggregation is carried out. For the high - risk mountain flood areas located at the rainstorm center, the size modeling of the basic grid of the basin is retained to finely describe the runoff generation and concentration processes within the area. For the medium - low - risk mountain flood areas or non - rainstorm center areas, the basic grid of the basin is aggregated into coarser grids of different multiples. Variable grid aggregation includes two parts: grid runoff generation area calculation and grid confluence flow direction extraction. The specific contents are as follows:

[0085] Step 3.1: Grid runoff generation area calculation. According to the grid level division results of different sub - regions, the basic grids of different sub - regions are aggregated into coarser grids of different levels and the grid runoff generation areas are calculated, as shown in the appendix. Figure 2 The total number of grids for variable grid hydrological modeling is 8141, among which the numbers of 1km, 3km, 5km, and 10km grids are 3329, 189, 3935, and 688 respectively.

[0086] Step 3.2: Grid confluence flow direction extraction. Calculate the river network flow directions of the variable grids and need to conduct multiple rounds of iterative search and correction for incorrect flow directions among various grid sizes to reconstruct the variable grid river network topological relationship. The flow direction extraction results are as shown in the appendix. Figure 3 as shown.

[0087] So far, the variable grid division work of the Jialing River Basin has been completed, and a variable grid distributed hydrological modeling scheme for the Jialing River Basin has been constructed.

[0088] The hardware device used for the operation of the variable grid distributed hydrological model is a laptop computer equipped with an 11th Gen Intel(R)Core(TM)i7 - 11800H processor with a base main frequency of 2.30 GHz and an 8 - core 16 - thread CPU. Compared with the 5min 12s required for the 1km grid model to simulate a flood runoff generation and concentration calculation, when using the variable grid distributed hydrological modeling method, the runoff generation and concentration calculation time for all sub - regions of the Jialing River is only 29s. The flood simulation accuracy of the variable grid distributed hydrological modeling is as Figures 5 to 14 shown, which significantly reduces the model calculation time while ensuring the refinement degree of model modeling.

[0089] The embodiments described above merely represent the implementation manners of the present invention, but should not be construed as limiting the scope of the patent for the present invention. It should be noted that for those skilled in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all fall within the protection scope of the present invention.

Claims

1. A large-scale watershed hydrological modeling method based on a variable grid, characterized in that, The large-scale watershed hydrological modeling method is as follows: First, determine the size of the basic grid of the watershed based on the spatial resolution of the available basic data and weighing the computing time-consuming of the distributed hydrological model at different resolutions; Second, determine the grid level according to the rainfall magnitude of a flood event and the mountain flood prevention level; Finally, according to the grid levels of different partitions, aggregate the basic grids into coarse grids of different levels to complete the variable-grid hydrological modeling of the watershed, including the following steps: Step 1, determine the size of the basic grid of the watershed; The basic grid of the watershed is the smallest grid obtained by analyzing the basic data of the watershed required for distributed hydrological modeling collected; Step 2, determine the grid levels of different regions in the watershed; Step 3, aggregate the basic grids of the watershed into coarse grids of different levels; According to the results of the grid level division of different partitions in the watershed obtained in Step 2, perform variable-grid aggregation to aggregate the basic grids of the watershed into coarse grids of different levels; For the high-risk mountain flood area located in the rainstorm center, retain the size of the basic grid of the watershed for modeling and finely describe the runoff generation and concentration processes in the area; For the medium and low-risk mountain flood areas or non-rainstorm center areas, aggregate the basic grids of the watershed into coarse grids of different multiples; The variable-grid aggregation includes two parts: calculating the runoff generation area of the grid and extracting the flow direction of grid confluence.

2. The large-scale watershed hydrological modeling method based on a variable grid according to claim 1, wherein, The specific content of Step 1 is as follows: Step 1.1, collect the basic data of the watershed required for distributed hydrological modeling; The basic data of the watershed includes meteorological and hydrological data and underlying surface data, Step 1.2, take the basic data of the watershed collected in Step 1.1 as the input data of the distributed hydrological model, classify according to the influence degree of the above input data on the distributed hydrological modeling, and respectively determine the requirements of various meteorological and hydrological data and underlying surface data for the grid size of the distributed hydrological modeling; Step 1.

3. Based on the requirements of the above various types of input data for the spatial resolution of distributed hydrological modeling, determine a reasonable size Grid of the basic basin grid, and the specific formula is as follows: 基础网格 , specifically, the formula is: , In summary, determine the size of the basic basin grid.

3. A large-scale watershed hydrological modeling method based on a variable grid according to claim 2, characterized in that In Step 1.1, the meteorological and hydrological data includes rainfall, evaporation, air temperature, wind speed data and their station locations and spatial distributions, and the underlying surface data includes remote sensing image data including topography, soil, vegetation, and land use.

4. A large-scale watershed hydrological modeling method based on a variable grid according to claim 2, characterized in that, The specific content of Step 1.2 is as follows: Step 1.2.1, for meteorological and hydrological data, the size of the basic grid of the watershed is not less than the spatial resolution of the meteorological and hydrological data: , where Grid 降雨 , Grid 蒸发 , Grid 气温 , Grid 风速 are the spatial resolutions of rainfall, evaporation, air temperature, and wind speed data, respectively; Step 1.2.2, the underlying surface data is obtained through remote sensing images. According to the spatial resolution of the underlying surface data, the requirements for the grid size are: , where Grid 地形 , Grid 土壤 , Grid 植被 , Grid 土地利用 respectively represent the spatial resolutions of terrain, soil, vegetation, and land use.

5. A large-scale watershed hydrological modeling method based on a variable grid according to claim 1, characterized in that The specific content of Step 2 is as follows: First, select a typical flood process within a basin. According to the spatial distribution of rainfall during this flood, statistically calculate the maximum rainfall P for 24 consecutive hours in different regions of the basin. max ,24h Then, collect the spatial distribution and flood prevention levels of mountain flood prevention areas within the basin. Conduct a spatial overlay of the rainfall spatial distribution and the mountain flood prevention areas. Based on P max ,24h and the flood prevention levels, determine the key areas of concern within the basin; and determine the grid level division rules for different regions.

6. The large-scale watershed hydrological modeling method based on variable grids according to claim 5, characterized in that In Step 2, the rules for dividing the grid levels of different regions are as follows: (1) For the high-risk mountain flood area: , where represents the spatial resolution of the i-th sub-region of the basin; P max ,24h Areas with heavy rain centers exceeding 100 mm belong to key flood control areas and require the use of basic grid modeling Grid 基础网格 ; For P max ,24h Non-heavy rain center areas less than 100 mm are modeled using coarse grids with different multiples of 3, 5, and 10; (2) For the medium and low-risk mountain flood areas: , when P max ,24h is less than or equal to 50 mm, model with a Grid size of 10 times 基础网格 ; when P max ,24h exceeds 50 mm, model with a Grid size of 5 times 基础网格 ; In summary, the results of the grid level division of different partitions in the watershed are obtained.

7. A large-scale watershed hydrological modeling method based on a variable grid according to claim 5, characterized in that The specific content of Step 3 is as follows: Step 3.1, calculate the runoff generation area of the grid; By establishing the mapping relationship between the coarse grid and the basin basic grid, counting the number of basin basic grids contained in each coarse grid, and accumulating the runoff generation area A of the basic grids contained in the coarse grid 基础网格,i The runoff generation area A of the coarse grid is obtained 粗网格 ; , where n represents the number of basic grids included in the coarse grid; Step 3.2, extract the flow direction of grid confluence; Perform multiple rounds of iterative search and correction among various grid sizes with incorrect flow directions, and reconstruct the topological relationship of the variable-grid river network.

8. A method for large-scale watershed hydrological modeling based on a variable grid according to claim 7, characterized in that Specifically, step 3.2 is as follows: First, identify the grid cells with interrupted or abnormal flow directions. Then, search for target cells with lower elevations and reasonable connection relationships in their adjacent areas, and adjust their flow directions to point to the correct downstream cells. Continuously correct the abnormal points through multiple iterations, and finally reconstruct a complete and connected river network topology to ensure that water can smoothly converge throughout the area.

Images