A rainfall-runoff-concentration-evolution whole-process dynamic simulation system and method

Abstract

The present application relates to the technical field of hydrological simulation, and more particularly to a rainfall-runoff-concentration-evolution whole-process dynamic simulation system and method. The method comprises the following steps: obtaining basin topographic information and extracting basin boundaries; dividing grid cells according to the basin boundaries and setting river nodes using the grid cells; collecting rainfall data using a rain measurement radar; calculating runoff using the rainfall data; inputting the runoff into a pre-trained distributed concentration model for concentration analysis and recording the concentration flow; inputting the concentration flow and the river nodes into a river flood evolution model to perform river evolution simulation and generate river concentration data; determining the amount of water entering the main river according to the river concentration data; generating a flow time series according to the amount of water entering the main river; and predicting the flood peak flow based on the flow time series. The present application realizes accurate simulation of the hydrological process of a basin based on hydrological simulation technology, and improves the accuracy and timeliness of flood prediction and early warning.

Description

Technical Field

[0001] This invention relates to the field of hydrological simulation technology, and in particular to a dynamic simulation system and method for the entire process of rainfall-runoff generation-confluence-evolution. Background Technology

[0002] Flood risk is prevalent in mountainous areas and their downstream basins, but existing watershed hydrological simulation methods still have many limitations in addressing the large variations in rainfall intensity, complex slope runoff distribution, and multi-node hydrodynamic evolution in rivers. Most studies rely on average rainfall duration or single-section water level observation data for simulation, lacking the ability to jointly analyze rainfall processes, slope runoff characteristics, and changes in water level and flow at river nodes, resulting in insufficient accuracy and timeliness in flood peak prediction. Relying solely on fixed-interval hydrological monitoring data or empirical formulas to calculate runoff generation and confluence ignores the instantaneous runoff surge under heavy rainfall conditions, the differences in inflow and confluence among different tributaries, and the longitudinal and cross-sectional characteristics of the river channel, leading to significant deviations between simulation results and actual river level and flow evolution. Existing studies based on distributed hydrological models or SWMMs mostly focus on rainfall-runoff or hydraulic calculations of a single river segment, using static parameters or fixed river node arrangements. This makes it difficult to accurately reflect the entire process of runoff generation, confluence, and river water level evolution in the entire basin under heavy rainfall conditions, and lacks adaptive modeling capabilities and dynamic simulation functions for complex basin environments. Summary of the Invention

[0003] Therefore, it is necessary for the present invention to provide a dynamic simulation system and method for the entire process of rainfall-runoff generation-convergence-evolution, in order to solve at least one of the above-mentioned technical problems.

[0004] To achieve the above objectives, a dynamic simulation method for the entire process of rainfall-runoff generation-confluence-evolution is proposed, comprising the following steps: Step S1: Obtain watershed topographic information and extract watershed boundaries; divide the watershed into grid cells based on the watershed boundaries and use the grid cells to set river nodes; Step S2: Collect rainfall data using a rain-measuring radar; calculate runoff using the rainfall data; input the runoff into a pre-trained distributed runoff model for runoff analysis and record the runoff flow. Step S3: Input the confluence flow and river nodes into the river flood evolution model, perform river evolution simulation, and generate river confluence data; Step S4: Determine the inflow volume into the main river channel based on the river confluence data; generate a flow time series based on the inflow volume into the main river channel; predict the peak flood flow based on the flow time series; Step S5: Simulate the evolution of river water level based on the peak flow, calculate the peak water level height, predict the peak arrival time, and generate a flood warning signal.

[0005] Preferably, this specification also provides a dynamic simulation system for the entire process of rainfall-runoff generation-confluence-evolution, used to execute the dynamic simulation method for the entire process of rainfall-runoff generation-confluence-evolution as described above. This dynamic simulation system for the entire process of rainfall-runoff generation-confluence-evolution includes: The watershed unit division module is used to acquire watershed topographic information, extract watershed boundaries, divide watershed boundaries into grid units, and set river nodes using grid units; The runoff analysis module is used to collect rainfall data using rain-measuring radar; calculate runoff using the rainfall data; input the runoff into a pre-trained distributed runoff model for runoff analysis; and record the runoff flow. The river evolution simulation module is used to input the confluence flow and river nodes into the river flood evolution model, perform river evolution simulation, and generate river confluence data; The flood peak prediction module is used to determine the inflow volume into the main river channel based on river confluence data; generate a flow time series based on the inflow volume; and predict the flood peak flow based on the flow time series. The early warning generation module is used to simulate the evolution of river water level based on the peak flow, calculate the peak water level height, predict the arrival time of the peak, and generate flood early warning signals.

[0006] The beneficial effects of this invention are as follows: (1) By comprehensively collecting and processing watershed topographic information, river nodes and rainfall data, combined with grid unit division and river node arrangement, a refined characterization of complex watershed hydrological features was achieved, ensuring the accuracy and completeness of watershed evolution simulation input data.

[0007] (2) In the process of generating and accumulating runoff, a distributed generating and accumulating runoff hydrological model is used to input the slope rainfall, soil moisture status and vegetation interception parameters, and dynamically calculate the upper surface runoff, infiltration water volume and slope runoff volume. This realizes the coupled simulation of the entire process of rainfall, infiltration and accumulating runoff, and improves the response capability to transient changes in hydrology under rainstorm conditions.

[0008] (3) In the river hydraulic simulation stage, the node water level and flow rate are calculated step by step based on the geometric characteristics of the river cross section, hydraulic radius, flow velocity, riverbed roughness and longitudinal slope. Combined with the lateral inflow superposition, a continuous river hydrodynamic evolution process is formed, realizing the unified modeling of river water level, flow rate and watershed confluence dynamics.

[0009] (4) In the calculation of flood peak prediction and water level evolution, the flow time series is generated by using the node water level and flow time series data to predict the flood peak flow and arrival time, realizing the dynamic quantitative characterization of the flood process of the entire basin and providing a scientific basis for flood early warning. Attached Figure Description

[0010] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a flowchart illustrating the steps of a dynamic simulation method for the entire process of rainfall-runoff generation-convergence-evolution according to the present invention. Figure 2 This is a schematic diagram of the river flood evolution model in this invention; Figure 3 This is a schematic diagram of the grid cells of the watershed boundary in this invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0011] The technical method of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention.

[0012] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0013] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0014] To achieve the above objectives, please refer to Figures 1 to 3 This invention provides a dynamic simulation method for the entire process of rainfall-runoff generation-confluence-evolution, the method comprising the following steps: Step S1: Obtain watershed topographic information and extract watershed boundaries; divide the watershed into grid cells based on the watershed boundaries and use the grid cells to set river nodes; In one embodiment, the basin elevation, slope aspect and basin boundary are analyzed using DEM data to generate 10m×10m grid cells; sub-basins are divided according to the basin's water catchment relationship, and nodes are arranged every 500m along the longitudinal direction of each river channel; cross-sectional width, riverbed roughness, longitudinal slope and lateral inflow coefficient are assigned to each node.

[0015] Step S2: Collect rainfall data using a rain-measuring radar; calculate runoff using the rainfall data; input the runoff into a pre-trained distributed runoff model for runoff analysis and record the runoff flow. In one embodiment, the rainfall radar is calibrated with a spatial resolution of 1 km and a temporal resolution of 5 min; radar reflectivity factors are collected and rainfall intensity distribution is inverted to generate rainfall grid data every 5 minutes; based on the initial soil moisture content, upper soil moisture state and vegetation interception parameters, the upper surface runoff, infiltration and slope runoff are calculated using the water balance method; the slope runoff is integrated to form the river inflow, and the runoff time series of each grid point and river node is recorded to form complete runoff data.

[0016] In another embodiment, it is assumed that rainfall data is collected continuously for 12 hours, with each hour divided into 12 5-minute time steps; the inversion yields rainfall ranges of 0–18 mm at each grid point, with an average rainfall intensity of 6 mm / h; and the total slope runoff is calculated to be approximately [missing information]. The peak inflow of the river channel was Low-flow rate The flow rate time series of each node was recorded, generating a total of 288 time steps of flow data. During the iteration process, the flow rate of each time step was compared with the historical measured data. If the deviation exceeded 5%, the infiltration parameters and slope flow coefficient were adjusted and recalculated to ensure the accuracy of the input data.

[0017] Of particular importance is that in step S2, the generated flow is input into a pre-trained distributed merging model for merging analysis, and the merging flow is recorded, including: Obtain slope units, allocate the runoff to the corresponding slope units to form the inflow of each slope unit; calculate the runoff velocity based on the inflow; record the slope length of the slope unit; use the slope length and runoff velocity as inputs into a pre-trained distributed runoff model to obtain the runoff duration; calculate the slope runoff volume based on the runoff duration. In one embodiment, the target watershed is first divided into 500 slope units, each with an area of ​​approximately [area missing]. The rainfall runoff was allocated according to the area ratio of each slope unit and the soil infiltration characteristics to obtain the inflow rate of each unit. (unit ) Calculate the surface runoff velocity using Manning's formula. Where R is the hydraulic radius, S is the longitudinal slope of the slope, and n is the roughness coefficient. The slope length L of each slope unit (average approximately 200m) is recorded, and V and L are input into a pre-trained distributed flow-collecting model (based on the time-area method or a dynamic-hydraulic model) to calculate the flow-collecting duration of each slope unit. Finally, according to and Calculate slope runoff The slope runoff curve is obtained by accumulating over time, which is then used for subsequent river runoff simulation.

[0018] In another embodiment, the watershed is assumed to be divided into 300 slope units, each with an area of ​​0.03 km²; the average inflow is 0.15 m³ / s; the slope length L is between 150 and 250 m; the slope roughness n is 0.035; and the longitudinal slope S is between 0.01 and 0.03. Using a pre-trained distributed runoff model, the runoff velocity V≈0.8–1.2 m / s is calculated, corresponding to a runoff duration T_c≈150–300 s. The runoff volume Q_out of each slope unit is calculated to be between 0.12 and 0.18 m³ / s, forming a continuous slope runoff time series for downstream river runoff analysis.

[0019] Calculate the river runoff based on the output flow; integrate the slope runoff and the river runoff as the total runoff.

[0020] In one embodiment, the aforementioned slope runoff is used as the inflow condition at the river boundary to establish a one-dimensional unsteady flow model for the river channel, and node-by-node calculations are performed on the river segment. Based on the river cross-sectional geometric parameters (width, gradient, roughness coefficient) and the upstream inflow process, the river velocity distribution and runoff at each cross-section are calculated. The spatiotemporal variation of the river runoff is obtained based on the matching relationship between runoff and runoff velocity; this is then integrated with the slope runoff to form a total runoff sequence. This runoff result is used for flood peak propagation simulation and downstream catchment response analysis.

[0021] In another embodiment, it is assumed that the upstream flow rate is After flowing into the river channel, the flow rate is reduced by approximately 2 m³ / s along the 2 km main river section; the total runoff from the midstream slope is... The downstream secondary tributary supplies approximately After river velocity correction and energy loss calculations, the final flow rate reaching the outlet section is approximately... This value is similar to the measured peak flood flow (approximately...). The difference is less than 2%, indicating that the confluence calculation based on slope-channel coupling has good accuracy and reliability.

[0022] Step S3: Input the confluence flow and river nodes into the river flood evolution model, perform river evolution simulation, and generate river confluence data; In one embodiment, the hydrological data of each node are input into the river flood evolution model to calculate the hydraulic radius, flow velocity and water level between nodes, and to simulate the longitudinal river flow advancement and the superposition of lateral inflows; the water level and flow time series of each node are output for flood peak prediction and water level evolution analysis.

[0023] In another embodiment, assuming the total area of ​​the watershed... The system was divided into 3000 grid cells, with 60 nodes per river channel. The node riverbed width was 1–15m, the roughness was 0.02–0.05, and the longitudinal slope was 0.003–0.008. The node water level was calculated using hydraulic formulas and a step-by-step iterative method. The initial iteration step size was set to 0.01m. In each iteration, the roughness or weighting coefficient of the river section was adjusted according to the water level convergence error (≤0.05m). The simulation yielded water levels of 0.4–1.7m in the upstream, 0.6–2.0m in the midstream, and 0.9–2.8m in the downstream. Each node generated 360 time-step water level sequences, with a flow range of 0–42m³ / s, which fully described the hydrodynamic evolution of the watershed.

[0024] Step S4: Determine the inflow volume into the main river channel based on the river confluence data; generate a flow time series based on the inflow volume into the main river channel; predict the peak flood flow based on the flow time series; In one embodiment, the inflow of the main channel is obtained by accumulating the inflow of each sub-basin according to the time step; a time series of the main channel flow is generated; the peak flow and arrival time are calculated by using flood peak analysis methods (such as SCS-CN method, instantaneous unit hydrograph method) in combination with the water storage characteristics and boundary conditions of the channel; and regression correction is performed by combining historical flow data to improve the accuracy of flood peak prediction.

[0025] In another embodiment, it is assumed that the peak inflow of the main channel is obtained by accumulating the confluence data. Average flow The generated flow time series contains 360 time steps; the peak flow is predicted through peak flow analysis. The arrival time at the downstream node is approximately the 175th time step (11:25 AM). During the iterative correction process, if the predicted flood peak deviates from historical observations by more than 5%, the sub-basin confluence coefficient and river hydrodynamic parameters are automatically adjusted to regenerate the flow time series.

[0026] Step S5: Simulate the evolution of river water level based on the peak flow, calculate the peak water level height, predict the peak arrival time, and generate a flood warning signal.

[0027] In one embodiment, the water level is iteratively calculated node by node using the characteristics of the river cross section, hydraulic radius, riverbed roughness and longitudinal slope to form a continuous river section water level evolution curve; the water level increment of each node is calculated in combination with the flood peak flow to obtain the flood peak arrival time; flood warning signals and possible inundation areas are generated based on the water level changes for flood control scheduling and emergency response.

[0028] In another embodiment, it is assumed that the river channel has 50 cross-sectional nodes, an average channel width of 10m, a longitudinal slope of 0.005, and a roughness coefficient of 0.03; the peak flow rate is... The calculation showed that the node water level varied from 0.8 to 2.7 m; the flood peak arrived downstream in about 3 hours; during the iteration process, the water level of each node was gradually corrected, and the longitudinal slope or roughness was adjusted to ensure that the deviation between the simulated water level and the measured water level was ≤0.1 m; finally, a flood warning signal was generated, marking the potential inundation area and time period.

[0029] Preferably, step S1, which involves dividing the watershed into grid cells based on the watershed boundary and setting the river nodes using these grid cells, includes: Identify the water flow direction of each grid cell based on the watershed boundary and determine the water catchment relationship; divide the grid cells using the water catchment relationship, and arrange the node positions in the grid cells according to the preset node spacing; In one embodiment, a digital elevation model (DEM) is used to acquire watershed topographic information. The flow direction of each grid cell is calculated using the D8 algorithm or a flow direction algorithm, and a watershed catchment tree is generated based on the catchment relationships. The watershed is then divided into 50m × 50m grid cells, and nodes are arranged at the center of each grid cell at a preset node spacing (e.g., 10m) for subsequent hydrological simulation and river network construction. Each node records its corresponding grid number, geographic coordinates, and catchment area identifier to establish flow path relationships between nodes.

[0030] In another embodiment, assuming a catchment area of ​​100 km², it is divided into 4000 grid cells, each with an average side length of 50 m. The flow direction of each grid cell is calculated using a DEM, and a catchment tree is generated according to the catchment area. The number of nodes in each grid cell is 4–9, with a random node spacing of 8–12 m to ensure coverage of varying terrain elevations. Approximately 15,000 nodes are ultimately generated, and the grid number and upstream node list for each node are recorded for use in establishing the river network.

[0031] Assign river cross-sectional features based on node locations and establish node connection relationships; map the node connection relationships to grid cells to generate a node network and set river nodes.

[0032] In one embodiment, for each deployed node, river cross-sectional features, such as cross-sectional width, depth, and river slope, are obtained using digital hydrological maps or river remote sensing data, and then assigned to the corresponding node. The flow connectivity between nodes is then calculated based on the catchment direction and river slope, forming a directed node graph. This node network is then mapped to grid cells, achieving spatial correspondence between river nodes in the grid. River nodes can be used as input for runoff calculation, water level simulation, and river transport models.

[0033] In another embodiment, it is assumed that the node network contains 15,000 nodes, of which approximately 3,000 are river nodes. Each river node is assigned cross-sectional characteristics: an average cross-sectional width of 6m, a depth of 2.5m, and a slope of 0.003. Node connections are established in an upstream-downstream order, forming a directed water flow network, with each river node connecting an average of 3 upstream nodes and 2 downstream nodes. After mapping to grid cells, the river nodes are distributed along the main river paths of the grid cells, with each grid cell corresponding to a maximum of 5 river nodes, used for subsequent watershed hydrological simulation, flood analysis, and water resource management.

[0034] Please see Figure 3 The compass (N) points north, and the scale bar at the bottom shows that 1 unit length on the map represents 12.5 kilometers in reality. In the legend, blue dots represent reservoirs, and blue curves represent water systems. The water system network formed by the blue curves covers the grid area, and there is a reservoir marked with a blue dot downstream. The overall map shows the distribution pattern of rivers and the location of reservoirs in the area.

[0035] Of particular importance is the allocation of river channel cross-sectional characteristics based on node locations, including: Identify the cross-sectional structure of the river channel at the node location, including regular and complex cross-sections; detect the river channel roughness based on the cross-sectional structure and assess the flow resistance based on the river channel roughness; measure the longitudinal slope at the node location. In one embodiment, a combination of manual measurement and drone aerial photography is used to obtain river cross-section data. For regular cross-sections, cross-sectional elevations are collected every 5m, and for compound cross-sections, cross-sectional elevations are collected every 2m. The riverbed roughness is obtained through field roughness tests, with a value of 0.03–0.05. The flow resistance of each cross-section is evaluated based on empirical formulas in fluid dynamics. The longitudinal slope measurement accuracy at nodes is 0.001–0.005.

[0036] In another embodiment, it is assumed that 10 regular cross sections and 5 complex cross sections are measured; the riverbed roughness is assumed to be [0.025, 0.04, 0.03, 0.035, 0.045, 0.05, 0.03, 0.04, 0.038, 0.042], and the longitudinal slope is assumed to be [0.002, 0.003, 0.0015, 0.004, 0.0035, 0.0025, 0.0045, 0.003, 0.0028, 0.0032]. These data can provide basic input for subsequent flow transfer calculations.

[0037] The lateral inflow at each node location is superimposed node by node according to the longitudinal slope to simulate the water flow along the river channel and the water flow velocity is recorded; the friction loss energy is calculated based on the water flow resistance and the water flow velocity; the river channel roughness is corrected based on the friction loss energy as the river channel cross-sectional characteristic.

[0038] In one embodiment, the lateral inflow of the tributary is collected for each node, and the total inflow of the node is obtained by adding it to the inflow from the previous node. The flow velocity is calculated based on the node inflow, the cross-sectional area of ​​the channel, and the slope. The friction loss energy is calculated using the water flow resistance, and then the friction loss is combined with the original channel roughness for iterative correction. The channel roughness of each node is updated as channel cross-sectional feature correction data.

[0039] In another embodiment, assume the lateral inflow rate (unit) of 10 consecutive nodes. The initial flow velocity is assumed to be [0.5, 0.8, 0.6, 0.7, 1.0, 0.9, 0.4, 0.6, 0.7, 0.5], and the initial flow velocity is assumed to be [0.3, 0.35, 0.32, 0.36, 0.4, 0.38, 0.31, 0.33, 0.34, 0.3] m / s; the initial riverbed roughness is [0.03, 0.035, 0.032, 0.034, 0.04, 0.038, 0.03, 0.033, 0.036, 0.032]. After correction for friction loss, the roughness of the nodes is assumed to be [0.031, 0.034, 0.033, 0.035, 0.039, 0.037, 0.031, 0.032, 0.035, 0.033]. These corrected data can be directly used for river confluence simulation or water level prediction.

[0040] Preferably, step S2, which involves collecting rainfall data using a rain-measuring radar, includes: The reflectivity factor is obtained using a rain-measuring radar; the rainfall intensity distribution is inverted based on the reflectivity factor; the rainfall intensity distribution is then processed into a time series to generate rainfall grid points at different time steps. In one embodiment, the rain-measuring radar is calibrated with a spatial resolution of 1 km and a temporal resolution of 5 min, and reflectivity factor Z data is collected; the reflectivity factor Z is then calculated using an empirical formula. Inversion into rainfall intensity (Unit: mm / h), where a=0.017, b=0.714; the rainfall intensity of each radar grid point is smoothed and filtered to eliminate isolated outliers; the rainfall intensity data is organized into a time series according to a preset time step (e.g., 5 min, 10 min, 15 min) to generate the rainfall sequence for each grid point; finally, the rainfall grid point matrix is ​​output, which includes spatial coordinate information and the rainfall at the corresponding time step, for subsequent runoff calculation and confluence analysis.

[0041] In another embodiment, it is assumed that radar data is collected continuously for 6 hours, with each hour divided into 12 5-minute time steps, generating a total of 72 time steps of rainfall grid data. Each time step has 2000 radar coverage watershed grid points, with the retrieved rainfall intensity ranging from 0 to 18 mm / h, and an average intensity of approximately 6 mm / h. After time-series processing, a cumulative rainfall sequence is generated for each grid point, with a maximum cumulative rainfall of approximately 72 mm. During data processing, a 3×3 grid neighborhood weighted average is used for filtering to remove outliers that are more than 3 times the standard deviation above the average. After filtering, the correlation coefficient between grid points reaches 0.92, ensuring the spatial continuity and temporal consistency of the rainfall grid data.

[0042] The geographic coordinate system of the rain-measuring radar is read and added to the rainfall grid to generate a spatial distribution map of rainfall; the cumulative rainfall is calculated from the spatial distribution map to obtain the rainfall data.

[0043] In one embodiment, the projected coordinate system information of the rainfall radar is obtained, including the latitude and longitude origin, projection type, and coordinate system parameters; each rainfall grid point is mapped to the corresponding geographic coordinate location to form a two-dimensional spatial matrix; the cumulative rainfall of each grid point is calculated using GIS methods, that is, the rainfall at each time step is accumulated, and a spatial distribution map is generated; in the spatial distribution map, the rainfall intensity and cumulative rainfall of each grid point are recorded to provide basic input for watershed runoff calculation.

[0044] In another embodiment, assuming a watershed area of ​​30 km², radar grid points are mapped to geographic coordinates to form a 30×30 grid with a total of 900 grid points. Rainfall is accumulated at each grid point over 72 consecutive time steps to obtain a cumulative rainfall matrix, with a maximum cumulative rainfall of 70 mm and an average cumulative rainfall of 45 mm. To ensure spatial continuity, bilinear interpolation is performed on the cumulative rainfall matrix, reducing the grid spacing from the original 1000 m to 500 m to form a high-resolution spatial distribution map of rainfall. Finally, the rainfall at each grid point can be used for subsequent watershed gridded simulation, runoff calculation, and confluence analysis.

[0045] Preferably, the calculation of runoff based on rainfall data in step S2 includes: Based on rainfall data, soil water evolution is simulated to update the upper soil moisture status; vegetation interception is assessed based on the upper soil moisture status, and upper surface runoff is calculated. In one embodiment, the watershed is divided into several soil units (e.g., each unit is 100m × 100m), the initial upper soil moisture content of each unit is obtained, and rainfall grid data is mapped to each soil unit. The upper soil moisture status of each soil unit is updated at set time steps (e.g., every 5 minutes), and vegetation interception and surface runoff are calculated based on soil saturation and vegetation cover. The updated upper soil moisture and surface runoff are used as input for lower soil infiltration.

[0046] In another embodiment, assuming a watershed area of ​​30 km², divided into 300 soil units, with an upper soil layer thickness of 0.3 m in each unit and an initial moisture content of [missing information]. The total rainfall over 6 hours was 72 mm, and soil moisture status was updated every 5 minutes. The simulation results showed that the average saturation of the upper soil layer increased from 0.25 to 0.42, the average vegetation interception was about 3 mm / h, and the surface runoff per unit ranged from 0 to 5 mm / h. During the peak rainfall, the runoff per unit reached 5 mm / h in some units, providing time-series water volume information for the infiltration of the lower soil layer.

[0047] The residual water is estimated by using the surface runoff in the upper layer, and the residual water is infiltrated into the lower soil layer to renew the soil moisture; the yield is calculated based on the soil moisture in the lower layer.

[0048] In one embodiment, for each soil unit, the residual water after evaporation and interception from the upper surface runoff is used as the lower soil infiltration rate to update the lower soil moisture status. Then, the runoff is calculated based on the lower soil saturation and topographic slope, resulting in a lower soil runoff sequence at each time step. By summing the runoff from all soil units, the runoff time series for the entire watershed can be obtained, providing fundamental data for runoff simulation and river flow analysis.

[0049] In another embodiment, assuming the total remaining moisture in the upper layer is 50 mm over 6 hours, the lower soil layer is 0.7 m thick, and the initial moisture content is... The infiltration rate varied from 0 to 4 mm / h across different time steps, and the average saturation of the lower soil layer increased from 0.20 to 0.35. The maximum yield per unit was approximately 8 mm / h, and the average was approximately 3.5 mm / h. After comprehensive calculation, the maximum total yield of the watershed occurred at the peak rainfall time step, reaching [amount missing] mm / h. By plotting heat maps of the flow rates of different units, the differences in water distribution can be visually displayed, providing a quantitative basis for flood peak prediction and hydrological analysis.

[0050] Preferably, the calculation of upper surface runoff based on the assessment of vegetation interception according to the upper soil moisture status includes: Obtain the vegetation type of the target area; determine the vegetation interception parameters based on the vegetation type; calculate the vegetation interception amount based on the vegetation interception parameters and the upper soil moisture status; calculate the upper surface runoff based on the vegetation interception amount.

[0051] In one embodiment, the vegetation type of the target area is identified using remote sensing imagery or UAV aerial photography, and mapped to pre-divided surface units (e.g., each unit is 50m × 50m). The corresponding interception coefficient is obtained from a table based on the vegetation type, for example, grassland 0.05–0.12, shrubs 0.10–0.18, and woodland 0.15–0.25. Combined with the current moisture content of the upper soil layer, the vegetation interception amount for each unit is calculated at each time step. The interception amount is then subtracted from the rainfall to obtain the upper surface runoff. The calculated surface runoff can form a surface runoff sequence for each time step, which is used for subsequent infiltration and lower soil moisture renewal.

[0052] In another embodiment, assuming the target area is 20 km², divided into 800 surface units; the vegetation type distribution is: forest 40%, shrubs 35%, and grassland 25%. The initial upper soil moisture content of each unit is... Four hours of continuous rainfall resulted in a total rainfall of 48 mm. Isolation and surface runoff were updated every 10 minutes. Simulation results showed that vegetation isolation ranged from 0.5 to 3 mm in each unit, with forest having the highest isolation at approximately 3 mm, followed by shrubs at 2.1 mm and grassland at 1.2 mm. Calculations indicated that upper surface runoff varied from 0 to 4.5 mm across time steps, with an average surface runoff of approximately 2.6 mm per unit. Time-series plots visually illustrated the differences in isolation and runoff among different vegetation types, providing input data for watershed hydrological models.

[0053] Preferably, the calculation of vegetation interception based on vegetation interception parameters and the upper soil moisture status includes: Determine canopy height based on vegetation type; assess maximum canopy interception based on canopy height; correct maximum canopy interception using vegetation interception parameters, and record the corrected canopy interception; In one embodiment, vegetation type information for the target area is acquired using remote sensing imagery or drone aerial photography, and the corresponding canopy height is determined for each surface unit (e.g., 50m × 50m). For example, the canopy height for forest is 12m, for shrubs it is 3m, and for grassland it is 0.5m. The maximum canopy interception amount for each vegetation type is obtained by looking up a table based on canopy height and rainfall characteristics; for example, 12–18mm for forest, 4–6mm for shrubs, and 1–2mm for grassland. Then, the maximum canopy interception amount is corrected according to actual rainfall conditions and vegetation interception parameters to obtain the corrected canopy interception amount, which is recorded in the attribute table of each surface unit, providing basic data for subsequent infiltration calculations.

[0054] In another embodiment, the target area is assumed to be divided into 500 units, with woodland accounting for 30%, shrubs for 50%, and grassland for 20%. Preliminary canopy heights were measured as follows: woodland 12m, shrubs 3m, and grassland 0.5m; corresponding maximum interception amounts were 15mm, 5mm, and 1.5mm, respectively. After correction for vegetation interception parameters, the recorded corrected canopy interception ranges were: woodland 14–16mm, shrubs 4.5–5.2mm, and grassland 1.2–1.6mm. The distribution of corrected canopy interception amounts in each unit can be displayed using a bar chart, providing quantitative data for hydrological simulation.

[0055] The corrected canopy interception is converted into infiltrative water volume to obtain the effective soil water volume; the vegetation interception is calculated based on the effective soil water volume and the moisture status of the upper soil layer.

[0056] In one embodiment, the amount of water that can be infiltrated is obtained by subtracting the amount of water returned by evaporation (which can be estimated using meteorological monitoring data or the Penman-Monteith method) from the corrected canopy interception. This infiltration is then combined with the current moisture state of the upper soil layer to calculate the actual amount of water intercepted by vegetation. For example, during continuous rainfall, the amount of water infiltrated per unit area in forest land decreases, while the amount of water infiltrated per unit area in shrubs and grasslands is relatively higher. This yields the actual vegetation interception amount for each unit, which can be used for hydrological simulation or runoff analysis.

[0057] In another embodiment, assuming that during a rainfall event, the corrected canopy interception in forest land is 15 mm, evaporation return is 3 mm, and infiltration is 12 mm; the interception in shrubland is 5 mm, evaporation return is 1 mm, and infiltration is 4 mm; and the interception in grassland is 1.5 mm, evaporation return is 0.2 mm, and infiltration is 1.3 mm. These infiltration amounts are then compared with the moisture content of the upper soil layer (assuming an initial value of...). Combining these findings, the effective soil water content for each unit was obtained as follows: 0.32 m³ / m³ for forest, 0.28 m³ / m³ for shrubs, and 0.26 m³ / m³ for grassland. Time series analysis provides a clear visual representation of the changes in effective soil water content under different vegetation types, offering a quantitative basis for further runoff and runoff simulation.

[0058] Preferably, step S3, which involves inputting the confluence flow and river nodes into the river flood evolution model, includes: The confluence flow and river nodes are input into the river flood evolution model, which includes an information acquisition layer, a computational simulation layer, and a verification and adjustment layer. In one embodiment, the coordinates, catchment area, cross-sectional features, and upstream flow of each node in the river node network are used as inputs and uniformly fed into the river flood evolution model. The model's information acquisition layer reads river cross-sectional information (such as width, depth, and roughness), obtains boundary conditions (upstream flow and rainfall intensity), and basic parameters (such as initial water level and infiltration parameters) in real time, and stores them in a node attribute database to support subsequent calculations and simulations.

[0059] In another embodiment, it is assumed that the watershed comprises 3000 river nodes, each with an average catchment area of ​​approximately The cross-sectional width is 4–8 m and the depth is 2–3 m. The runoff volume under high rainfall events is... The model is input through node number and upstream flow rate. The cross-sectional roughness of each node is initially set to 0.03–0.04. Boundary conditions include historical water level and rainfall sequence of upstream nodes, which are input to the model information acquisition layer to prepare for simulation.

[0060] The information acquisition layer is used to read river cross-section information, obtain boundary conditions, and adjust basic parameters; the calculation and simulation layer is used to calculate hydraulic elements, simulate the confluence evolution of water flow between nodes along the longitudinal channel and lateral inflow, and form the dynamic spatiotemporal process of river level and flow; the verification and adjustment layer is used to verify the rationality of the calculated water level and flow, and adjust the river section roughness and weighting coefficient when the calculation results are unreasonable, and iteratively calculate to form a complete evolution process of river cross-section water level and flow.

[0061] In one embodiment, the computational simulation layer performs time-step simulation based on a node network. The hydraulic elements (such as velocity, flow rate, and water level) of each node are calculated according to the longitudinal transport formula of the river channel and the lateral inflow and confluence model. After the simulation is completed, the verification and adjustment layer compares the water level with historical water levels or water levels of neighboring nodes. If the water level deviation is found to exceed a threshold (such as ±5cm), the roughness of the river section or the weighting coefficient is automatically adjusted, and the calculation is iterated again until the simulated water level and flow rate results are reasonable, thus obtaining the complete spatiotemporal evolution process and outputting the water level and flow rate sequence of each node at each time step.

[0062] In another embodiment, assuming a simulation time step of 1 hour and a simulation duration of 72 hours, the initial water level at each node is 1.2 m. The simulation yields the following example of the water level (in meters) sequence for each node: the water level at node A1 changes over time as [1.2, 1.35, 1.5, 1.62, 1.7, 1.68, 1.55], and the flow rate (in m³ / s) is [0, 12, 25, 33, 40, 37, 28]. The verification adjustment layer found that the water level at node A1 exceeded the reasonable threshold of 0.1 m for adjacent sections in the fourth time period. Therefore, the roughness was adjusted from 0.03 to 0.028, and after iterative calculation, the water level was corrected to a reasonable range. Similar operations were performed on 3000 nodes, ultimately forming a complete dynamic spatiotemporal evolution process of river section water level and flow rate, which is used for flood prediction and water resource management.

[0063] Please see Figure 2 This demonstrates the process system for calculating and verifying river water level and flow. First, river information is read, boundary conditions are determined, and weighting coefficients and river roughness can be adjusted. Then, hydraulic element calculations, node and cross-sectional water level and flow calculations are performed sequentially to obtain water level and flow results. After that, the results are verified. If the results are unreasonable, they are fed back to the previous stage to adjust the parameters. If they are reasonable, the river cross-sectional water level and flow process is presented. Then, the next time period can be calculated to simulate and analyze the river hydraulic process.

[0064] Preferably, when the calculation results are unreasonable, the roughness of the river section and the weighting coefficient are adjusted, and the iterative calculation to form a complete process of the evolution of water level and flow in the river section includes: Based on the deviation between the calculation results and the measured data of the river channel, the adjustment range of the roughness of the river section and the weighting coefficient is determined; according to the preset iteration step size and convergence threshold, the roughness of the river section and the weighting coefficient are dynamically corrected and the hydrodynamic calculation is re-executed to update the water level and flow rate of each node. In one embodiment, the system first compares the simulated water level with the measured water level at each river node, and then determines the adjustment amount of the roughness coefficient and weighting coefficient for each river segment based on the magnitude of the difference. After setting the initial iteration step size and water level convergence threshold, the river segment parameters are dynamically corrected according to the magnitude of the difference, and hydrodynamic calculations are performed again to update the water level and flow rate of all nodes. After multiple iterations, the water level deviation of all nodes reaches the preset convergence standard, forming a stable river water level and flow rate evolution sequence.

[0065] In another embodiment, it is assumed that the watershed contains 500 nodes, with an initial river roughness coefficient between 0.025 and 0.04, and weighting coefficients between 0.8 and 1.2. After the first iteration, the node water level deviation ranges from -0.15 meters to 0.12 meters. After five iterations, the water level deviation of most nodes converges to between -0.02 meters and 0.02 meters, with an average adjustment of approximately 0.003 for the river roughness coefficient and an average adjustment of approximately 0.05 for the weighting coefficients, with maximum adjustments reaching 0.007 and 0.12, respectively. Throughout the iteration process, the node water level and flow rate gradually approach the measured values, ensuring stable convergence of the calculation.

[0066] In the process of calculating and correcting the roughness of the river section and the weighting coefficient through multiple iterations, the iteration step size is gradually reduced according to the preset attenuation coefficient to ensure the stable convergence of the calculation process and to obtain the water level and flow evolution results that conform to the actual situation of the basin.

[0067] In one embodiment, after each iteration, the system reduces the adjustment range to make the river segment parameter correction more detailed and avoid instability caused by excessive one-time adjustments. As the iteration proceeds, the river segment water level and flow gradually approach the measured data, the entire simulation process remains stable, and a reliable river channel evolution sequence is obtained.

[0068] In another embodiment, assuming an initial iteration step size of 0.02, the iteration step size is gradually reduced in each round according to the decay ratio. After 7 rounds of iteration, the water level deviations at 500 nodes in the basin converge to between -0.018 meters and 0.019 meters, the flow deviations are between -1.2 and 1.5 cubic meters per second, the river roughness is adjusted by an average of approximately 0.004, and the weighting coefficients are adjusted by an average of approximately 0.06. The water level and flow evolution sequences after iterations highly match the actual measurement data and can be directly used for flood forecasting and water resource management.

[0069] Preferably, the confluence evolution of water flow between simulated nodes along the longitudinal channel and lateral inflow includes: Calculate the hydraulic radius between nodes and simulate the longitudinal river flow propagation process to form water transfer between nodes; collect the lateral inflow of tributaries and superimpose the lateral inflow into the longitudinal river channel to update the energy of the water level nodes; In one embodiment, the system acquires cross-sectional data of each node in the river channel and calculates the hydraulic radius of each node by combining the riverbed slope and channel width. Subsequently, the system simulates the water flow process longitudinally along the river channel, calculating the water transfer between nodes. Simultaneously, the system obtains lateral inflow from tributaries and superimposes it onto the corresponding river channel nodes, updating the node water level energy. This method provides real-time water level and flow changes for each node, offering fundamental data for subsequent hydrological analysis.

[0070] In another embodiment, it is assumed that the river channel has 200 longitudinal nodes, each with a hydraulic radius between 0.5 meters and 3.0 meters. The tributary inflow rate ranges from 0.2 cubic meters per second to 3.5 cubic meters per second. Calculations show that the water level change at each node ranges from 0.05 meters to 0.6 meters, and the longitudinal water transfer velocity is approximately 0.15 meters per second to 1.8 meters per second. Simulations demonstrate that the water volume at different nodes in the river channel is updated every minute, reflecting the combined effects of longitudinal flow and lateral inflow.

[0071] The simulation of longitudinal river flow propulsion is set with a hydraulic radius of 0.5m-3.0m and a flow velocity of 0.1m / s-2.0m / s; the riverbed roughness is set to 0.02-0.06, the longitudinal slope to 0.001-0.01, and the river width to 1m-20m.

[0072] In one embodiment, the system sets initial ranges for the hydraulic radius and flow velocity, and performs simulations in conjunction with riverbed roughness, longitudinal slope, and channel width. As the calculations proceed, the system continuously adjusts the hydraulic radius and flow velocity parameters to make the water level and flow rate results at each node close to the measured data, thereby obtaining a stable hydrological evolution sequence.

[0073] In another embodiment, the hydraulic radius is assumed to be between 0.5 m and 3.0 m, and the flow velocity is between 0.1 m / s and 2.0 m / s. The riverbed roughness is set to 0.02 to 0.06, the longitudinal slope to 0.001 to 0.01, and the channel width to 1 m to 20 m. Simulations show that the node water level varies between 0.03 m and 0.65 m, and the flow rate varies between 0.1 m / s and 3.8 m / s. By adjusting these parameters, the system achieves a reasonable distribution of node water level and flow rate, and ensures the stability of the longitudinal channel flow.

[0074] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0075] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A dynamic simulation method for the entire process of rainfall-runoff generation-confluence-evolution, characterized in that, Includes the following steps: Step S1: Obtain watershed topographic information and extract watershed boundaries; divide the watershed into grid cells based on the watershed boundaries and use the grid cells to set river nodes; Step S2: Collect rainfall data using a rain-measuring radar; Calculate runoff using rainfall data; The generated flow is input into a pre-trained distributed flow-in-flow model for flow analysis, and the flow rate is recorded. Step S3: Input the confluence flow and river nodes into the river flood evolution model, perform river evolution simulation, and generate river confluence data; Step S4: Determine the inflow volume into the main river channel based on the river confluence data; generate a flow time series based on the inflow volume into the main river channel; predict the peak flood flow based on the flow time series; Step S5: Simulate the evolution of river water level based on the peak flow, calculate the peak water level height, predict the peak arrival time, and generate a flood warning signal.

2. The dynamic simulation method for the entire process of rainfall-runoff generation-convergence-evolution according to claim 1, characterized in that, Step S1, which involves dividing the watershed into grid cells based on the watershed boundary and setting the river nodes using these grid cells, includes: Identify the water flow direction of each grid cell based on the watershed boundary and determine the water catchment relationship; divide the grid cells using the water catchment relationship, and arrange the node positions in the grid cells according to the preset node spacing; Assign river cross-sectional features based on node locations and establish node connection relationships; map the node connection relationships to grid cells to generate a node network and set river nodes.

3. The dynamic simulation method for the entire process of rainfall-runoff generation-convergence-evolution according to claim 1, characterized in that, Step S2, which involves collecting rainfall data using a rain-measuring radar, includes: The reflectivity factor is obtained using a rain-measuring radar; the rainfall intensity distribution is inverted based on the reflectivity factor; the rainfall intensity distribution is then processed into a time series to generate rainfall grid points at different time steps. The geographic coordinate system of the rain-measuring radar is read and added to the rainfall grid to generate a spatial distribution map of rainfall; the cumulative rainfall is calculated from the spatial distribution map to obtain the rainfall data.

4. The dynamic simulation method for the entire process of rainfall-runoff generation-convergence-evolution according to claim 1, characterized in that, Step S2, which involves calculating runoff using rainfall data, includes: Based on rainfall data, soil water evolution is simulated to update the upper soil moisture status; vegetation interception is assessed based on the upper soil moisture status, and upper surface runoff is calculated. The residual water is estimated by using the surface runoff in the upper layer, and the residual water is infiltrated into the lower soil layer to renew the soil moisture; the yield is calculated based on the soil moisture in the lower layer.

5. The dynamic simulation method for the entire process of rainfall-runoff generation-convergence-evolution according to claim 4, characterized in that, Based on the assessment of vegetation interception according to the upper soil moisture status, the calculation of upper surface runoff includes: Obtain the vegetation type of the target area; determine the vegetation interception parameters based on the vegetation type; calculate the vegetation interception amount based on the vegetation interception parameters and the upper soil moisture status; calculate the upper surface runoff based on the vegetation interception amount.

6. The dynamic simulation method for the entire process of rainfall-runoff generation-convergence-evolution according to claim 5, characterized in that, The calculation of vegetation interception based on vegetation interception parameters and upper soil moisture status includes: Determine canopy height based on vegetation type; assess maximum canopy interception based on canopy height; correct maximum canopy interception using vegetation interception parameters, and record the corrected canopy interception; The corrected canopy interception is converted into infiltrative water volume to obtain the effective soil water volume; the vegetation interception is calculated based on the effective soil water volume and the moisture status of the upper soil layer.

7. The dynamic simulation method for the entire process of rainfall-runoff generation-convergence-evolution according to claim 1, characterized in that, Step S3 involves inputting the confluence flow and river node data into the river flood evolution model, including: The confluence flow and river nodes are input into the river flood evolution model, which includes an information acquisition layer, a computational simulation layer, and a verification and adjustment layer. The information acquisition layer is used to read river cross-section information, obtain boundary conditions, and adjust basic parameters; the calculation and simulation layer is used to calculate hydraulic elements, simulate the confluence evolution of water flow between nodes along the longitudinal channel and lateral inflow, and form the dynamic spatiotemporal process of river level and flow; the verification and adjustment layer is used to verify the rationality of the calculated water level and flow, and adjust the river section roughness and weighting coefficient when the calculation results are unreasonable, and iteratively calculate to form a complete evolution process of river cross-section water level and flow.

8. The dynamic simulation method for the entire process of rainfall-runoff generation-convergence-evolution according to claim 7, characterized in that, When the calculation results are unreasonable, the roughness of the river section and the weighting coefficient are adjusted, and iterative calculations are performed to form a complete process of the evolution of water level and discharge in the river section, including: Based on the deviation between the calculation results and the measured data of the river channel, the adjustment range of the roughness of the river section and the weighting coefficient is determined; according to the preset iteration step size and convergence threshold, the roughness of the river section and the weighting coefficient are dynamically corrected and the hydrodynamic calculation is re-executed to update the water level and flow rate of each node. In the process of calculating and correcting the roughness of the river section and the weighting coefficient through multiple iterations, the iteration step size is gradually reduced according to the preset attenuation coefficient to ensure the stable convergence of the calculation process and to obtain the water level and flow evolution results that conform to the actual situation of the basin.

9. The dynamic simulation method for the entire process of rainfall-runoff generation-convergence-evolution according to claim 7, characterized in that, The simulation of the confluence evolution of water flow between nodes along the longitudinal channel and lateral inflow includes: Calculate the hydraulic radius between nodes and simulate the longitudinal river flow propagation process to form water transfer between nodes; collect the lateral inflow of tributaries and superimpose the lateral inflow into the longitudinal river channel to update the energy of the water level nodes; The simulation of longitudinal river flow propulsion is set with a hydraulic radius of 0.5m-3.0m and a flow velocity of 0.1m / s-2.0m / s; the riverbed roughness is set to 0.02-0.06, the longitudinal slope to 0.001-0.01, and the river width to 1m-20m.

10. A dynamic simulation system for the entire process of rainfall-runoff generation-convergence-evolution, characterized in that, For executing the dynamic simulation method of the entire process of rainfall-runoff generation-confluence-evolution as described in claim 1, the dynamic simulation system of the entire process of rainfall-runoff generation-confluence-evolution includes: The watershed unit division module is used to acquire watershed topographic information, extract watershed boundaries, divide watershed boundaries into grid units, and set river nodes using grid units; The runoff analysis module is used to collect rainfall data using rain-measuring radar; calculate runoff using the rainfall data; input the runoff into a pre-trained distributed runoff model for runoff analysis; and record the runoff flow. The river evolution simulation module is used to input the confluence flow and river nodes into the river flood evolution model, perform river evolution simulation, and generate river confluence data; The flood peak prediction module is used to determine the inflow volume into the main river channel based on river confluence data; generate a flow time series based on the inflow volume; and predict the flood peak flow based on the flow time series. The early warning generation module is used to simulate the evolution of river water level based on the peak flow, calculate the peak water level height, predict the arrival time of the peak, and generate flood early warning signals.

Images