Slope water and soil loss process simulation method based on digital twinning

Abstract

This invention discloses a method for simulating slope soil erosion processes based on digital twins, specifically in the field of digital twins. Through spatiotemporal registration of multi-source slope observation data, it unifies the spatial grid and time step of topography, soil, vegetation, rainfall, runoff, and sediment data, enabling time-backtracking of outlet monitoring data and establishing a precise correspondence between slope internal units and observation data. Based on the construction of the slope digital twin state volume, a unified state vector containing hydrological and erosion attributes is generated for each computational unit, achieving dynamic data integration. Through runoff-triggered state identification, it distinguishes between three states: no runoff, local water storage, and runoff, accurately defining the boundary of erosion simulation. It adopts a progressive erosion and sediment transport simulation, forming a complete computational chain of runoff identification, shear drive, soil stripping, sediment transport, and temporary retention. Based on digital twin closed-loop correction, it uses measured data to correct infiltration, erodibility, and transport parameters, achieving dynamic matching between simulation and measurement.

Description

Technical Field

[0001] This invention relates to the field of digital twin technology, and more specifically, to a method for simulating slope soil erosion processes based on digital twins. Background Technology

[0002] Existing slope soil erosion simulation technologies typically use slope topography, rainfall processes, soil properties, and vegetation cover data as basic inputs. First, the slope is divided into computational units based on a digital elevation model (DEM), and slope gradient, slope length, and runoff direction are extracted. Then, rainfall data is input into the model according to the simulation time step. At each simulation time point, the model calculates slope runoff based on rainfall, soil infiltration capacity, and slope runoff relationships. It further estimates soil stripping and sediment transport based on slope gradient, soil erodibility parameters, and runoff intensity. Simulation results are usually output as slope outlet runoff processes, slope outlet sediment processes, slope erosion intensity distribution, or soil loss distribution, reflecting the soil erosion process of the target slope under rainfall conditions.

[0003] Existing technologies typically use slope outlet monitoring data as model validation data during operation, lacking a process to extrapolate outlet runoff and sediment observation results back to the internal runoff generation time of the slope and participate in the calculation state update. This makes it difficult to form a closed-loop correlation between the internal calculation units of the slope and the outlet monitoring results. Existing technologies rely heavily on rainfall thresholds, infiltration thresholds, or single hydrological conditions to determine runoff generation status, making it difficult to simultaneously express the linkage between rainfall input, upstream water inflow, surface water depth, and effective infiltration capacity. In erosion and sediment transport simulation, existing technologies usually estimate sediment volume directly based on slope, runoff, and soil erodibility parameters, lacking a process to link runoff generation status, shear drive, initiation threshold, vegetation cover constraints, upstream sediment input, and temporary retention status into a progressive calculation chain. This results in an incomplete expression of the dynamic transmission relationship between slope hydrological processes, erosion processes, and sediment transport processes. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, the present invention provides a method for simulating slope soil erosion processes based on digital twins, which solves the problems mentioned in the background art through the following scheme.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for simulating slope soil erosion processes based on digital twins, comprising: S1, execute the spatiotemporal registration step of multi-source slope observation data, obtain the digital elevation model of the target slope, slope boundary data, soil sampling point data, vegetation cover remote sensing image, rainfall monitoring sequence, runoff outlet monitoring sequence and sediment outlet monitoring sequence, divide the target slope into multiple slope calculation units, and generate slope registration dataset, slope flow direction connection table and time-aligned observation data. S2, execute the slope digital twin state body construction step, establish the slope digital twin state body database based on the slope registration dataset, slope flow direction connection table and time-aligned observation data, and establish a digital twin state vector for each slope calculation unit; S3, execute the slope runoff triggering state identification step, read the digital twin state vector, and identify the non-runoff state, local water storage state and runoff state of the slope calculation unit based on rainfall input, upstream runoff input, effective infiltration capacity and surface water depth; S4, execute the slope erosion and sediment transport progression simulation step, calculate the slope shear drive, unit soil stripping, unit sediment volume and downstream transfer for the slope calculation unit in the runoff state, and temporarily store the upstream sediment input for the slope calculation unit in the local water storage state. S5 executes the slope digital twin closed-loop correction and simulation result output steps, compares the simulated runoff and simulated sediment volume at the slope outlet with the measured runoff and measured sediment volume at the slope outlet in the time-aligned observation data, calculates the digital twin closed-loop residual, and corrects the effective infiltration capacity, soil erodibility parameters and transport coefficients in the slope digital twin state database according to the runoff deviation direction and sediment deviation direction, and outputs the simulation results of the slope soil and water loss process.

[0006] The technical effects and advantages of this invention are as follows: This scheme employs a spatiotemporal registration process using multi-source slope observation data to uniformly map digital elevation models, soil sampling point data, vegetation cover remote sensing images, rainfall monitoring sequences, runoff outlet monitoring sequences, and sediment outlet monitoring sequences to slope calculation units and simulation times. Furthermore, it uses the confluence path length and runoff propagation velocity to perform time-based back-calculation of measured runoff and sediment volume at the slope outlet, establishing a correspondence between the outlet monitoring data and the runoff generation times within the slope's internal calculation units. Consequently, the slope's digital twin state can utilize time-aligned observation data to correct effective infiltration capacity, soil erodibility parameters, and transport coefficients during subsequent closed-loop calibration. This transforms the simulation process from a simple one-way calculation and result verification into a state update process driven by observation data. This scheme, through the steps of constructing a digital twin state volume of the slope and identifying the slope runoff triggering state, organizes soil moisture content, surface water depth, unit runoff, unit sediment load, vegetation cover, surface roughness, slope, and upstream input runoff into a unified digital twin state vector. It then uses a slope runoff triggering index to distinguish between non-runoff states, localized water storage states, and runoff states. This approach allows runoff determination to simultaneously correlate rainfall input, upstream input runoff, effective infiltration capacity, and surface water depth. It enables the continuous representation of the slope's process from rainfall input, infiltration absorption, surface water storage, to runoff formation within a single state volume, providing clear state boundaries and data entry points for subsequent erosion and sediment transport progression simulations. This scheme employs a progressive simulation step for slope erosion and sediment transport, treating the slope calculation unit with runoff generation as the sediment calculation object. It sequentially performs calculations of slope shear drive, initiation shear threshold determination, unit soil stripping calculation, upstream sediment input accumulation, temporary retention status recording, and downstream transport record writing. This approach ensures that sediment simulation is no longer directly determined by a single slope or runoff volume, but rather forms a progressive calculation process under the constraints of a slope flow direction connection table. This process includes runoff identification, shear drive, soil stripping, sediment transport, temporary retention, and outlet collection, thus providing a consistent data source and update path for the water and sediment transport relationships between slope calculation units, the sediment temporary storage relationships under local water storage conditions, and the outlet sediment process. Attached Figure Description

[0007] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0008] Figure 2 This is a flowchart of the spatiotemporal registration process for multi-source observation data of slopes according to the present invention.

[0009] Figure 3 This is a flowchart of the slope runoff triggering state identification process of the present invention.

[0010] Figure 4 This is a flowchart illustrating the slope erosion and sediment transport progression simulation of the present invention.

[0011] Figure 5 This is a flowchart of the slope digital twin closed-loop correction and simulation result output of the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] refer to Figures 1-5 The digital twin-based slope soil erosion process simulation method shown includes: S1: Spatiotemporal registration step of multi-source observation data of slope: used to unify the measured rainfall data, topographic data, soil data, vegetation cover data, surface roughness data, slope runoff data and slope sediment data into the same spatial grid and the same simulation time step, forming a slope registration dataset that can be called by subsequent steps; In this step, the slope is divided into several slope calculation units. Each slope calculation unit has a fixed spatial number and corresponds to slope, slope length, aspect, soil texture, initial moisture content, vegetation cover, surface roughness, and elevation. Rainfall data is segmented according to the simulation time step, and runoff and sediment data are mapped to adjacent simulation time points according to the sampling time. For runoff and sediment observation data with time delays, time backtracking is performed using the slope runoff path length and runoff propagation velocity to bring the runoff and sediment data back to their corresponding slope runoff generation time.

[0014] When implementing the spatiotemporal registration step of the multi-source observation data of the slope, the digital elevation model, slope boundary data, soil sampling point data, vegetation cover remote sensing image, rainfall monitoring sequence, runoff outlet monitoring sequence and sediment outlet monitoring sequence of the target slope are first acquired, and the digital elevation model, slope boundary data, soil sampling point data and vegetation cover remote sensing image are converted to the same coordinate system.

[0015] The target slope is trimmed according to the slope boundary data. The trimmed target slope is divided into multiple slope calculation units according to the preset grid size. Each slope calculation unit is configured with a unique slope calculation unit number. The slope calculation unit number is kept consistent in the steps of spatiotemporal registration of multi-source observation data of slope, construction of digital twin state of slope, identification of slope runoff triggering state, simulation of slope erosion and sediment transport progression, and closed-loop correction and simulation result output of digital twin of slope.

[0016] For each slope calculation unit, its elevation, slope, aspect, slope length, confluence direction, and confluence path length are extracted from the digital elevation model. The downstream slope calculation unit number is determined based on the confluence direction. All slope calculation unit numbers, their corresponding downstream slope calculation unit numbers, and confluence path lengths are written into a slope flow direction connection table. For slope calculation units located at the slope outlet, their downstream slope calculation unit numbers are marked as the slope outlet number. This slope flow direction connection table is used in subsequent steps to determine the source relationship between upstream runoff and upstream sediment input.

[0017] When spatially mapping soil sampling point data, the location of each soil sampling point is overlaid with a slope calculation unit. When a soil sampling point falls into a slope calculation unit, the soil texture, soil bulk density, saturated water content, initial water content, and soil erodibility parameters of that soil sampling point are written into that slope calculation unit. When multiple soil sampling points exist within a slope calculation unit, the arithmetic mean of the same soil parameter from multiple soil sampling points within that slope calculation unit is taken. When no soil sampling points exist within a slope calculation unit, the three soil sampling points closest to the center of that slope calculation unit are selected, and the soil parameters of that slope calculation unit are calculated according to the inverse distance weight, and the calculated soil parameters are written into the slope registration dataset.

[0018] When spatially mapping vegetation cover remote sensing images, the images are overlaid with slope calculation cells, and the number of vegetation-covered pixels and the total number of pixels within each cell are counted. The vegetation cover rate of each slope calculation cell is determined based on the number of vegetation-covered pixels and the total number of pixels, and this rate is then entered into the slope registration dataset. For slope calculation cells with invalid pixels in the vegetation cover remote sensing image, only valid pixels are used to calculate the vegetation cover rate. When no valid pixels exist within a slope calculation cell, the average vegetation cover rate of spatially adjacent slope calculation cells with the same slope aspect is entered into that slope calculation cell, and a missing vegetation cover data marker is added to that slope calculation cell.

[0019] When performing time registration on rainfall monitoring sequences, a simulation time step is set, and the rainfall monitoring sequences are resampled according to the simulation time step. For any simulation moment, the rainfall monitoring values ​​within the corresponding time interval are accumulated to form the rainfall input, and the rainfall input is written into the slope registration dataset. The rainfall input is written into all slope calculation units at the same simulation moment as input data for the subsequent slope runoff triggering state identification step. When there is a missing rainfall monitoring sequence within the corresponding time interval of a certain simulation moment, linear interpolation is performed using the previous and next valid rainfall input values ​​of that simulation moment, and a rainfall data missing marker is added to that simulation moment.

[0020] When performing time registration on the runoff outlet monitoring sequence and the sediment outlet monitoring sequence, the runoff outlet monitoring sequence and the sediment outlet monitoring sequence are first resampled according to the simulated time step to form the measured runoff and measured sediment volume at the slope outlet. Since the measured runoff and measured sediment volume at the slope outlet correspond to the slope outlet monitoring time, they need to be extrapolated back to the corresponding runoff generation time inside the slope. For the i-th slope calculation unit, the runoff propagation time is calculated based on the confluence path length from the slope calculation unit to the slope outlet and the runoff propagation velocity. ; in, The runoff propagation time from the i-th slope calculation unit to the slope outlet; The length of the confluence path from the i-th slope calculation unit to the slope outlet; Let be the runoff propagation velocity corresponding to the i-th slope calculation unit, where i is the slope calculation unit number.

[0021] In obtaining Then, subtract the outlet monitoring time corresponding to the measured runoff and measured sediment volume at the slope outlet. The time backtracking time of the i-th slope calculation unit is obtained. If the time backtracking time falls between two adjacent simulation times, the measured runoff and measured sediment volume at the slope outlet are allocated to the two adjacent simulation times according to the time distance; if the time backtracking time coincides with a certain simulation time, the measured runoff and measured sediment volume at the slope outlet are written into that simulation time.

[0022] After completing the time backtracking, time-aligned observation data is generated, which includes the slope calculation unit number, the corresponding simulation time number, the measured runoff at the slope outlet, the measured sediment volume at the slope outlet, and the time backtracking marker.

[0023] When generating the slope registration dataset, the slope calculation unit number, elevation, slope, aspect, slope length, confluence direction, confluence path length, downstream slope calculation unit number, soil texture, soil bulk density, saturated moisture content, initial moisture content, soil erodibility parameters, vegetation cover, rainfall input, measured runoff at the slope outlet, measured sediment load at the slope outlet, and data missing markers for each slope calculation unit are uniformly written into the same data structure.

[0024] The slope registration dataset is indexed according to the slope calculation unit number and the simulation time number, so that the slope digital twin state body construction step can read static attributes according to the slope calculation unit number and read dynamic input data according to the simulation time number.

[0025] When outlier data is found in the slope registration dataset, it is marked and its inclusion in subsequent calculations is restricted. Outlier data includes spatial data extending beyond the slope boundary, data with timestamps earlier than the simulation start time, data with timestamps later than the simulation end time, data with soil moisture content exceeding saturation, data with vegetation cover less than zero, and data with vegetation cover greater than one. For marked outlier data, the outlier marker is retained in the slope registration dataset and treated as a missing input in the slope digital twin state body construction step. The spatiotemporal registration step of the multi-source slope observation data outputs the slope registration dataset, the slope flow direction connection table, and time-aligned observation data for use in the slope digital twin state body construction step.

[0026] S2: Construction steps of slope digital twin state body: used to construct a slope digital twin state body that is synchronously updated with the real slope. The slope digital twin state body stores static attributes, dynamic hydrological attributes, and dynamic erosion attributes according to slope calculation units, and provides a unified data entry point for S3 and S4; This step organizes the data for each slope computational unit into a digital twin state vector. The state vector simultaneously includes topography, soil, vegetation, rainfall, infiltration, runoff generation, shearing, soil stripping, and sediment transport. At each simulation time point, the slope digital twin state volume receives the slope registration dataset output by S1 and updates the upstream inflow and the state of the current unit according to the slope flow direction connection table.

[0027] When implementing the slope digital twin state body construction step, the slope registration dataset, slope flow direction connection table and time-aligned observation data output from the slope multi-source observation data spatiotemporal registration step are received, and a slope digital twin state body database is established in the storage layer of the computing device.

[0028] The slope digital twin state database includes a slope calculation unit table, a slope flow direction connection table, a slope dynamic hydrology table, and a slope dynamic erosion table. The slope calculation unit table stores the slope calculation unit number, elevation, slope, aspect, slope length, runoff path length, soil texture, soil bulk density, saturated moisture content, initial moisture content, soil erodibility parameters, vegetation cover, and surface roughness. The slope flow direction connection table stores the slope calculation unit number, downstream slope calculation unit number, upstream slope calculation unit number, and runoff path length. The slope dynamic hydrology table stores soil moisture content, surface water depth, unit runoff, upstream input runoff, rainfall input, and effective infiltration capacity according to the slope calculation unit number and simulation time number. The slope dynamic erosion table stores unit sediment load, upstream sediment input, slope shear drive, unit soil stripping load, transport coefficient, and temporary retention state according to the slope calculation unit number and simulation time number.

[0029] After the slope digital twin state database is established, the static attributes from the slope registration dataset are written into the slope computational unit table. These static attributes include elevation, slope, aspect, slope length, runoff path length, soil texture, soil bulk density, saturated water content, soil erodibility parameters, vegetation cover, and surface roughness. The slope computational unit number is used as the primary index during writing, ensuring that each slope computational unit has a unique data storage location within the slope digital twin state. For static attributes in the slope registration dataset marked with outlier data, the corresponding missing data flag is recorded in the slope computational unit table, and the connection between the slope computational unit number and the slope flow direction join table is maintained.

[0030] When writing the slope flow direction connection table into the slope digital twin state database, the downstream slope calculation unit number of each slope calculation unit is determined according to the slope calculation unit number, and the upstream slope calculation unit number set of that slope calculation unit is generated in reverse. For slope calculation units that do not have an upstream slope calculation unit number, their upstream slope calculation unit number set is set to empty. For slope calculation units whose downstream slope calculation unit number is the slope outlet number, that slope calculation unit is regarded as the adjacent unit of the slope outlet. The slope digital twin state database calculates the upstream input runoff and upstream sediment input based on this connection relationship in subsequent simulation times.

[0031] At the first simulation time point, the slope digital twin state volume reads the initial moisture content of each slope computational unit from the slope registration dataset and writes it into the soil moisture content field of the slope dynamic hydrology table. The initial surface water depth, initial unit runoff, initial upstream input runoff, initial unit sediment load, initial upstream sediment input, initial slope shear drive, and initial unit soil stripping load of each slope computational unit are set to zero and written into the slope dynamic hydrology table and the slope dynamic erosion table. The rainfall input from the slope registration dataset is written into the slope dynamic hydrology table according to the simulation time point number, ensuring that each slope computational unit has a corresponding rainfall input at the same simulation time point.

[0032] For the i-th slope calculation unit at the t-th simulation time, establish a digital twin state vector: ; in, Let be the digital twin state vector of the i-th slope calculation unit at the t-th simulation time; Let be the soil moisture content of the i-th slope calculation unit at the t-th simulation time; Let be the surface water depth of the i-th slope calculation unit at the t-th simulation time; Let be the unit runoff of the i-th slope calculation unit at the t-th simulation time; Let be the sediment volume of the i-th slope calculation unit at the t-th simulation time; Let be the vegetation coverage of the i-th slope calculation unit; Let be the surface roughness of the i-th slope calculation unit; Let be the slope of the i-th slope calculation unit; Let i be the upstream runoff of the i-th slope calculation unit at the t-th simulation time, where i is the slope calculation unit number and t is the simulation time number.

[0033] When updating the digital twin state of the slope at each simulation time, the set of upstream slope computational unit numbers flowing towards the i-th slope computational unit is first read from the slope flow direction connection table. If this set is empty, the upstream input runoff of the i-th slope computational unit at simulation time t is then used. Set to zero. If the set is not empty, read the unit runoff output of each upstream slope calculation unit in the set at the previous simulation time, accumulate them, and write them into the upstream input runoff field of the i-th slope calculation unit at the t-th simulation time. After the upstream input runoff is written, the slope digital twin state body synchronously updates the digital twin state vector. .

[0034] During dynamic hydrological status updates, the rainfall input at the t-th simulation time is read from the slope registration dataset and written into the slope dynamic hydrological table. The slope digital twin state body reads the soil moisture content of the previous simulation time, the rainfall input at the current simulation time, the upstream input runoff, saturated moisture content, and effective infiltration capacity at the current simulation time from the ith slope calculation unit, and updates the soil moisture content and surface water depth at the current simulation time based on this data. If the updated soil moisture content reaches the saturated moisture content, the water volume exceeding the effective infiltration capacity is written into the surface water depth field and participates in the slope runoff triggering index calculation in the subsequent slope runoff triggering state identification step. If the updated soil moisture content does not reach the saturated moisture content, the surface water depth at that simulation time is written into the slope dynamic hydrological table according to the remaining water volume after infiltration.

[0035] During dynamic erosion state updates, the slope digital twin state body reads the set of upstream slope calculation unit numbers flowing towards the i-th slope calculation unit from the slope flow direction connection table. If this set is empty, the upstream sediment input of the i-th slope calculation unit at simulation time t is set to zero. If the set is not empty, the unit sediment input of each upstream slope calculation unit in the set at the previous simulation time is read and written into the slope dynamic erosion table as the upstream sediment input source for the current simulation time. The upstream sediment input is used together with the unit soil stripping amount in the slope erosion and sediment transport progression simulation step to calculate the unit sediment input.

[0036] When a slope calculation unit in the slope registration dataset has missing rainfall data, missing vegetation cover data, missing soil parameter data, or abnormal data markers at the current simulation time, the slope digital twin state body does not delete the slope calculation unit, nor does it change the upstream and downstream connection relationships of the slope calculation unit in the slope flow direction connection table. For missing dynamic input data, the slope digital twin state body retains the corresponding state value of the slope calculation unit at the previous simulation time and writes a missing marker in the slope dynamic hydrology table or slope dynamic erosion table. For missing static attributes, the slope digital twin state body calls the data that has been interpolated in the slope registration dataset and retains the interpolation markers. The missing markers and interpolation markers are used to determine the runoff residual weights and sediment residual weights in the slope digital twin closed-loop correction and simulation result output steps.

[0037] After completing the state writing at the t-th simulation time, the slope digital twin state body generates digital twin state vectors one by one according to the slope calculation unit number, and writes the digital twin state vectors into the slope digital twin state body database. Each digital twin state vector is bound and stored with the slope calculation unit number, simulation time number, and data missing marker. The slope runoff triggering state identification step reads the digital twin state vector according to the slope calculation unit number and simulation time number, and uses the soil moisture content, surface water depth, unit runoff, vegetation cover, surface roughness, slope, and upstream input runoff to determine the subsequent runoff state. The slope digital twin state body construction step outputs the slope digital twin state body database and the set of digital twin state vectors at the current simulation time for the slope runoff triggering state identification step to call.

[0038] S3: Slope Runoff Triggering State Identification Step: This step determines whether each slope calculation unit has entered a runoff generating state at the current simulation moment. This step incorporates rainfall input, soil moisture content, infiltration capacity, upstream water inflow, and surface water depth into the judgment. Through this progressive judgment process, the slope digital twin state can distinguish between non-runoff generating state, localized water storage state, and runoff generating state. This step reads the digital twin state vector of the slope from S2, first calculates the effective infiltration capacity at the current simulation moment, and then determines whether the rainfall input and upstream runoff input exceed the absorption capacity of the slope calculation unit. When the absorption capacity is exceeded, the slope calculation unit is marked as a runoff generation trigger unit; when the absorption capacity is not exceeded, only the soil moisture content and surface water depth are updated, and erosion and sediment transport calculations are not performed.

[0039] When implementing the slope runoff triggering state identification step, the digital twin state vector of each slope calculation unit at the current simulation time is first read from the slope digital twin state body database output by the slope digital twin state body construction step. The digital twin state vector includes soil moisture content, surface water depth, unit runoff, upstream input runoff, vegetation cover, surface roughness, slope, and effective infiltration capacity. For each slope calculation unit, the effective infiltration capacity is calculated based on its current soil moisture content and vegetation cover. The effective infiltration capacity It is obtained by combining soil moisture content and saturated moisture content, vegetation cover and surface roughness, and is used to reflect the ability of the slope calculation unit to absorb rainfall and upstream runoff at the current simulation moment.

[0040] Subsequently, based on the rainfall input of the slope calculation unit and upstream input runoff and the corrected effective infiltration capacity and current surface water depth Calculate the slope runoff triggering index : ; in, Let be the slope runoff triggering index of the i-th slope calculation unit at the t-th simulation time; This refers to the amount of rainfall input; For upstream input runoff; For effective infiltration capacity; λ is the surface water depth; λ is the surface water depth amplification factor, used to reflect the amplification effect of surface water depth on runoff triggering; the constant 1 in the denominator is used to avoid division by zero when the effective infiltration capacity is zero, and to maintain dimensional consistency.

[0041] The slope runoff triggering index was calculated. Then, it is compared with the preset flow trigger threshold. Comparison. When At that time, the slope calculation unit is marked as a non-runoff state, and the soil moisture content in the digital twin state vector is... and surface water depth Updated to the calculated value at the current simulation moment, unit runoff. and unit sediment volume Keep it at zero. When At that time, further determine the surface water depth at the current simulation moment. Has the preset surface water depth threshold been reached? ,like The slope calculation unit is marked as a local water storage state, and only the soil moisture content and surface water depth in the digital twin state vector are updated, while the unit runoff and unit sediment content remain zero; if The slope calculation unit is marked as the runoff-generating state, and its digital twin state vector is passed to the slope erosion and sediment transport progression simulation step to perform sediment stripping and transport calculations.

[0042] When updating the runoff generation status of slope calculation units, the runoff generation status of each slope calculation unit is simultaneously written into the status field of the slope digital twin state database. This allows for the selection of slope calculation units involved in soil stripping and sediment transport based on the runoff generation status during the slope erosion and sediment transport progression simulation step. For slope calculation units in non-runoff generation or locally impounded states, only their soil moisture content and surface water depth are updated to ensure that the hydrological status can be used to calculate effective infiltration capacity and slope runoff triggering index at the next simulation time step.

[0043] In the slope runoff triggering state identification step, for slope calculation units with missing or abnormal data markers, the slope digital twin state volume retains the state value from the previous simulation moment, and a missing marker is added to the digital twin state vector to ensure that missing or abnormal data does not interrupt the entire slope runoff triggering state identification process. The missing marker is used to adjust the runoff state weight and correction coefficient in the slope digital twin closed-loop correction and simulation result output step.

[0044] After completing the calculation of the slope runoff triggering index and state determination for all slope calculation units, the updated digital twin state vector set is output to the progressive slope erosion and sediment transport simulation step. This is used to progressively calculate the slope shear drive, unit soil stripping, and unit sediment load for each runoff-generating slope calculation unit, and then transfer these values ​​to downstream slope calculation units along the slope flow direction connection table. The slope runoff triggering state identification step ensures that the slope digital twin state is synchronized with the actual slope conditions in terms of hydrological and runoff states, providing accurate data input for subsequent progressive simulations.

[0045] When implementing the slope erosion and sediment transport progression simulation step, the updated digital twin state vector set output from the slope runoff generation trigger state identification step is received, and runoff-generating slope calculation units are selected from this digital twin state vector set. For slope calculation units in non-runoff-generating state and slope calculation units in locally impounded state, no new soil stripping amount calculation is performed; for slope calculation units in runoff-generating state, the upstream slope calculation unit number set and downstream slope calculation unit number at the current simulation time are determined according to the slope flow direction connection table, and the calculation is performed in the order from upstream slope calculation units to downstream slope calculation units, so that the unit sediment amount of the current slope calculation unit can be used as the upstream sediment input amount of the downstream slope calculation unit in subsequent calculations.

[0046] For the i-th slope calculation unit with runoff generation status, at the t-th simulation time, the unit runoff of that slope calculation unit is read from the slope digital twin state database. ,slope Vegetation coverage , surface roughness Soil erodibility parameters The upstream sediment input and temporary retention status. The unit runoff volume. The hydrological status update result is derived from the runoff generation trigger state identification step on the slope; the slope Vegetation coverage , surface roughness and soil erodibility parameters The upstream sediment input is derived from the slope calculation unit table written in the slope digital twin state body construction step; the upstream sediment input is derived from the unit sediment output of the upstream slope calculation unit flowing to the i-th slope calculation unit in the slope flow direction connection table at the previous simulation time.

[0047] After reading the input data of the slope calculation unit for runoff generation, the slope shear drive of the i-th slope calculation unit at the t-th simulation time is calculated. The slope shear drive is used to characterize the degree of runoff's stripping effect on the slope soil at the current simulation time. The calculation incorporates unit runoff volume, slope, and surface roughness into the same calculation process, specifically: ; in, Let be the slope shear drive quantity of the i-th slope calculation unit at the t-th simulation time; This is the shear conversion factor; Let be the unit runoff of the i-th slope calculation unit at the t-th simulation time; Let be the slope of the i-th slope calculation unit; Let be the surface roughness of the i-th slope calculation unit.

[0048] After the slope shear drive is calculated, With the starting shear threshold Comparison. When Not achieved When the t-th simulation time is t, the newly added soil stripping amount of the i-th slope calculation unit is set to zero, and the upstream sediment input amount of the slope calculation unit continues to be read; when achieve At that time, the unit soil stripping amount of the slope calculation unit is calculated. The unit soil stripping amount is determined by the portion of the slope shear drive exceeding the initiation shear threshold, the soil erodibility parameter, and the vegetation cover rate, specifically as follows: ; in, Let be the amount of soil stripping in the i-th slope calculation unit at the t-th simulation time; For the i-th slope calculation unit, the soil erodibility parameter is denoted as . Let be the slope shear drive quantity of the i-th slope calculation unit at the t-th simulation time; This is the threshold for initiating shearing; Let be the vegetation cover rate of the i-th slope calculation unit. (Function) This is used to limit the amount of new soil stripping to zero when the slope shear drive amount does not reach the initiation shear threshold.

[0049] After obtaining the unit soil stripping amount, the set of upstream slope calculation unit numbers flowing to the i-th slope calculation unit is read from the slope flow direction connection table. .when When the set is empty, the upstream sediment input of the i-th slope calculation unit is set to zero. Read the collection when it is not empty. The sediment load of each upstream slope calculation unit at the previous simulation time is calculated, and the corresponding transport coefficient is read. The upstream sediment load processed by the transport coefficient is accumulated to obtain the upstream sediment input of the i-th slope calculation unit at the current simulation time. The sediment load of the i-th slope calculation unit at the t-th simulation time is calculated by the following formula: ; in, Let be the sediment volume of the i-th slope calculation unit at the t-th simulation time; Let be the amount of soil stripping in the i-th slope calculation unit at the t-th simulation time; Let j be the set of upstream slope calculation unit numbers flowing towards the i-th slope calculation unit; j is the set The upstream slope calculation unit number within; The sediment transport coefficient is used to transfer sediment from the j-th slope calculation unit to the i-th slope calculation unit; Let be the sediment volume of the j-th slope calculation unit at the previous simulation time.

[0050] When the i-th slope calculation unit is in a runoff-producing state and the slope shear drive reaches the initiation shear threshold, the calculated unit soil stripping amount will be... Slope shear drive Unit sediment volume and transport coefficient Write the dynamic erosion table of the slope and record the unit sediment volume. The sediment input is transferred to the downstream slope calculation unit according to the slope flow direction connection table, serving as the upstream sediment input source for the downstream slope calculation unit at the current simulation time or the next simulation time. During the transfer, if the downstream slope calculation unit number of the i-th slope calculation unit is the slope outlet number, then the unit sediment quantity is transferred. Write the simulated sediment volume record at the slope outlet; if the downstream slope calculation unit number of the i-th slope calculation unit corresponds to a slope calculation unit inside the slope, then record the sediment volume of the unit. Write the upstream sediment input field into the downstream slope calculation unit.

[0051] When the i-th slope calculation unit is in a runoff-producing state but the slope shear drive has not reached the initiation shear threshold, the unit soil stripping amount of this slope calculation unit is... The current sediment load is set to zero, and it is determined whether there is upstream sediment input in the slope calculation unit. If upstream sediment input exists, the upstream sediment input is written into the temporary retention state of the slope calculation unit, and the temporary retention state is bound and stored with the slope calculation unit number and simulation time number. If there is no upstream sediment input, the current unit sediment load of the slope calculation unit is set to zero, and the zero new sediment state is written into the slope dynamic erosion table. The temporary retention state continues to be retained in subsequent simulation time steps until the slope calculation unit re-enters the runoff generation state and participates in the unit sediment load calculation.

[0052] When the i-th slope calculation unit is in a locally impounded state, the slope shear drive and newly added soil stripping are not calculated, and the unit soil stripping amount at the current simulation time is written as zero. If the locally impounded slope calculation unit receives upstream sediment input, the upstream sediment input is written to the temporary retention state; if no upstream sediment input is received, the unit sediment amount at the current simulation time is kept at zero. This processing ensures that the locally impounded slope calculation unit does not participate in sediment output before a runoff-generating state is formed, and only saves the upstream sediment input information related to it.

[0053] After all slope calculation units in the runoff generation state have completed the calculation of slope shear drive, unit soil stripping, unit sediment volume, and downstream transport, the slope erosion and sediment transport progression simulation step writes the slope shear drive, unit soil stripping, unit sediment volume, upstream sediment input, temporary retention status, and downstream transport records of each slope calculation unit at the current simulation time back to the slope digital twin state volume database. The slope digital twin state volume database saves this result according to the slope calculation unit number and simulation time number, enabling the slope digital twin closed-loop correction and simulation result output steps to read the simulated sediment volume at the slope outlet, the sediment distribution at the slope calculation unit level, the slope soil stripping distribution, and the slope sediment transport path at the current simulation time.

[0054] S5: Slope Digital Twin Closed-Loop Correction and Simulation Result Output Step: This step compares the simulated runoff and simulated sediment volume obtained in S4 with the measured runoff and measured sediment volume after time alignment, and writes the deviation back into the slope digital twin state body. This step adjusts the effective infiltration capacity, transport coefficient, and soil erodibility parameters through closed-loop correction, so that the slope digital twin state body at the next simulation time can continue to participate in runoff generation trigger identification and erosion and sediment transport progression simulation; This step does not directly replace the simulated values ​​with measured values, but rather adjusts the hydrological and erosion parameters based on runoff and sediment deviations, respectively. Runoff deviation is mainly used to correct for effective infiltration capacity and surface water depth; sediment deviation is mainly used to correct for soil erodibility parameters and transport coefficients.

[0055] During the implementation of the slope digital twin closed-loop correction and simulation result output steps, the current simulation time data written back from the slope erosion and sediment transport progression simulation step is received from the slope digital twin state volume database, and the simulated runoff at the slope outlet at simulation time t is read from the slope digital twin state volume database. Simulated sediment volume at slope outlet The simulated runoff at the slope outlet. The simulated sediment volume at the slope outlet is obtained by collecting the unit runoff output from the downstream slope calculation unit numbered as the slope outlet number. The sediment volume is collected from the downstream slope calculation unit numbered with the slope outlet number. During reading, the simulation time number t is used as the time index, and the slope outlet number is used as the spatial index, so that the simulated runoff and simulated sediment volume at the slope outlet correspond to the same simulation time in the time-aligned observation data.

[0056] Subsequently, the measured runoff at the slope outlet at time t is read from the time-aligned observation data generated by the spatiotemporal registration step of the multi-source slope observation data. Measured sediment volume at the slope outlet When time-aligned observation data contains time-backtracking markers, the corresponding measured runoff and sediment load at the slope outlet are read according to the time-backtracking time determined in the spatiotemporal registration steps of the multi-source slope observation data. When time-aligned observation data does not contain time-backtracking markers, the measured runoff and sediment load at the slope outlet are read directly according to the simulated time number. For measured runoff or sediment loads at the slope outlet with missing markers, they are not directly used as complete observation values ​​for correction. Instead, the corresponding runoff residual weights or sediment residual weights are reduced according to the missing markers to ensure the closed-loop correction process remains continuous.

[0057] After reading the simulated slope outlet data and time-aligned observation data, the digital twin closed-loop residual at the t-th simulation time is calculated. This digital twin closed-loop residual, composed of runoff and sediment residuals, is used to express the deviation between the digital twin state of the slope at the current simulation time and the actual slope outlet monitoring results. Specifically: ; in, The digital twin closed-loop residual at simulation time t; Let t be the measured runoff at the slope outlet at the t-th simulation time. Let be the simulated runoff at the slope outlet at simulation time t; The measured sediment load at the slope outlet at the t-th simulation time; Let t be the simulated sediment load at the slope outlet at the t-th simulation time. The weights are the runoff residuals. The weights are the sediment residuals, and .

[0058] The runoff residual weight The sediment residual weights are determined based on missing data markers for measured runoff at the slope outlet and the confidence level of the runoff monitoring equipment. The weighting is determined based on missing data markers for measured sediment load at the slope outlet and the confidence level of the sediment monitoring equipment. When both the measured runoff volume and the measured sediment load at the slope outlet have missing data markers, the runoff residual weighting is increased. Reduce sediment residual weight When there are missing markers for the measured runoff at the slope outlet but no missing markers for the measured sediment load at the slope outlet, reduce the weight of the runoff residual. Increase the weight of sediment residuals When neither of them has missing markers, the weight ratio of the two is determined according to the confidence level of the runoff monitoring equipment and the confidence level of the sediment monitoring equipment, and the sum of their weights is kept to be one.

[0059] After completing the closed-loop residual calculation of the digital twin, the digital twin state of the slope is corrected according to the direction of runoff deviation and the direction of sediment deviation, respectively. For the direction of runoff deviation, a comparison is made... and .when When the simulated runoff at the slope outlet is determined to be less than the measured runoff at the slope outlet at the current simulation time, the digital twin state of the slope reduces the effective infiltration capacity of the corresponding slope calculation unit in the next simulation time, and writes the reduced effective infiltration capacity into the slope dynamic hydrological table; when When the simulated runoff at the slope outlet is greater than the measured runoff at the slope outlet at the current simulation time, the digital twin state of the slope increases the effective infiltration capacity of the corresponding slope calculation unit in the next simulation time, and writes the increased effective infiltration capacity into the slope dynamic hydrological table; when At that time, the effective infiltration capacity of the corresponding slope calculation unit remains unchanged, and the current effective infiltration capacity is used as the initial value of the effective infiltration capacity at the next simulation time.

[0060] Regarding the direction of sediment deviation, compare and .when When the simulated sediment volume at the slope outlet is determined to be less than the measured sediment volume at the slope outlet at the current simulation time, the digital twin state of the slope increases the soil erodibility parameter or transport coefficient of the corresponding slope calculation unit in the next simulation time, and writes the increased soil erodibility parameter or transport coefficient into the slope calculation unit table or the slope dynamic erosion table; when When the simulated sediment volume at the slope outlet is greater than the measured sediment volume at the slope outlet at the current simulation time, the digital twin state of the slope will reduce the soil erodibility parameter or transport coefficient of the corresponding slope calculation unit in the next simulation time, and write the reduced soil erodibility parameter or transport coefficient into the slope calculation unit table or the slope dynamic erosion table; when At that time, the soil erodibility parameters and transport coefficients of the corresponding slope calculation unit are kept unchanged, and the current soil erodibility parameters and current transport coefficients are used as the initial values ​​of the erosion state at the next simulation time.

[0061] When identifying slope calculation units requiring correction, the slope digital twin state body traces upstream from the slope outlet number based on the slope flow direction connection table to obtain a set of slope calculation unit numbers that participate in the simulated runoff and sediment volume at the current slope outlet. For each slope calculation unit in this set, the correction target is determined according to its unit runoff, unit sediment volume, unit soil stripping, and transport path records at the current simulation time. For slope calculation units with a high degree of participation in unit runoff, effective infiltration capacity is adjusted first; for slope calculation units with a high degree of participation in unit soil stripping and unit sediment volume, soil erodibility parameters or transport coefficients are adjusted first. For slope calculation units with missing markers, the missing markers are retained, and the parameter adjustment range is reduced to avoid missing inputs directly changing the main parameters of the slope digital twin state body.

[0062] After correcting the effective infiltration capacity, soil erodibility parameters, and transport coefficients, the corrected parameters are bound and stored with the corresponding slope calculation unit number, simulation time number, and correction source marker. The correction source marker is used to record whether the correction was triggered by the runoff deviation direction, the sediment deviation direction, or both. Subsequently, the slope digital twin state body writes the corrected effective infiltration capacity into the slope dynamic hydrological table for the next simulation time, writes the corrected soil erodibility parameters or transport coefficients into the slope dynamic erosion table for the next simulation time, and uses the soil moisture content, surface water depth, unit runoff, unit sediment load, temporary retention state, and runoff generation state at the current simulation time as the initial basis for the state update at the next simulation time.

[0063] When outputting the simulation results of slope soil erosion, the slope digital twin state body generates slope calculation unit-level runoff distribution, slope calculation unit-level sediment distribution, slope runoff triggering state, slope soil stripping distribution, slope sediment transport path, and slope outlet runoff-sediment process line according to the slope calculation unit number and simulation time number. The slope calculation unit-level runoff distribution is formed by the unit runoff volume of each slope calculation unit; the slope calculation unit-level sediment distribution is formed by the unit sediment volume of each slope calculation unit; the slope runoff triggering state is formed by the non-runoff state, local water storage state, and runoff state written in the slope runoff triggering state identification step; the slope soil stripping distribution is formed by the unit soil stripping volume calculated by the slope erosion and sediment transport progression simulation step; the slope sediment transport path is formed by the slope flow direction connection table and downstream transfer records; and the slope outlet runoff-sediment process line is formed by arranging the slope outlet simulated runoff volume and slope outlet simulated sediment volume at each simulation time in chronological order.

[0064] After completing the slope digital twin closed-loop correction and simulation result output steps at simulation time t, it is determined whether there is a next simulation time. If there is a next simulation time, the corrected slope digital twin state volume is used as the input state for the next simulation time, and the process returns to the slope multi-source observation data spatiotemporal registration step or continues to read the data of the next simulation time that has already been registered. If there is no next simulation time, the simulation results of the slope soil and water loss process within the complete simulation cycle are output, and the slope digital twin state volume database, slope outlet runoff and sediment process line, slope sediment transport path, and digital twin closed-loop residuals for each simulation time are saved.

[0065] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other. In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for simulating a slope water and soil loss process based on digital twinning, characterized in that, include: S1, execute the spatiotemporal registration step of multi-source slope observation data, obtain the digital elevation model of the target slope, slope boundary data, soil sampling point data, vegetation cover remote sensing image, rainfall monitoring sequence, runoff outlet monitoring sequence and sediment outlet monitoring sequence, divide the target slope into multiple slope calculation units, and generate slope registration dataset, slope flow direction connection table and time-aligned observation data. S2, execute the slope digital twin state body construction step, establish the slope digital twin state body database based on the slope registration dataset, slope flow direction connection table and time-aligned observation data, and establish a digital twin state vector for each slope calculation unit; S3, execute the slope runoff triggering state identification step, read the digital twin state vector, and identify the non-runoff state, local water storage state and runoff state of the slope calculation unit based on rainfall input, upstream runoff input, effective infiltration capacity and surface water depth; S4, execute the slope erosion and sediment transport progression simulation step, calculate the slope shear drive, unit soil stripping, unit sediment volume and downstream transfer for the slope calculation unit in the runoff state, and temporarily store the upstream sediment input for the slope calculation unit in the local water storage state. S5 executes the slope digital twin closed-loop correction and simulation result output steps, compares the simulated runoff and simulated sediment volume at the slope outlet with the measured runoff and measured sediment volume at the slope outlet in the time-aligned observation data, calculates the digital twin closed-loop residual, and corrects the effective infiltration capacity, soil erodibility parameters and transport coefficients in the slope digital twin state database according to the runoff deviation direction and sediment deviation direction, and outputs the simulation results of the slope soil and water loss process. 2.The slope water and soil loss process simulation method based on digital twinning according to claim 1, characterized in that, The spatiotemporal registration steps for multi-source observation data of the slope include: The digital elevation model, slope boundary data, soil sampling point data, and vegetation cover remote sensing image are converted to the same coordinate system; the target slope is cropped according to the slope boundary data; the cropped target slope is divided into multiple slope calculation units according to the preset grid size; a unique slope calculation unit number is assigned to each slope calculation unit; the elevation, slope, aspect, slope length, confluence direction, and confluence path length of each slope calculation unit are extracted from the digital elevation model; the downstream slope calculation unit number of each slope calculation unit is determined according to the confluence direction; and the slope calculation unit number, downstream slope calculation unit number, and confluence path length are written into the slope flow direction connection table. 3.The slope water and soil erosion process simulation method based on digital twinning according to claim 1, characterized in that, The spatiotemporal registration step for multi-source observation data of the slope also includes: Soil sampling point data is overlaid with slope calculation unit for judgment; when soil sampling point falls into slope calculation unit, soil texture, soil bulk density, saturated water content, initial water content and soil erodibility parameters are written into the slope calculation unit. When there are multiple soil sampling points within a slope calculation unit, the arithmetic mean of the same soil parameter is taken. When there are no soil sampling points within a slope calculation unit, the three soil sampling points closest to the center of the slope calculation unit are selected, and the soil parameters of the slope calculation unit are calculated according to the inverse distance weight. The vegetation cover remote sensing image is overlaid with the slope calculation unit, and the vegetation cover rate is determined based on the number of effective pixels covered by vegetation and the total number of effective pixels within the slope calculation unit. The vegetation cover rate is then written into the slope registration dataset. 4.The slope water and soil erosion process simulation method based on digital twinning according to claim 1, characterized in that, The spatiotemporal registration step for multi-source observation data of the slope also includes: The rainfall monitoring sequence was resampled according to the simulated time step, and the rainfall monitoring values ​​within the time interval corresponding to the simulated time were accumulated to form the rainfall input. The runoff outlet monitoring sequence and sediment outlet monitoring sequence were resampled according to the simulated time step to form the measured runoff and measured sediment volume at the slope outlet. The runoff propagation time was determined based on the confluence path length from the slope calculation unit to the slope outlet and the runoff propagation velocity. The outlet monitoring time corresponding to the measured runoff and measured sediment volume at the slope outlet was back-estimated to the runoff generation time corresponding to the slope calculation unit according to the runoff propagation time to generate time-aligned observation data. 5.The slope water-soil erosion process simulation method based on digital twinning according to claim 1, wherein, The steps for constructing the digital twin state of the slope include: A digital twin state database of slope is established in the storage layer of the computing device. The digital twin state database of slope includes a slope calculation unit table, a slope flow direction connection table, a slope dynamic hydrology table, and a slope dynamic erosion table. The slope calculation unit table stores the slope calculation unit number, elevation, slope, aspect, slope length, runoff path length, soil texture, soil bulk density, saturated moisture content, initial moisture content, soil erodibility parameters, vegetation cover, and surface roughness. The slope dynamic hydrology table stores soil moisture content, surface water depth, unit runoff, upstream input runoff, rainfall input, and effective infiltration capacity according to the slope calculation unit number and simulation time number. The slope dynamic erosion table stores unit sediment load, upstream sediment input, slope shear drive, unit soil stripping load, transport coefficient, and temporary retention status according to the slope calculation unit number and simulation time number. 6.The slope water-soil erosion process simulation method based on digital twinning according to claim 1, wherein, The steps for constructing the digital twin state of the slope also include: The upstream slope calculation unit number set is read from the slope flow direction connection table to the current slope calculation unit. When the upstream slope calculation unit number set is empty, the upstream input runoff and upstream sediment input of the current slope calculation unit at the current simulation time are set to zero. When the upstream slope calculation unit number set is not empty, the unit runoff output by the upstream slope calculation unit in the set at the previous simulation time is read and accumulated as the upstream input runoff of the current slope calculation unit. The unit sediment volume of the upstream slope calculation unit in the set at the previous simulation time is read and written into the upstream sediment input source of the current slope calculation unit. Soil moisture content, surface water depth, unit runoff, unit sediment volume, vegetation cover, surface roughness, slope, and upstream input runoff are combined into a digital twin state vector. 7.The slope water-soil erosion process simulation method based on digital twinning according to claim 1, wherein, The slope runoff triggering state identification step includes: The digital twin state vector of each slope computational unit at the current simulation moment is read from the slope digital twin state volume database; the effective infiltration capacity is determined based on soil moisture content, saturated moisture content, vegetation cover, and surface roughness; the slope runoff generation trigger index is determined based on rainfall input, upstream runoff input, effective infiltration capacity, and surface water depth; when the slope runoff generation trigger index does not reach the preset runoff generation trigger threshold, the corresponding slope computational unit is marked as non-runoff generation; when the slope runoff generation trigger index reaches the preset runoff generation trigger threshold and the surface water depth does not reach the preset surface water depth threshold, the corresponding slope computational unit is marked as locally impounded; when the slope runoff generation trigger index reaches the preset runoff generation trigger threshold and the surface water depth reaches the preset surface water depth threshold, the corresponding slope computational unit is marked as runoff generation, and its digital twin state vector is transferred to the slope erosion and sediment transport progression simulation step.

8. The method for simulating slope soil erosion based on digital twins according to claim 1, characterized in that, The steps for simulating slope erosion and sediment transport progression include: The algorithm selects runoff-generating slope calculation units from the updated digital twin state vector set; determines the upstream and downstream slope calculation unit number sets according to the slope flow direction connection table; reads the unit runoff, slope, vegetation cover, surface roughness, soil erodibility parameters, upstream sediment input, and temporary retention status in the order from upstream to downstream slope calculation units; calculates the slope shear drive based on the unit runoff, slope, and surface roughness; when the slope shear drive reaches the initiation shear threshold, calculates the unit soil stripping based on the slope shear drive, initiation shear threshold, soil erodibility parameters, and vegetation cover, and calculates the unit sediment volume in conjunction with the upstream sediment input; when the slope shear drive does not reach the initiation shear threshold and there is upstream sediment input, writes the upstream sediment input into the temporary retention status; when the downstream slope calculation unit number is the slope outlet number, writes the unit sediment volume into the slope outlet simulated sediment volume record.

9. The method for simulating slope soil erosion based on digital twins according to claim 1, characterized in that, The steps for outputting the digital twin closed-loop correction and simulation results of the slope include: Read the simulated runoff and simulated sediment volume at the slope outlet at the current simulation time; read the measured runoff and measured sediment volume at the slope outlet at the same simulation time from the time-aligned observation data; determine the digital twin closed-loop residual based on the measured runoff, simulated runoff, measured sediment volume, simulated sediment volume, runoff residual weight, and sediment residual weight; When the measured runoff at the slope outlet is greater than the simulated runoff at the slope outlet, the effective infiltration capacity of the slope calculation unit corresponding to the next simulation time is reduced; when the measured runoff at the slope outlet is less than the simulated runoff at the slope outlet, the effective infiltration capacity of the slope calculation unit corresponding to the next simulation time is increased; when the measured sediment load at the slope outlet is greater than the simulated sediment load at the slope outlet, the soil erodibility parameter or transport coefficient of the slope calculation unit corresponding to the next simulation time is increased; when the measured sediment load at the slope outlet is less than the simulated sediment load at the slope outlet, the soil erodibility parameter or transport coefficient of the slope calculation unit corresponding to the next simulation time is reduced; the output includes the runoff distribution at the slope calculation unit level, the sediment distribution at the slope calculation unit level, the slope runoff generation triggering state, the slope soil stripping distribution, the slope sediment transport path, and the slope outlet runoff and sediment process curve.

Images