Method for estimating soil loss amount in karst region

Abstract

The invention discloses a method for estimating the soil loss amount in a karst region, and the method comprises the steps: determining a negative exponential function relation between a rock exposure rate and the soil loss amount based on field in-situ observation; according to the negative exponential function relation, stony desertification factors used for quantifying the inhibition effect of rock exposure on soil erosion are constructed; introducing the stony desertification factor as a multiplicative factor into an RUSLE model, and establishing an RUSLE-D model; and obtaining each RUSLE factor data and rock exposure rate data of the target area, inputting the RUSLE-D model, and calculating the soil erosion amount of the target area. According to the invention, the accuracy and practicability of model estimation and prediction are improved.

Description

Technical Field

[0001] This invention belongs to the field of soil loss calculation technology in karst areas, and particularly relates to a method for estimating soil loss in karst areas. Background Technology

[0002] Karst regions are characterized by complex topography, fragmented surfaces, shallow and discontinuous soils, and extensive exposure of bedrock, resulting in a distinctly heterogeneous landscape. The unique geological structure and karst processes make soil erosion processes highly complex in these areas. Currently used soil erosion models, such as USLE (Universal Soil Erosion Model) / RUSLE (Modified Universal Soil Erosion Model), are mostly empirical models built upon long-term field observations in homogeneous areas (non-karst regions). However, when these models, based on homogeneous soil regions, are directly applied to karst areas, they often overestimate soil erosion. This is because these models misclassify exposed rock areas as exposed soil areas, leading to a significant discrepancy between the model's output erosion amount and the actual observed values—specifically, an overestimation of soil erosion. To address this issue, some studies have attempted to incorporate rock exposure rate into the RUSLE model, for example, by introducing (1-α) (where α is the rock exposure rate). The initial introduction of (1-α) can, to some extent, compensate for the applicability issues of traditional models in karst regions. However, the aforementioned correction methods assume a simple linear relationship between rock exposure rate and soil erosion. In reality, the relationship between soil erosion and exposed bedrock in karst regions is highly complex, not a simple linear one, but typically exhibits a complex nonlinear response. Therefore, existing RULSE models, even after similar (1-α) revisions, still suffer from insufficient accuracy. Furthermore, simply introducing bedrock exposure rate to correct the RULSE model does not conform to fundamental scientific and rigorous principles. Therefore, there is an urgent need to develop a method for estimating soil loss in karst regions through long-term scientific field observations, taking into account the unique geological environment of this area. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention proposes a method for estimating soil loss in karst regions, thereby improving the accuracy and practicality of model estimation and prediction.

[0004] To achieve the above objectives, the present invention provides a method for estimating soil loss in karst regions, comprising: Based on in-situ field observations, a negative exponential function relationship between rock exposure rate and soil loss was determined. Based on the negative exponential function relationship, a rocky desertification factor is constructed to quantify the inhibitory effect of exposed rocks on soil erosion. The rocky desertification factor was introduced as a multiplicative factor into the RUSLE model to establish the RUSLE-D model; Data on each RUSLE factor and rock exposure rate in the target area are obtained and input into the RUSLE-D model to calculate the soil erosion in the target area.

[0005] Optionally, determining the negative exponential function relationship based on in-situ field observations includes: Multiple in-situ runoff plots with different rock exposure rates were set up in a typical karst watershed. Under natural rainfall conditions, the runoff plots are continuously monitored to collect rainfall, runoff, and sediment data; Based on collected rainfall, runoff, and sediment data, soil loss under different rock exposure rates was calculated, and a negative exponential function relationship between rock exposure rate and soil loss was obtained by fitting relative soil loss as an indicator.

[0006] Optionally, constructing rocky desertification factors includes: The relative soil loss in the negative exponential function relationship is converted into a rocky desertification factor; The rocky desertification factor is expressed as a negative exponential function of the rock exposure rate. Where D is the rocky desertification factor, a is the coefficient, b is the index reflecting the degree of influence of rock exposure rate on soil erosion, and E is the rock exposure rate.

[0007] Optionally, the constructed RUSLE-D model is as follows: ; in, The model is RUSLE-D; It is the erosivity factor of rainfall; The soil erosion coefficient; For slope length and steepness coefficient; Factors related to vegetation cover and management practices; For soil and water conservation measurement factors; It is a factor contributing to rocky desertification.

[0008] Optionally, obtaining the RUSLE factor data for the target region includes: Raster data for calculating rainfall erosivity factors based on meteorological data and spatial interpolation; raster data for calculating soil erodibility factors based on soil texture data; raster data for calculating slope and slope length factors based on digital elevation model (DEM) data; and raster data for determining vegetation cover and management factors and soil and water conservation measures factors based on remote sensing imagery or land use data.

[0009] Optionally, obtaining rock exposure rate data for the target area includes: Obtain the bedrock exposure rate vector data of the target area and convert the bedrock exposure rate vector data into raster data; normalize the raster data so that the value of the rock exposure rate E is between 0 and 1.

[0010] Optionally, calculating the soil erosion in the target area includes: The raster data of each RUSLE factor and the raster data of the rocky desertification factor calculated from the rock exposure rate data are spatially registered in the GIS platform; the raster calculator is used to perform the multiplication operation of the RUSLE-D model on each cell to obtain the spatial distribution results representing the soil erosion modulus.

[0011] Optionally, the method further includes: combining the soil loss calculated based on the RUSLE-D model with the sediment transport ratio (SDR) model to estimate the sediment yield at the watershed outlet.

[0012] Technical effects of the invention: 1. Fundamentally Improved Estimation Accuracy: This invention overcomes the major flaw of existing general soil loss models (RUSLE) that systematically overestimate soil erosion when applied to karst regions. By scientifically quantifying the impact of exposed bedrock on soil erosion through rocky desertification factors, the output results of the modified RUSLE-D model are in high agreement with field observation data, significantly reducing the deviation between simulated and measured values, and greatly improving the accuracy and reliability of soil loss estimation under this special geological background.

[0013] 2. Scientific Characterization of Nonlinear Relationships, Enhancing the Theoretical Foundation of the Model: Compared to the simple and linear methods of correcting rock exposure rates in existing technologies, this invention, based on long-term field observations, constructs a mathematical model that accurately reflects the complex nonlinear relationship between soil erosion and bedrock exposure rates in karst regions. This core improvement makes the model's characterization of natural laws more scientific and rigorous, moving from empirical corrections to mechanistic simulations, laying a solid scientific foundation for the model's application in karst regions, and avoiding the subjectivity and inaccuracy of previous correction methods.

[0014] 3. Optimizing Regional Assessment Capabilities and Enhancing the Practical Value of Results: While inheriting the spatial analysis advantages of the RUSLE model, this method integrates rocky desertification factors to achieve a more refined depiction of the spatial differentiation patterns of soil erosion in karst regions. This method is easily implemented in a Geographic Information System (GIS) platform and can efficiently generate high-precision regional soil erosion distribution maps. The results can directly serve water and soil conservation planning, rocky desertification control effectiveness assessment, and ecological security pattern construction in karst regions, providing more reliable technical support for relevant decision-making and possessing extremely high value for widespread application. Attached Figure Description

[0015] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of five runoff plots with different rock exposure rates according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the relationship between suspended sediment concentration and rock exposure rate in an embodiment of the present invention. Figure 3 This is a schematic diagram comparing the sediment yield simulated by the RUSLE model in an embodiment of the present invention with the sediment yield observed in the SBT catchment area. Figure 4 This is a schematic diagram comparing the sediment yield simulated by the RUSLE-D model in an embodiment of the present invention with the sediment yield observed in the SBT catchment area. Figure 5 This is a schematic diagram comparing the sediment yield simulated by the RUSLE model in an embodiment of the present invention with the sediment yield observed in the GC catchment area. Figure 6 This is a schematic diagram comparing the sediment yield simulated by the RUSLE-D model in an embodiment of the present invention with the sediment yield observed in the GC catchment area. Figure 7 This is a flowchart illustrating a method for estimating soil loss in karst regions according to an embodiment of the present invention. Detailed Implementation

[0016] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0017] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0018] like Figure 7 As shown, this embodiment provides a method for estimating soil loss in karst areas, including: determining the negative exponential function relationship between rock exposure rate and soil loss based on in-situ field observations; Based on the negative exponential function relationship, a rocky desertification factor is constructed to quantify the inhibitory effect of exposed rocks on soil erosion. The rocky desertification factor was introduced as a multiplicative factor into the RUSLE model to establish the RUSLE-D model; Data on each RUSLE factor and rock exposure rate in the target area are obtained and input into the RUSLE-D model to calculate the soil erosion in the target area.

[0019] Furthermore, determining the negative exponential function relationship based on in-situ field observations includes: Multiple in-situ runoff plots with different rock exposure rates were set up in a typical karst watershed. Under natural rainfall conditions, the runoff plots are continuously monitored to collect rainfall, runoff, and sediment data; Based on collected rainfall, runoff, and sediment data, soil loss under different rock exposure rates was calculated, and a negative exponential function relationship between rock exposure rate and soil loss was obtained by fitting relative soil loss as an indicator.

[0020] Specifically, one implementation of this embodiment is as follows: like Figure 1 The study investigates the impact of exposed rocks on soil erosion and fits the relationship between rock exposure rate and soil erosion: (1) Five in-situ runoff plots were established in a typical karst watershed. Given that the length of continuous slopes in karst areas is generally no more than 15m, the runoff plots are 12m long and 3m wide. All plots face south and have a slope of 15°.

[0021] (2) In order to minimize the impact of stone size, limestone with a diameter of 30-40 cm and a height of 40-60 cm is selected. Natural limestone is transplanted and planted to simulate rocky desertification slopes, buried 30 cm or more below the surface soil in contact with the bedrock.

[0022] (3) Set the rock exposure rate to 0%, 10%, 30%, 50% and 70%. Use drone photography and Photoshop to determine the rock exposure rate.

[0023] (4) A small weather station was set up near the site to monitor rainfall. Field observations began four months after natural settlement.

[0024] (5) Rainfall, runoff, and sediment data were collected after each rainfall event: Rainfall data were collected using rain gauges, while runoff and sediment data were measured manually. Runoff depth was measured directly using a water level gauge in a runoff collector, and runoff volume was calculated. Sediments deposited in each runoff pool were thoroughly stirred, and turbidity samples were collected and brought back to the laboratory. Samples were filtered using qualitative filter paper, and the filtered sediment was then placed in a drying oven at 105°C for 12 hours. Finally, the sediment yield for each rainfall event was determined based on the ratio of sediment weight to runoff volume in the sample. During the observation period, weeds were manually removed weekly to eliminate vegetation impact. A total of 32 erosive rainfall datasets were collected over the two rainy seasons.

[0025] like Figure 2As shown, the relationship between rock exposure rate and soil loss is as follows: suspended sediment concentration decreases with increasing rock exposure rate. Specifically, the runoff plot with a rock exposure rate of 0% has the highest suspended sediment concentration, followed by runoff plots with rock exposure rates of 10%, 30%, 50%, and 70%.

[0026] Furthermore, the construction of rocky desertification factors includes: The relative soil loss in the negative exponential function relationship is converted into a rocky desertification factor; The rocky desertification factor is expressed as a negative exponential function of the rock exposure rate. Where D is the rocky desertification factor, a is the coefficient, b is the index reflecting the degree of influence of rock exposure rate on soil erosion, and E is the rock exposure rate.

[0027] Specifically, one implementation of this embodiment is as follows: To quantify the impact of rock exposure rate on soil erosion, a rocky desertification factor (D) was constructed, using relative soil loss as the basis for quantitative analysis. Relative soil loss is the ratio of soil loss in areas with exposed rock to soil loss in areas without exposed rock (i.e., an exposure rate of 0%). ; Where A i This is a relative soil erosion measure, reflecting the relative impact of exposed rock on soil erosion; it is dimensionless. i S0 represents soil loss in g for rocky (0–70%) areas; S0 represents soil loss in areas without rock.

[0028] Establish a functional relationship between rock exposure rate (0–70%, as the independent variable) and relative soil loss (the dependent variable). Try the following functional models in turn: Linear model: ; Polynomial model: ; Exponential model: ; Logarithmic model: ; Power function model: ; The relationships between relative soil erosion and rock exposure were fitted using the above functions. The exponential function showed the highest R-value. 2 The value is 0.834, while the power function exhibits the lowest R. 2 The value is 0.609, as shown in Table 1.

[0029] Table 1

[0030] The desertification factor D is represented by a negative exponential function, as shown below: ; Where D is the rocky desertification factor, a is the coefficient, b is the index reflecting the degree of influence of rock exposure rate on soil erosion, and E is the rock exposure rate, with a value range of 0-100 (excluding %). According to the field observation results, a is 0.967 and b is 0.036.

[0031] Furthermore, the constructed RUSLE-D model is as follows: The rocky desertification factor (D) was introduced into the RUSLE model to establish the RUSLE-D model. The rocky desertification factor represents the influence of rock exposure rate on soil erosion. The constructed RUSLE-D model is as follows: ; In the formula, This represents the RUSLE-D model; R represents the rainfall erosivity factor (MJ⋅mm⋅ha). -1 ⋅h -1 ⋅yr -1 K represents the soil erosion coefficient (t⋅hm). -2 ⋅h⋅MJ -1 ⋅mm -1 ⋅hm -2 ); LS represents slope length and steepness coefficient (dimensionless); C represents vegetation cover and management measures factor (dimensionless); P represents soil and water conservation measurement factor (dimensionless); D represents rocky desertification factor, which is dimensionless.

[0032] Furthermore, obtaining the RUSLE factor data for the target region includes: Raster data for calculating rainfall erosivity factors based on meteorological data and spatial interpolation; raster data for calculating soil erodibility factors based on soil texture data; raster data for calculating slope and slope length factors based on digital elevation model (DEM) data; and raster data for determining vegetation cover and management factors and soil and water conservation measures factors based on remote sensing imagery or land use data.

[0033] Specifically, one implementation of this embodiment is as follows: Rainfall erosivity factor (R): Data Acquisition: Obtain multi-year daily precipitation data (such as daily precipitation or hourly precipitation) from various meteorological stations in the study area from the National Meteorological Data Platform.

[0034] Rainfall erosivity calculation: Missing data are removed and time series integrity is checked to ensure calculation accuracy. The processed precipitation data is then substituted into the rainfall erosivity calculation formula: ; Among them, R jRepresents MJ⋅mm⋅ha -1 ⋅h -1 Monthly rainfall erosivity; j ranges from 1 to 12, P d This indicates that the daily erosion rainfall is ≥12mm.

[0035] Spatialization: Since the calculated R-values ​​are based on meteorological station data, spatial interpolation (such as inverse distance weighting (IDW) or kriging interpolation) needs to be performed on the GIS platform to obtain continuously distributed R-factor raster data for the study area. The final generated R-factor raster is a dimensionless spatial layer that can be directly used as input data for the RUSLE-D model.

[0036] Soil erodibility factor (K): The sand content, silt content, clay content, and organic carbon content of each soil type need to be obtained and then calculated using the following formula.

[0037] ; In the formula, SAN represents sand content (0.05–2.00 mm), %; SIL represents mud content (0.002–0.05 mm), %; CLA represents clay content (<0.002 mm), %; SN1 = 1–SAN / 100; C o Represents organic carbon content, %. The result value is multiplied by 0.1317 to convert to SI units, with the final unit being t⋅hm. 2 ⋅h⋅MJ -1 ⋅mm -1 ⋅hm -2 .

[0038] Slope Length and Steepness Coefficient (LS): DEM data of the study area was acquired at a resolution consistent with other model factors. The DEM was filled with depressions to ensure the continuity of the flow path. The confluence length of each cell was obtained using flow direction and cumulative discharge analysis. The LS factor was calculated using the spatial analysis tool – Raster Calculator, with the following formula: ; ; ; ; In the formula, L represents the slope length factor, which is dimensionless; s represents the slope steepness factor, which is also dimensionless; θ is the slope value extracted from the DEM data; n and m are the slope index and the ratio of gully erosion to intergully erosion, respectively. The above calculations are performed in a GIS platform, and the results generate continuously distributed LS factor raster data, which serves as the input to the RUSLE-D model.

[0039] Vegetation cover and management measures factor (C) and soil and water conservation measures factor (P): The vegetation cover and management measures factor (C) is an important indicator for evaluating the ability of vegetation to resist soil erosion, ranging from 0 to 1. The soil and water conservation measures factor (P) reflects the inhibitory effect of soil and water conservation measures on soil erosion, ranging from 0 to 1, and is calculated using an evaluation method.

[0040] Data source: Remote sensing images or land use data of the study area are obtained. After importing the land use data, the vegetation cover is retrieved using the remote sensing vegetation index (NDVI), or the vegetation cover level is determined according to the land use type.

[0041] Spatialization processing: Various land use units or pixels are assigned corresponding C / P values ​​to obtain continuously distributed C / P factor raster as shown in Table 2.

[0042] Table 2

[0043] Furthermore, obtaining data on the rock exposure rate in the target area includes: Obtain the bedrock exposure rate vector data of the target area and convert the bedrock exposure rate vector data into raster data; normalize the raster data so that the value of the rock exposure rate E is between 0 and 1.

[0044] Specifically, one implementation of this embodiment is as follows: The calculation process for the rocky desertification factor is as follows: 1) Bedrock exposure rate data acquisition and rasterization: The bedrock exposure rate vector layer in the original rocky desertification monitoring data is converted into raster data, and the exposure rate raster data is generated with the "bedrock exposure rate" field as the value field.

[0045] 2) Missing value handling: Missing values ​​(null cells) in raster data are uniformly assigned a value of 0 to ensure the integrity of subsequent calculations.

[0046] 3) Mask extraction of the study area: The vector boundary of the study area is used as a mask to extract the bedrock exposure rate raster corresponding to the area from the rocky desertification data.

[0047] 4) Normalization and Zero Value Correction: The bedrock exposure rate is normalized to unify its value range within the [0,1] interval. For zero values ​​appearing in the normalization results, smaller values ​​are used for substitution (e.g., 0 is replaced with 0.005) to avoid numerical anomalies caused by zero values ​​in subsequent model calculations. The processed data is then used to obtain the rocky desertification raster data.

[0048] 5) Results: The final data is a dimensionless raster, representing the bedrock exposure rate corresponding to each pixel. After normalization and numerical correction, it can be directly used as the spatial input data for the rocky desertification factor (D).

[0049] 6) Calculation of the rocky desertification factor (D): The rocky desertification factor is calculated using processed rock exposure rate data. The exponential relationship is entered into the raster calculator. E represents raster data of rock exposure rate.

[0050] Furthermore, calculating the soil erosion in the target area includes: The raster data of each RUSLE factor and the raster data of the rocky desertification factor calculated from the rock exposure rate data are spatially registered in the GIS platform; the raster calculator is used to perform the multiplication operation of the RUSLE-D model on each cell to obtain the spatial distribution results representing the soil erosion modulus.

[0051] Specifically, one implementation of this embodiment is as follows: Soil erosion simulation based on the RUSLE-D model: Factor rasterization: The calculation results of each factor (R, K, LS, C, P) of RUSLE are converted into raster data with a unified resolution and coordinate system. The same spatial resolution (e.g., 30m) and projected coordinates are used to ensure that pixel-level operations can be performed between factors. The rocky desertification factor raster constructed in step 3 is used as the input of the rocky desertification factor (D).

[0052] Spatial calculation process: In the GIS platform, the raster calculator is used to perform formula calculations pixel by pixel: A = R × K × LS × C × P × D; Spatial distribution simulation of soil erosion: The calculated result, raster A, represents the annual soil erosion modulus (MJ⋅mm⋅ha) of each pixel in the study area. -1 ⋅h -1 ⋅yr -1 ).

[0053] Results: The simulation results provide an estimate of regional soil loss, intuitively revealing the spatial variability of soil loss in karst regions. This method integrates the RUSLE model with rocky desertification factors in a GIS environment, overcoming the limitation of traditional RUSLE models that overestimate soil erosion in karst regions due to the presence of bare rock.

[0054] Model validation includes the following: Selection of validation areas: The SBT and GC hydrological stations in typical karst watersheds were selected as research objects. Measured runoff and sediment yield data of the two watersheds from 2010 to 2020 were obtained, and a comparative validation scenario was constructed by combining the regional DEM and rocky desertification distribution map.

[0055] Model operation: The annual soil and water loss in the two watersheds was calculated using the original RUSLE model and the improved RUSLE-D model, respectively.

[0056] The formula for calculating the RUSLE model is: A = R × K × LS × C × P; The RUSLE-D model then introduces a rocky desertification factor D into the above formula: A = R × K × LS × C × P × D.

[0057] Furthermore, the method also includes: combining the soil loss calculated based on the RUSLE-D model with the sediment transport ratio (SDR) model to estimate the sediment yield at the watershed outlet.

[0058] Specifically, one implementation of this embodiment is as follows: By combining the calculated soil loss with the sediment transport ratio (SDR) model, the annual sediment yield at the watershed outlet is estimated: ; in, It is the sediment yield (t) at the outlet cross section of the basin. It is the sediment transport rate (dimensionless). It is the amount of soil and water loss (t) in the upstream watershed of the control section.

[0059] An improved SDR model for karst areas was used for estimation: a pixel-level weighting factor AW was constructed using parameters such as rainfall, slope, vegetation cover, and rock exposure rate. i And based on the connectivity index KIC f Calculate the SDR value of the spatial distribution; ; Among them: AW i R is the integral weighting factor for the i-th pixel (dimensionless). t It is the normalized rainfall erosion coefficient for period t; E i C is the rock exposure rate of the i-th pixel; t S represents the vegetation management factor in period t; i It is the slope (m / m) of the i-th pixel; ; In the formula, KIC f It is the connectivity index (dimensionless); D up This represents the downhill component, while D... dn Indicates the downhill component; It is the average normalized rainfall erosivity over period t; It is the average rock exposure rate; It is the average vegetation management coefficient within period t; This is the average slope (m / m). A is the drainage area on the upslope (m²). 2 ). d i It is the length (m) of the i-th unit along the downhill path. The subscript m indicates that each pixel i has its own KIC. f The value, n, represents the total number of pixels. To avoid calculation errors, an AW value less than 0.005 is used. i The value will be adjusted to 0.005. KIC f The value ranges from –∞ to ∞.

[0060] Finally, a pixel-by-pixel SDR raster is formed within the watershed, and it is overlaid with the soil loss raster to obtain the total sediment yield of the watershed. ; in: This represents the SDR of the i-th pixel. It is the theoretical maximum SDR, with a value of 1. and These are calibration parameters, set to 0 and 1 respectively.

[0061] Model performance evaluation: Using the coefficient of determination R 2 The simulation results were validated using the Nash efficiency coefficient (NSE), percentage bias (PBIAS), and model comparison index (MCI).

[0062] The evaluation criteria are: R 2 A score >0.5, NSE >0.5, and |PBIAS| <25% indicate that the model has acceptable accuracy; MCI is used to compare the performance of RUSLE and RUSLE-D models.

[0063] ; ; ; ; Where n represents the number of data samples; O i and S i These are the observed and simulated sediment values ​​for the watershed in year i, respectively; O and S are the average values ​​of the observed and simulated data, respectively; S ri and S oi These are the values ​​simulated by the RUSLE-D model and the original RUSLE model, respectively.

[0064] In the SBT watershed, the simulation results of the RUSLE-D model show high accuracy in fitting the measured sediment yield, as evidenced by R... 2=0.78, NSE=0.55, PBIAS=13.87%, all superior to the original RUSLE model. In the GC watershed, the RUSLE-D model also significantly improved the simulation accuracy of soil loss and sediment yield. The results show that the RUSLE-D model, which incorporates the rocky desertification factor D, can effectively correct the overestimation of erosion caused by the widespread presence of bare rock in karst areas, and has stronger applicability and accuracy. A comparison of observed and simulated sediment yields in the SBT and GC watersheds is shown below. Figures 4-6 As shown, where Figure 3 This is a schematic diagram comparing the sediment yield simulated by the RUSLE model in an embodiment of the present invention with the sediment yield observed in the SBT catchment area. Figure 4 This is a schematic diagram comparing the sediment yield simulated by the RUSLE-D model in an embodiment of the present invention with the sediment yield observed in the SBT catchment area. Figure 5 This is a schematic diagram comparing the sediment yield simulated by the RUSLE model in an embodiment of the present invention with the sediment yield observed in the GC catchment area. Figure 6 This is a schematic diagram comparing the sediment yield simulated by the RUSLE-D model in an embodiment of the present invention with the sediment yield observed in the GC catchment area.

[0065] This embodiment completely overcomes the major flaw of existing general soil loss models (RUSLE) that systematically overestimate soil erosion when applied to karst regions. By scientifically quantifying the inhibitory effect of exposed bedrock on soil erosion through rocky desertification factors, the output results of the modified RUSLE-D model are in high agreement with field observation data, significantly reducing the deviation between simulated and measured values, and greatly improving the accuracy and reliability of soil loss estimation under this special geological background.

[0066] Compared to the simple, linear methods for correcting bedrock exposure rates in existing technologies, this invention, based on long-term field observations, reveals for the first time and successfully constructs a mathematical model that accurately reflects the complex nonlinear relationship between soil erosion and bedrock exposure rates in karst regions. This core improvement makes the model's representation of natural laws more scientific and rigorous, moving from empirical correction to mechanistic simulation, laying a solid scientific foundation for the model's application in karst regions, and avoiding the subjectivity and inaccuracy of previous correction methods.

[0067] This embodiment, while inheriting the spatial analysis advantages of the RUSLE model, achieves a more refined characterization of the spatial differentiation patterns of soil erosion in karst regions by integrating rocky desertification factors. This method is easily implemented in a Geographic Information System (GIS) platform, capable of efficiently generating high-precision regional soil erosion distribution maps. Its results can directly serve water and soil conservation planning, rocky desertification control effectiveness assessment, and ecological security pattern construction in karst regions, providing more reliable technical support for relevant decision-making and possessing extremely high value for widespread application.

[0068] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for estimating soil loss in karst regions, characterized in that, include: Based on in-situ field observations, a negative exponential function relationship between rock exposure rate and soil loss was determined. Based on the negative exponential function relationship, a rocky desertification factor is constructed to quantify the inhibitory effect of exposed rocks on soil erosion. The rocky desertification factor was introduced as a multiplicative factor into the RUSLE model to establish the RUSLE-D model; Data on each RUSLE factor and rock exposure rate in the target area are obtained and input into the RUSLE-D model to calculate the soil erosion in the target area.

2. The method for estimating soil loss in karst areas as described in claim 1, characterized in that, Determining the relationship of the negative exponential function based on in-situ field observations includes: Multiple in-situ runoff plots with different rock exposure rates were set up in a typical karst watershed. Under natural rainfall conditions, the runoff plots are continuously monitored to collect rainfall, runoff, and sediment data; Based on collected rainfall, runoff, and sediment data, soil loss under different rock exposure rates was calculated, and a negative exponential function relationship between rock exposure rate and soil loss was obtained by fitting relative soil loss as an indicator.

3. The method for estimating soil loss in karst areas as described in claim 2, characterized in that, The factors that constitute rocky desertification include: The relative soil loss in the negative exponential function relationship is converted into a rocky desertification factor; The rocky desertification factor is expressed as a negative exponential function of the rock exposure rate. Where D is the rocky desertification factor, a is the coefficient, b is the index reflecting the degree of influence of rock exposure rate on soil erosion, and E is the rock exposure rate.

4. The method for estimating soil loss in karst areas as described in claim 1, characterized in that, The constructed RUSLE-D model is as follows: ； in, It is the RUSLE-D model; It is the erosivity factor of rainfall; The soil erosion coefficient; For slope length and steepness coefficient; Factors related to vegetation cover and management practices; For soil and water conservation measurement factors; It is a factor contributing to rocky desertification.

5. The method for estimating soil loss in karst areas as described in claim 1, characterized in that, Obtaining the RUSLE factor data for the target region includes: Raster data for calculating rainfall erosivity factors based on meteorological data and spatial interpolation; raster data for calculating soil erodibility factors based on soil texture data; raster data for calculating slope and slope length factors based on digital elevation model (DEM) data; and raster data for determining vegetation cover and management factors and soil and water conservation measures factors based on remote sensing imagery or land use data.

6. The method for estimating soil loss in karst areas as described in claim 1, characterized in that, Obtaining rock exposure rate data for the target area includes: Obtain the bedrock exposure rate vector data of the target area and convert the bedrock exposure rate vector data into raster data; normalize the raster data so that the value of the rock exposure rate E is between 0 and 1.

7. The method for estimating soil loss in karst areas as described in claim 6, characterized in that, Calculating the soil erosion in the target area includes: The raster data of each RUSLE factor and the raster data of the rocky desertification factor calculated from the rock exposure rate data are spatially registered in the GIS platform; the raster calculator is used to perform the multiplication operation of the RUSLE-D model on each cell to obtain the spatial distribution results representing the soil erosion modulus.

8. The method for estimating soil loss in karst areas as described in claim 1, characterized in that, The method further includes: combining the soil loss calculated based on the RUSLE-D model with the sediment transport ratio (SDR) model to estimate the sediment yield at the watershed outlet.

Images