Soil erosion assessment and prediction method based on public data set

Abstract

The invention provides a soil erosion evaluation and prediction method based on a public data set, and belongs to the field of soil erosion. An improved RUSLE model is constructed by obtaining public land utilization data, rainfall data, vegetation coverage data, soil texture data and DEM data of a target area, and the soil erosion condition of the target area is evaluated; four prediction scenes including a natural development scene, a city extension suppression scene, an ecological priority scene and a cultivated land protection scene are put forward according to policies, adaptive atlas under the four scenes are made, existing land utilization data and the adaptive atlas of all the scenes are input into a CA-Markov model, and land utilization conditions under all the scenes in the future are obtained; and on the basis of the predicted land utilization data, a water and soil conservation measure factor under each scene is obtained in combination with an assignment method, and then an improved RUSLE model is constructed in the grid, so that the soil erosion condition of the target area under multiple scenes in the future can be obtained.

Description

Technical Field

[0001] The present invention belongs to the field of soil erosion, and in particular relates to a soil erosion assessment and prediction method based on a public data set. Background Art

[0002] Soil is the most basic environmental element on earth and the basic resource for people to survive. However, in the process of development, human beings will inevitably cause damage to the soil. In addition to the influence of nature itself, soil erosion problems will occur in some areas. In severe cases, it will cause a series of ecological and environmental problems such as soil nutrient loss, land barrenness, floods and waterlogging, thereby reducing the quality of the soil and causing great negative impacts on the ecological environment, human life and social and economic development. The soil erosion problem has been identified by the United Nations as one of the major challenges facing the world in the 21st century, and it needs to be solved by the world together. At present, the method for estimating soil erosion mainly adopts the Revised Universal Soil Loss Equation (RUSLE). This model takes into account many factors that affect soil erosion and has a wide range of applicability. Therefore, this model is used to estimate soil erosion. Soil is the key to human survival. Therefore, accurately predicting the development trend of soil erosion is the key to controlling soil erosion and formulating macro-soil and water conservation measures. Related research has become a hot topic in soil research. At present, there are two main methods in the research on soil erosion prediction. One is to directly predict based on the existing soil erosion situation, and the other is to predict various soil erosion factors and then input them into the soil erosion model to achieve the purpose of predicting soil erosion. The first method ignores the impact of natural and human factors on soil erosion during the prediction process, and the prediction results are relatively rough; while the second method has been widely used, and its rationality and accuracy have been effectively verified.

[0003] At present, the prediction of soil erosion is mainly for a single city or region, and less consideration is given to urban agglomerations and economic zones such as the Tianshan North Slope Urban Agglomeration. As the main economic carrier of Xinjiang Uygur Autonomous Region, the Tianshan North Slope Urban Agglomeration accounts for 12.3% of Xinjiang's area, with a population of 11.72 million, accounting for more than 45.3% of Xinjiang's total population, and GDP accounts for more than 57.11% of Xinjiang's total population. Affected by seasonal precipitation, soil erosion is relatively serious. The spatial and temporal distribution of vegetation in the Tianshan North Slope Urban Agglomeration is quite different, with high vegetation coverage in mountainous areas and low vegetation coverage in plains. In addition to natural factors, human factors (land use, nomadic activities, and tourism) have become an important factor that cannot be ignored in aggravating soil erosion in the Tianshan North Slope Urban Agglomeration. Land use affects runoff and sediment transport processes by changing the surface morphology, thereby affecting soil erosion. Soil erosion and land use affect each other. For the Tianshan North Slope Urban Agglomeration, economic activities are relatively active, the proportion of building area is relatively large, and natural factors are relatively complex and diverse.

[0004] The actual measurement of soil erosion data requires a lot of time, manpower and material resources, and is difficult to obtain. The future soil erosion conditions are even more impossible to measure, and there is no reasonable basis for the formulation of soil and water conservation work plans. Therefore, this patent proposes a method framework that can simply and efficiently evaluate and predict the soil erosion conditions in the study area.

[0005] At present, the prediction of soil erosion is mostly focused on a single scenario, and there is little comprehensive consideration of multiple scenarios. In addition, the multi-scenario prediction method of soil erosion has certain limitations in predicting erosion factors. When predicting land use, the possibility of social development is not fully considered, the applicability factors of mutual conversion between various land types are not considered; nor is the impact of existing government policies on land use changes considered. Therefore, this patent innovatively integrates urban development, national policies, and the internal mechanism of land use conversion, and proposes four prediction scenarios: natural development scenario, urban expansion suppression scenario, ecological priority scenario, and farmland protection scenario. Summary of the invention

[0006] In view of the shortcomings of existing technologies, a soil erosion assessment and prediction method based on public data sets is proposed. By using easily accessible public data sets, the soil erosion conditions in the study area can be simply and efficiently assessed and predicted, providing a reference for the government to formulate soil and water conservation plans.

[0007] To achieve the above technical objectives, the present invention discloses a soil erosion assessment and prediction method based on a public data set, which is characterized by the following steps:

[0008] Step 1: Combined with national and local government policies and the influencing factors of soil erosion, the study area was set into different development scenarios, including: natural development scenario, urban expansion suppression scenario, ecological priority scenario, and farmland protection scenario;

[0009] Step 2: Preprocess the long-term land use type data and the geographic scenario constraint data corresponding to each scenario. The geographic scenario constraint data includes road, water system, slope, and elevation information. Through preprocessing, the land use data is reclassified into 7 primary categories, including: cultivated land, forest land, grassland, water body, ice / snow, bare land, and building land; unify the scope, spatial resolution, and coordinate system of the study area, and mark the slope data of the study area;

[0010] Step 3: Soil erosion assessment based on the improved RUSLE model, divide the study area into 1km grids, build an improved RUSLE model for soil erosion in each grid unit, obtain the soil erosion amount of a single grid, and summarize the results to assess the soil erosion status of the study area. Use resampling to divide the urban agglomeration-level study area into multiple 1km grids, build an improved RUSLE model in each grid unit, calculate the soil erosion amount of a single grid, and summarize the results to assess the soil erosion status of the study area;

[0011] Step 4: Based on the existing policies, the possible future development direction of the research area was simulated and four development models were set, namely natural development scenario, urban expansion suppression scenario, ecological priority scenario, and farmland protection scenario; restrictive conditions were set for the four development scenarios to make the simulation process as consistent as possible with future development, and then predict the future land use status data.

[0012] Step 5: Predict the land use adaptability atlas of the study area in a certain year by scenario. The land use adaptability atlas includes constraints and adaptability factors. The constraints are of Boolean type and are set between land types. 0 indicates that the conversion is not allowed and 1 indicates that the conversion is allowed. The adaptability factor is used to restrict the detailed rules of the use of the surrounding environment of the land.

[0013] Step 6: Construct the CA-markov model, which is a coupling model of Cellular Automata (CA) and Markov model. The Markov model is used to analyze the land use type data and land use adaptability atlas after data preprocessing to obtain the transition probability matrix under the corresponding state; the CA model is used to process the land use transformation on this basis;

[0014] Step 7: Combine land use data and adaptability atlas to predict the land use pattern at the urban agglomeration level using the CA-Markov model. Use the predicted multi-scenario land use data and soil and water conservation measures to obtain soil and water conservation measures factors for future time series.

[0015] Step 8: Input the future land use conditions into the RUSLE model improved for the public data set to obtain the soil erosion status assessment under multiple scenarios in the future: calculate the soil and water conservation measures factor by assigning values ​​to the land use data, and calculate other influencing factors in the soil erosion model RUSLE, and finally calculate the predicted soil erosion amount for each scenario in the target year.

[0016] Furthermore, several possible future development models are set to see the differences in the development of soil erosion under various models, providing reference for the formulation of soil and water conservation measures:

[0017] The CA-Markov model sets the conversion relationship and conversion priority between different land types according to the preset development scenarios, and predicts the land use type in the target year in combination with the restriction data.

[0018] CA Model:

[0019] S (t+1) =f(S t ,N)

[0020] S (t+1) , S t are the cell states at time t+1 and time t respectively, N is the neighborhood of the cell, and f is the conversion rule.

[0021] Markov Model:

[0022] S t+1 =P ij ×S t

[0023] S t , S t+1 are the states of ecosystem land use at time t and time t+1, P ji is a state transfer matrix.

[0024] Furthermore, according to the current urbanization development model and the existing land use conversion rate, a natural development scenario is set without restricting the conversion between land types in the conversion rules; a scenario of suppressing urban expansion: controlling the expansion of building land, allowing the conversion of building land to other land uses, but not allowing the conversion of other land to building land; an ecological priority scenario: taking ecological protection as the principle, practicing the concept of protecting the natural environment, and prohibiting the conversion of four types of land with high ecological benefits, namely forest land, grassland, water area, and ice / snow, to other land uses; a scenario of protecting cultivated land: prohibiting the conversion of cultivated land to other land uses, while allowing other land types to be transferred to cultivated land and converted to each other.

[0025] Furthermore, based on the multi-scenario prediction of land use change, other influencing factors in the improved RUSLE model refer to the slope and slope length DEM data, vegetation cover height data, soil texture data, precipitation data, and land use data of the calculation base year. The calculation formula is as follows:

[0026] DEM data:

[0027] LS=H max -H min

[0028] Among them, LS is the slope length factor, H max is the highest elevation value, H min is the minimum value of elevation;

[0029] Vegetation cover height data:

[0030]

[0031] Among them, C is the vegetation coverage factor, c is the vegetation coverage;

[0032] Soil texture data:

[0033]

[0034] Among them, S AN , S IL , C LA , C represent the contents of sand, silt, clay and organic matter, respectively; S N1 is the calculation parameter;

[0035] Precipitation data:

[0036]

[0037] Among them, P i is the rainfall in the ith year (mm), and the model parameters are α=0.0534, β=1.6548.

[0038] Land use data: After the land use data are assigned, the soil and water conservation measures factor is formed. Specifically, the cultivated land is assigned a value of 0.5, the forest land and grassland are assigned a value of 0.9, the water body, ice / snow, and construction land are assigned a value of 0, the bare land is assigned a value of 1, and the vegetation cover factor is C.

[0039] Beneficial effects: This method makes the soil erosion situation easier and more efficient to understand; it also takes into account the factors of conversion between different land types, such as road factors, water system factors, slope factors, elevation factors, etc., and uses the CA-Markov model to predict the conversion of land use under various scenarios, making the prediction results more in line with the actual situation in the future and improving the accuracy of the prediction. It is of great significance for the formulation of soil and water conservation measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the improved RUSLE model in the present invention;

[0041] Figure 2 It is a schematic diagram of the soil erosion assessment and prediction method based on a public data set of the present invention. DETAILED DESCRIPTION

[0042] The specific embodiments of the present invention are further described below in conjunction with the accompanying drawings:

[0043] like Figure 1 and Figure 2 A soil erosion assessment and prediction method based on a public data set is shown. The specific technical solution is as follows:

[0044] Soil erosion factor calculation based on public data sets, including slope and length factor, vegetation cover factor, rainfall erosion factor, soil erodibility factor, and soil and water conservation measures factor. The economic development, natural conditions, and location factors of the research area are comprehensively studied, and different soil erosion factors are calculated using the soil erosion factor calculation method.

[0045] Calculate each soil erosion factor. The calculation method of each factor within the framework is as follows:

[0046] Slope length factor:

[0047] LS=H max -H min

[0048] Among them, LS is the slope length factor, H max is the highest value of elevation, H min The lowest elevation value.

[0049] Vegetation cover factor:

[0050]

[0051] Among them, C is the vegetation coverage factor and c is the vegetation coverage.

[0052] Precipitation erosion factor:

[0053]

[0054] Among them, P i is the rainfall in the ith year (mm), and the model parameters are α=0.0534, β=1.6548.

[0055] Soil erodibility factor:

[0056]

[0057] Among them, S AN , S IL , C LA , C represent the contents of sand, silt, clay and organic matter, respectively; S N1 is the calculation parameter.

[0058] Soil and water conservation measures factors:

[0059] The soil and water conservation measures factor is assigned from land use data. Specifically, cultivated land is assigned a value of 0.5, forest land and grassland are assigned a value of 0.9, water bodies, ice / snow, and construction land are assigned a value of 0, and bare land is assigned a value of 1.

[0060] Preprocess the long-term land use type data and the corresponding geographic scenario data (roads, water systems, slopes, elevations, etc.) of each scenario. Preprocessing refers to the unification of the coordinate system, spatial range, and resolution of the long-term land use type data and the corresponding geographic scenario restriction data (roads, water systems, slopes, elevations, etc.), and also includes the binarization of the restriction conditions (each land use type) (0 represents the background value and 1 represents the data).

[0061] Soil erosion assessment based on the improved RUSLE model. The study area is divided into 1km grids, and an improved RUSLE model is constructed for the soil erosion of each grid unit to obtain the soil erosion amount of a single grid. The results are summarized and used to assess the soil erosion status of the study area. The urban agglomeration-level study area is divided into multiple 1km grids by resampling. An improved RUSLE model is constructed within each grid unit to calculate the soil erosion amount of a single grid. The summary results are used to assess the soil erosion status of the study area.

[0062] Based on the internal mechanism of urban development, national policies and land use conversion, four forecast scenarios are proposed: natural development scenario, urban expansion suppression scenario, ecological priority scenario and farmland protection scenario. Four possible future development scenarios of urban agglomerations are formulated, namely natural development scenario, urban expansion suppression scenario, ecological priority scenario and farmland protection scenario.

[0063] According to the development model set in each scenario, the corresponding Boolean type adaptability atlas is produced. The adaptability atlas of each land use type under each scenario refers to the atlas used to specify the conversion rules of a certain land type, including constraints (Boolean type, set between land types, 0 means no conversion, 1 means conversion) and adaptability factors (referring to the selection of appropriate CA-Markov built-in fuzzy normalization functions for restricted data, that is, the impact of the restriction on this land use type).

[0064] The CA-Markov model is used to predict the land use pattern under different scenarios in the future. The CA-Markov model is used to predict the land use pattern at the urban agglomeration level by combining land use data and adaptability atlas. The CA-Markov model refers to a coupling model that combines cellular automata (CA) and the Markov model. The Markov model is used to analyze the data and obtain the transition probability matrix under the corresponding state; the CA model processes the land use transition on this basis.

[0065] CA Model:

[0066] S (t+1) =f(S t ,N)

[0067] S (t+1) , S t are the cell states at time t+1 and time t respectively, N is the neighborhood of the cell, and f is the conversion rule.

[0068] Markov Model:

[0069] S t+1 =P ji ×S t

[0070] S t , S t+1 are the states of ecosystem land use at time t and time t+1, P ji is a state transfer matrix;

[0071] Calculate the soil and water conservation measures factors for multiple scenarios in the future. Using the multi-scenario land use data predicted in step 6 and the soil and water conservation measures factor calculation method, the soil and water conservation measures factors for the future time series are obtained.

[0072] Evaluate the soil erosion status of multiple scenarios in the future. Input the predicted soil and water conservation measures factor and other factors (slope length factor, vegetation cover factor, rainfall erosion factor, soil erodibility factor) into the RUSLE model, divide it into 1km grids, and obtain the soil erosion amount of a single grid. The results are summarized to evaluate the future soil erosion status of the study area.

[0073] Calculation method of soil and water conservation measures factors: Use the assignment method to obtain soil and water conservation measures factors under multiple future scenarios at the urban agglomeration level.

[0074] In order to facilitate understanding of the implementation process and basic principles of the present invention, the Tianshan North Slope Urban Agglomeration is selected as an example object to specifically explain a soil erosion assessment and prediction method based on a public data set described in the present invention:

[0075] Collect data sets, the data sources are shown in the following table:

[0076]

[0077] S1. The collected public DEM data are used to calculate the slope and length of the slope in the study area; the public vegetation coverage data are combined with the vegetation coverage factor calculation method to obtain the vegetation coverage factor of the study area; the public annual precipitation data are combined with the precipitation erosion factor calculation method to obtain the precipitation erosion factor of the study area; the public soil texture data are combined with the soil erodibility factor calculation method to obtain the soil erodibility factor of the study area; based on the existing land use data, the soil and water conservation measures factor of the study area is obtained using the assignment method.

[0078] S2. Use GIS tools to unify the spatial range, coordinate system and resolution of the prepared land use data, road buffer data, railway buffer data, water system data, slope data and elevation data. The role of the road buffer and railway buffer is to restrict the conversion of land types in a certain range of areas around roads and railways. Binarize the restricted data and assign the restricted data a value of 1. The rest are 0.

[0079] S3. Divide the study area into 1km grids and construct an improved RUSLE model in each grid unit. Multiply the slope and slope length factors, vegetation coverage factors, rainfall erosivity factors, soil erodibility factors, and soil and water conservation measures factors in each grid to obtain the soil erosion amount of a single grid. The results are summarized to evaluate the soil erosion status of the study area.

[0080] S4. Based on the integration of urban development, national policies, and the internal mechanism of land use conversion, the following four scenarios for the possible future development of the study area at the city cluster level are proposed: the natural development scenario is a scenario setting that does not restrict the conversion between land types in the conversion rules based on the current urbanization development model and the existing land use conversion rate without considering the policy planning for future development; the urban expansion suppression scenario will control the expansion of building land, allowing the conversion of building land to other land uses, but not allowing the conversion of other land uses to building land; the ecological priority scenario is based on the principle of ecological protection, prohibiting the conversion of four types of land with high ecological benefits, namely forest land, grassland, water area, and ice / snow, to other land uses; the cultivated land protection scenario will stop the conversion of cultivated land to other land uses, while allowing other land types to be transferred to cultivated land and converted to each other.

[0081] The transfer matrix results of the average soil erosion of the urban agglomeration on the northern slope of Tianshan Mountains. The transfer matrix results of the average soil erosion (t km -2 ·a -1 )as follows:

[0082]

[0083] S5. Preparation of adaptability atlas: Prepare adaptability atlas for each land use type in each scenario. According to the scenario setting and the nature of the land type itself, select corresponding constraints and restrictions to prepare the adaptability atlas, and select the built-in fuzzy normalization function of the model for each restriction. Take the adaptability atlas of each land type in each scenario as a set, which is the adaptability atlas for that scenario.

[0084] S6. Use the Markov model to calculate the transfer area matrix between two periods of existing land use type data as an input data for prediction; the adaptability atlas of the scenario is also used as an input data; the time span of the two periods of data should be equal to the time span between the base year and the forecast year.

[0085] S7. Use the assignment method to obtain the soil and water conservation measures factors for multiple scenarios in the future. The value range of the soil and water conservation measures factor is between 0 and 1. The closer it is to 0, the better the effect of the soil and water conservation measures. The value of cultivated land is assigned to 0.5, forest land, shrubs, and grassland are all assigned to 0.9, bare land is 1, and water bodies, ice / snow, and construction land are assigned to 0.

[0086] S8. Evaluate the soil erosion status under multiple scenarios in the future. Consistent with the process of S1 and S3, replace the soil and water conservation measures factor with the soil and water conservation measures factor under multiple scenarios in the future obtained in S7 to obtain the soil erosion amount under multiple scenarios in the future, which is used to evaluate the soil erosion status under multiple scenarios in the future.

[0087] The results of the actual application of this method to the urban agglomeration on the northern slope of the Tianshan Mountains show that from the perspective of short-term development, the scenario of suppressing urban expansion has the best effect; from the perspective of long-term development, the average soil erosion reduction under the scenario of protecting cultivated land is almost the same as that under the scenario of suppressing urban expansion, but the scenario of suppressing urban expansion begins to show an upward trend. In summary, under the premise of restricting urban land expansion and protecting cultivated land, the soil erosion situation has improved significantly, and these two development models can be considered comprehensively. The results of soil erosion in each scenario obtained by applying this method to the urban agglomeration on the northern slope of the Tianshan Mountains (t km -2 ·a -1 ):

[0088]

[0089] The methods demonstrated above can predict the amount of soil erosion in the study area under various development scenarios in a relatively simple, efficient and accurate manner, providing certain theoretical support for the future development planning of the study area and the formulation of soil and water conservation measures.

Claims

1. A soil erosion assessment and prediction method based on a public data set, characterized in that: Here are the steps: Step 1: Combined with international and local government policies and the influencing factors of soil erosion, the study area was set into different development scenarios, including: natural development scenario, urban expansion suppression scenario, ecological priority scenario, and farmland protection scenario; Step 2: Preprocess the long-term land use type data and the geographic scenario constraint data corresponding to each scenario. The geographic scenario constraint data includes road, water system, slope, and elevation information. Through preprocessing, the land use data is reclassified into 7 primary categories, including: cultivated land, forest land, grassland, water body, ice / snow, bare land, and building land; unify the scope, spatial resolution, and coordinate system of the study area, and mark the slope data of the study area; Step 3: Based on the soil erosion assessment of the improved RUSLE model, the study area is divided into 1km grids, and the improved RUSLE model is constructed for the soil erosion of each grid unit to obtain the soil erosion amount of a single grid. The results are summarized to assess the soil erosion status of the study area; the urban agglomeration-level study area is divided into multiple 1km grids by resampling, and the improved RUSLE model is constructed in each grid unit to calculate the soil erosion amount of a single grid. The summary results are used to assess the soil erosion status of the study area; Step 4: Based on the existing policies, the possible future development direction of the research area was simulated and four development models were set, namely natural development scenario, urban expansion suppression scenario, ecological priority scenario, and farmland protection scenario; restrictive conditions were set for the four development scenarios to make the simulation process as consistent as possible with future development, and then predict the future land use status data. Step 5: Predict the land use adaptability atlas of the study area in a certain year by scenario. The land use adaptability atlas includes constraints and adaptability factors. The constraints are of Boolean type and are set between land types. 0 indicates that the conversion is not allowed and 1 indicates that the conversion is allowed. The adaptability factor is used to restrict the detailed rules of the use of the surrounding environment of the land. Step 6: Construct the CA-markov model, which is a coupling model of Cellular Automata (CA) and Markov model. The Markov model is used to analyze the land use type data and land use adaptability atlas after data preprocessing to obtain the transition probability matrix under the corresponding state; the CA model is used to process the land use transformation on this basis; Step 7: Combine land use data and adaptability atlas to predict the land use pattern at the urban agglomeration level using the CA-Markov model. Use the predicted multi-scenario land use data and soil and water conservation measures to obtain soil and water conservation measures factors for future time series. Step 8: Input the future land use conditions into the RUSLE model improved for the public data set to obtain the soil erosion status assessment under multiple scenarios in the future: calculate the soil and water conservation measures factor by assigning values ​​to the land use data, and calculate other influencing factors in the soil erosion model RUSLE, and finally calculate the predicted soil erosion amount for each scenario in the target year.

2. The soil erosion assessment and prediction method based on a public data set according to claim 1, characterized in that: Based on multi-scenario prediction of land use change, several possible future development models are set to see the differences in the development of soil erosion under various models, providing reference for the formulation of soil and water conservation measures: The CA-Markov model sets the conversion relationship and conversion priority between different land types according to the preset development scenarios, and predicts the land use type in the target year in combination with the restriction data; CA Model: S (t+1) =f(S t ,N) S (t+1) , S t are the states of the cell at time t+1 and t respectively, N is the neighborhood of the cell, and f is the conversion rule; Markov Model: S t+1 =P ji ×S t S t , S t+1 are the states of ecosystem land use at time t and time t+1, P ji is a state transfer matrix.

3. The soil erosion assessment and prediction method based on a public data set according to claim 1, characterized in that: The natural development scenario is: according to the current urbanization development model and the existing land use conversion rate, a scenario setting is made in which the conversion between land types is not restricted in the conversion rules; the urban expansion suppression scenario: controls the expansion of building land, allows the conversion of building land to other land uses, but does not allow the conversion of other land uses to building land; the ecological priority scenario: takes ecological protection as the principle, practices the concept of protecting the natural environment, and prohibits the conversion of four types of land with high ecological benefits, namely forest land, grassland, water area, and ice / snow, to other land uses; the cultivated land protection scenario: prohibits the conversion of cultivated land to other land uses, while allowing other land types to be transferred to cultivated land and converted into each other.

4. The soil erosion assessment and prediction method based on public data sets according to claim 1, characterized in that: Other influencing factors in the improved RUSLE model refer to the slope and slope length DEM data, vegetation cover height data, soil texture data, precipitation data, and land use data of the calculation base year. The calculation formula is as follows: DEM data: LS=H max -H min Among them, LS is the slope length factor, H max is the highest elevation value, H min is the minimum value of elevation; Vegetation cover height data: Among them, C is the vegetation coverage factor, c is the vegetation coverage; Soil texture data: Among them, S AN , S IL , C LA , C represent the contents of sand, silt, clay and organic matter, respectively; S N1 is the calculation parameter; Precipitation data: Among them, P i is the rainfall in the i-th year (mm), model parameters α=0.0534, β=1.6548; Land use data: After the land use data are assigned, the soil and water conservation measures factor is formed. Specifically, the cultivated land is assigned a value of 0.5, the forest land and grassland are assigned a value of 0.9, the water body, ice / snow, and construction land are assigned a value of 0, the bare land is assigned a value of 1, and the vegetation cover factor is C.

Images