Priority evaluation method for soil and water loss control in small watershed based on soil and water loss pattern recognition

Abstract

The present invention discloses a method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss spot identification, comprising the steps of: S1, obtaining soil and water loss spots of key loss land types in the test area based on soil erosion modulus raster data and land use vector data based on a multi-scale segmentation algorithm; S2, spatially superimposing slope raster data, vegetation coverage raster data and rocky desertification intensity grade raster data, calculating the average slope, average vegetation coverage and rocky desertification intensity grade information of the soil and water loss spots; S3, spatially superimposing vector boundary data of the small watershed range, calculating the comprehensive index M of the severity of soil and water loss in each small watershed; S4, calculating the comprehensive index S of the suitability for control of each small watershed based on the conditions for easy implementation of control measures in each land type; S5, calculating the priority P of soil and water loss control in the small watershed based on the comprehensive indexes M and S. The present invention can realize the identification of key control areas for soil and water loss and the precise spatial positioning of control land types.

Description

Technical Field

[0001] The invention belongs to the technical field of soil and water conservation and remote sensing monitoring, and specifically relates to a method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss pattern recognition. Background Art

[0002] Existing spatial evaluation units are mostly concentrated on the macro scale or regional scale, and there are problems such as the evaluation scale is too large and the spatial distribution boundary of erosion intensity is unclear. Therefore, it is urgent to study a small watershed soil erosion control priority evaluation method based on soil erosion patch identification to achieve more precise control of soil erosion. Summary of the invention

[0003] The purpose of the present invention is to overcome the shortcomings and deficiencies existing in the prior art, and to provide a small watershed soil erosion control priority evaluation method based on soil erosion pattern recognition, so as to overcome the problems of large spatial scale of existing soil erosion remote sensing monitoring, difficulty in accurately locating the spatial position of soil erosion, and insufficient consideration of accurate identification of soil erosion areas that are not suitable for control or have no control potential, and to achieve the identification of key soil erosion control areas in the target area and the accurate spatial positioning of the control land types.

[0004] The purpose of the present invention is achieved through the following technical solutions:

[0005] The priority evaluation method for soil and water loss control in small watersheds based on soil and water loss pattern recognition includes the following steps:

[0006] S1. Based on the soil erosion modulus raster data and land use vector data, the soil and water loss patches of key loss land types including cultivated land, garden land, forest land and grassland in the tested area are obtained based on the multi-scale segmentation algorithm;

[0007] S2, spatially superimposing slope raster data, vegetation coverage raster data and rocky desertification intensity grade raster data, calculating the average slope, average vegetation coverage and rocky desertification intensity grade information of each soil erosion map, and obtaining the first superimposed soil erosion map;

[0008] S3, spatially superimpose the vector boundary data of the small watershed range to obtain the second superimposed soil erosion map with the small watershed as the evaluation unit, calculate the comprehensive index M of the severity of soil erosion in each small watershed based on the second superimposed soil erosion map, and obtain the data M' after normalization, M'∈[0,1] and dimensionless:

[0009] M i =aL i +bQ i +cH i , formula (1),

[0010] Among them, i is the number of the small watershed, and the evaluation indicators of the severity of soil erosion in the small watershed include L, Q and H. The evaluation index L represents the proportion of the area of ​​soil erosion spots in the small watershed to the total area of ​​soil erosion spots in the tested area. The evaluation index Q represents the proportion of moderate and above soil erosion spots in the small watershed to the total area of ​​moderate and above soil erosion spots in the tested area. The evaluation index H represents the proportion of the area of ​​rocky desertification soil erosion spots in the small watershed to the total area of ​​rocky desertification soil erosion spots in the tested area. a, b and c represent the weight coefficients of the evaluation indicators L, Q and H respectively.

[0011] S4. According to the conditions for implementing governance measures on cultivated land, gardens, forests and grasslands, calculate the comprehensive index S of the governance suitability of each small watershed, and obtain the data S' after normalization, S'∈[0,1] and dimensionless;

[0012] S5. Based on the normalized data M' of the comprehensive index M of soil and water loss severity in each small watershed and the normalized data S' of the comprehensive index S of suitability for control, calculate the priority P of soil and water loss control in the small watershed. After normalization, the data P' is obtained, and P'∈[0,1] is dimensionless.

[0013] Preferably, step S1 specifically includes segmenting and clustering the soil erosion modulus raster data within the land use vector boundary by using two parameters, namely, segmentation scale and heterogeneity factor, and merging adjacent homogeneous pixels of the soil erosion modulus raster data to obtain patch objects of different land use types. The specific method is as follows:

[0014] S11, superimposing land use vector data and soil erosion modulus raster data to obtain an image object;

[0015] S12, setting the cutting scale ε;

[0016] S13, merge image pixels and calculate the overall heterogeneity value F of the merged object:

[0017] F=W c ·h c + W s ·h s , formula (2),

[0018] Among them, W c is the weight corresponding to the spectral information, W s is the weight corresponding to the shape information, W c +W s =1; h c is the spectral heterogeneity factor, h s is the shape heterogeneity factor, h c + h s =1,h s According to the compactness factor h cmpctand the smoothness factor h smooth Two parameters are used to represent h cmpct +h smooth =1;

[0019] S14. Compare whether the overall heterogeneity value F is greater than the cutting scale ε:

[0020] If not, the adjacent image pixels are determined to be homogeneous pixels and continue to be merged, returning to step S13;

[0021] If yes, stop merging and get the segmentation result of soil erosion patch;

[0022] S15. Calculate the average soil erosion modulus A of each image block according to the segmentation results, and determine that the cultivated land, garden land, forest land and grassland image blocks with an average erosion modulus value of A ≥ 500 are the final soil and water loss image blocks.

[0023] Preferably, step S2 specifically includes:

[0024] S21. Obtain slope raster data and vegetation coverage raster data of the area to be measured based on the remote sensing monitoring data of soil and water loss, and obtain desertification intensity raster data based on the results of desertification monitoring in karst areas;

[0025] S22, serially numbering each soil and water loss map spot;

[0026] S23. Spatially overlay slope raster data and vegetation coverage raster data, use professional geographic information software, distinguish statistical tools with tables, and obtain the corresponding average slope, average vegetation coverage and average rocky desertification intensity table data for each soil erosion map according to the number field;

[0027] S24. Using the number field as a link, link the map attribute information counted in the form of a table back to the soil erosion map vector data to obtain the soil erosion map results with average slope, average vegetation coverage and desertification intensity level information.

[0028] Preferably, in step S3, the calculation formulas of the evaluation indicators L, Q and H of the severity of soil erosion in a small watershed are:

[0029] L = B i / B s , formula (3),

[0030] Q = G i / G s , formula (4),

[0031] H= J i / J s , formula (5),

[0032] Among them, B i is the area of ​​soil erosion in each small watershed, B s is the total area of ​​soil erosion patches in the area to be tested, G i is the area of ​​moderate and severe soil erosion in each small watershed, G s is the total area of ​​moderate or above soil erosion in the area to be measured, J i is the area of ​​rocky desertification and soil erosion in each small watershed, J s It is the total area of ​​rocky desertification and soil erosion in the area to be measured.

[0033] Preferably, the calculation formula of the comprehensive index S of the suitability for management of each small watershed in step S4 is:

[0034] S=E e (S g +S y )+ E v (S l +S c ), formula (6),

[0035] , formula (7),

[0036] E v =1-E e , formula (8),

[0037] In the formula, S g , S y , S l and S c are the proportions of the areas of soil and water loss control patches of cultivated land, gardens, forests and grasslands in the small watershed to the total areas of soil and water loss control in the measured area, %; cultivated land and gardens belong to the types of land for implementing engineering measures, forests and grasslands belong to the types of land for implementing enclosure measures or biological measures, E e is the ratio of the area treated by engineering measures to the total treated area of ​​the area to be measured, %; j N is the type of engineering measure, which includes high-standard farmland, horizontal terraces, slope-to-terrace conversion, ridges and horizontal steps; j is the area of ​​control by different types of engineering measures; E is the total area of ​​comprehensive control of soil erosion in the tested area, E v It is the ratio of the sum of the areas under closure and biological measures to the total area under test, %. The types of biological measures include: tending, forest economy, transformation of inefficient forests and artificial replanting measures.

[0038] Preferably, the conditions for easy implementation of treatment measures for the cultivated land, garden land, forest land and grassland in step S4 are:

[0039] The average slope of the cultivated land patches corresponding to the patches suitable for soil and water loss control is less than 25°. Such patches are easy to construct high-standard farmland, horizontal terraces or slope-to-terrace engineering measures;

[0040] The average slope of the garden plot corresponding to the plot suitable for soil and water loss control is less than 25°, and such plots are easy to implement ridge or horizontal terrace land preparation measures;

[0041] Forest land suitable for water and soil erosion control includes: forest land with an average vegetation coverage greater than or equal to 75%, such land is easy to implement closure measures; forest land with an average vegetation coverage less than 75% and an average slope less than 25°, such land is easy to implement tending, forest economy or low-efficiency forest transformation measures;

[0042] Grassland patches suitable for soil and water loss control include: grassland patches with an average vegetation coverage greater than or equal to 60%, and such patches are easy to implement enclosure measures; grassland patches with an average vegetation coverage less than 60% and an average slope less than 25°, and such patches are easy to implement tending or artificial replanting measures.

[0043] Preferably, the calculation formula for the soil and water loss control priority P of the small watershed in step S5 is:

[0044] P = M' + S', formula (9),

[0045] After normalizing P, we get the data P', where the larger the P' value is, the higher the priority of small watershed management is.

[0046] Preferably, in steps S3 to S5, the formula for normalization processing is:

[0047] Z=(XX min ) / ( X max -X min ), formula (10),

[0048] Among them, X is the data before normalization, including comprehensive indicators M, S and priority P, Z is the corresponding data after normalization, including M', S' and P', X max and X min are the maximum and minimum values ​​of the corresponding data samples before normalization.

[0049] Preferably, after step S2, the method further includes: verifying the soil and water loss image segmentation result according to a relative error evaluation method:

[0050] Q=|K a -K b | / K b 100%, formula (11),

[0051] Among them, Q is the relative error of area, Ka K is the soil erosion area calculated based on the soil erosion map; b It is the area of ​​soil erosion calculated based on remote sensing monitoring results.

[0052] Preferably, in step S3, the weight coefficients a, b and c are 0.5, 0.3 and 0.2 respectively.

[0053] Compared with the prior art, the present invention has the following advantages and effects:

[0054] (1) The method for evaluating the priority of soil erosion control in small watersheds based on soil erosion spot identification proposed in the present invention makes full use of the results of soil erosion remote sensing monitoring, realizes effective connection with the current soil and water conservation rate value of the area to be tested, identifies soil erosion spots of key erosion land types through a multi-scale segmentation algorithm, and uses soil erosion spots as a scale to judge the degree of suitability for control. Finally, the small watershed is used as the evaluation unit, and the severity of soil erosion and the degree of suitability for control of soil erosion spots are used as parameter indicators to jointly determine the priority of soil erosion control in small watersheds. It can accurately locate the priority control key areas of soil erosion and the spatial location of soil erosion spots that are suitable for control. It overcomes the problems of large spatial scale of soil erosion remote sensing monitoring in the prior art, difficulty in accurately locating the spatial location of soil erosion, and failure to fully consider the accurate judgment of soil erosion areas that are not suitable for control or have no control potential.

[0055] (2) The method of the present invention combines the results of remote sensing monitoring to determine the priority of soil and water loss control in small watersheds from two dimensions: the severity of soil and water loss and the suitability of soil and water loss map for control. It comprehensively considers the natural attributes of soil and water loss in the measured area and the operability of control, and provides technical support for promoting the scientific, precise and efficient decomposition and phased soil and water loss control tasks, key control project layout, soil and water conservation planning and other work in the target area. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The present invention is a flow chart of a method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss pattern recognition. DETAILED DESCRIPTION

[0057] The present invention will be further described in detail below in conjunction with embodiments and drawings, but the embodiments of the present invention are not limited thereto.

[0058] Example

[0059] like Figure 1 The figure shows a flow chart of a method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss patch identification, including the following steps:

[0060] S1. Based on the soil erosion modulus raster data and land use vector data, the soil and water loss patches of key loss land types including cultivated land, garden land, forest land and grassland in the tested area are obtained based on the multi-scale segmentation algorithm;

[0061] S2, spatially superimposing slope raster data, vegetation coverage raster data and rocky desertification intensity grade raster data, calculating the average slope, average vegetation coverage and rocky desertification intensity grade information of each soil erosion map, and obtaining the first superimposed soil erosion map;

[0062] S3, spatially superimpose the vector boundary data of the small watershed range to obtain the second superimposed soil erosion map with the small watershed as the evaluation unit, calculate the comprehensive index M of the severity of soil erosion in each small watershed based on the second superimposed soil erosion map, and obtain the data M' after normalization, M'∈[0,1] and dimensionless:

[0063] M i =aL i +bQ i +cH i , formula (1),

[0064] Among them, i is the number of the small watershed, and the evaluation indicators of the severity of soil erosion in the small watershed include L, Q and H. The evaluation index L represents the proportion of the area of ​​soil erosion spots in the small watershed to the total area of ​​soil erosion spots in the tested area. The evaluation index Q represents the proportion of moderate and above soil erosion spots in the small watershed to the total area of ​​moderate and above soil erosion spots in the tested area. The evaluation index H represents the proportion of the area of ​​rocky desertification soil erosion spots in the small watershed to the total area of ​​rocky desertification soil erosion spots in the tested area. a, b and c represent the weight coefficients of the evaluation indicators L, Q and H respectively.

[0065] S4. According to the conditions for implementing governance measures on cultivated land, gardens, forests and grasslands, calculate the comprehensive index S of the governance suitability of each small watershed, and obtain the data S' after normalization, S'∈[0,1] and dimensionless;

[0066] S5. Based on the normalized data M' of the comprehensive index M of soil and water loss severity in each small watershed and the normalized data S' of the comprehensive index S of suitability for control, calculate the priority P of soil and water loss control in the small watershed. After normalization, the data P' is obtained, and P'∈[0,1] is dimensionless.

[0067] Specifically, based on the soil erosion modulus raster data obtained from the remote sensing monitoring of soil erosion in the industry and the soil erosion map obtained from the land use vector data, the results can be connected to the soil erosion remote sensing monitoring data and the current soil and water conservation rate value of the area to be tested in the industry, so as to achieve accurate identification and positioning of soil erosion maps divided by land type and erosion intensity level. In this embodiment, the weight coefficients a, b and c in step S3 are respectively taken as 0.5, 0.3 and 0.2. The comprehensive index M of soil and water loss severity is normalized to obtain data M', which is dimensionless and between 0 and 1. The larger the value, the more serious the soil and water loss in the small watershed. Therefore, the result can evaluate the priority of soil and water loss severity of each small watershed in the test area, and truly reflects the natural distribution characteristics of soil and water loss in the small watershed; and the comprehensive index S of the suitability of small watershed governance is normalized to obtain data S', which is dimensionless and between 0 and 1. The larger the value, the higher the suitability of small watershed governance. Under normal circumstances, it may appear that the soil and water loss severity of small watersheds in the test area is ranked first, but the area that can be governed or suitable for governance is small. Therefore, through the comprehensive index S of suitability for governance, the spatial distribution position of small watersheds suitable for governance in the test area can be accurately located. The priority of soil and water loss governance in small watersheds P is normalized to obtain data P', which is dimensionless and between 0 and 1. The larger the value, the more priority the small watershed should be given to governance. By comprehensively considering the two dimensions of natural characteristics of soil erosion and suitability for management of small watersheds in the tested area, it is possible to accurately guide the selection of small watersheds for priority management in the tested area, and provide a more scientific and operational basis for management decision-making to achieve the phased soil and water conservation rate targets in the tested area.

[0068] Step S1 specifically includes segmenting and clustering the soil erosion modulus raster data within the land use vector boundary by using two parameters, namely, segmentation scale and heterogeneity factor, and merging adjacent homogeneous pixels of the soil erosion modulus raster data to obtain patch objects of different land use types. The specific method is as follows:

[0069] S11, superimposing land use vector data and soil erosion modulus raster data to obtain an image object;

[0070] S12, setting the cutting scale ε;

[0071] S13, merge image pixels and calculate the overall heterogeneity value F of the merged object:

[0072] F=W c ·h c + W s ·h s , formula (2),

[0073] Among them, W c is the weight corresponding to the spectral information, W s is the weight corresponding to the shape information, W c+W s =1; h c is the spectral heterogeneity factor, h s is the shape heterogeneity factor, h c + h s =1,h s According to the compactness factor h cmpct and the smoothness factor h smooth Two parameters are used to represent h cmpct +h smooth =1;

[0074] S14. Compare whether the overall heterogeneity value F is greater than the cutting scale ε:

[0075] If not, the adjacent image pixels are determined to be homogeneous pixels and continue to be merged, returning to step S13;

[0076] If yes, stop merging and get the segmentation result of soil erosion patch;

[0077] S15. Calculate the average soil erosion modulus A of each image block according to the segmentation results, and determine that the cultivated land, garden land, forest land and grassland image blocks with an average erosion modulus value of A ≥ 500 are the final soil and water loss image blocks.

[0078] Specifically, the soil erosion modulus raster data does not contain land use type information. In order to prevent the raster pixels of different land use types from being confused when clustering the soil erosion modulus, the land use vector data and the soil erosion modulus raster data are superimposed. The cutting scale parameter represents the standard of object segmentation size, which refers to the maximum heterogeneity when merging rasters. The larger the cutting scale parameter, the larger the boundary range of the resulting patch object.

[0079] According to step S13, the spectral heterogeneity factor h c The specific calculation method is:

[0080] ,

[0081] Among them, b is the image band, w b is the weight value of the b-th band, σ b is the spectral standard deviation of band b; the soil erosion modulus raster data has only one band, w b The value is 1;

[0082] The specific calculation process is illustrated as follows: Assume that the standard deviations of two adjacent soil erosion modulus grid pixels c1 and c2 are σ c1 and σ c2 , the areas are a c1 and a c2 , a merge represents the total area after merging, σ mergeis the standard deviation after merging, then the spectral heterogeneity of the object after merging c1 and c2 is:

[0083] ,

[0084] Shape heterogeneity factor h s The calculation method is:

[0085] h s =w cmpct h cmpct + (1-w cmpct )h smooth ,

[0086] In the formula, w cmpct is the compactness weight, 1-w cmpct is the smoothness weight, h cmpct is the compactness factor, h smooth are the smoothness factors, and their calculation methods are:

[0087] ,

[0088] h smooth = l / d,

[0089] In the formula, l is the perimeter of the merged object, N is the total number of pixels of the merged object, and d is the perimeter of the minimum circumscribed rectangle of the merged object.

[0090] Using the Ecognition multi-scale segmentation software, based on the calculation principle of the overall heterogeneity value F in step S13, it is only necessary to set the cutting scale ε and the corresponding weight W of the shape information s and the compactness weight w cmpct There are 3 parameters in total. Usually, high-intensity erosion is spatially fragmented. To avoid high-intensity erosion being merged during clustering, which leads to the loss of high-intensity erosion, in this embodiment, the cutting scale ε is determined to be 200, and the shape information corresponds to the weight W. s and the compactness weight w cmpct The parameter values ​​are 0.1 and 0.5 respectively, and the parameters have good versatility.

[0091] The multi-scale segmentation is based on the object of the first segmentation, and the heterogeneity between adjacent soil erosion modulus grids is calculated in a cyclic iterative manner, and the adjacent homogeneous pixels are merged. When the heterogeneity exceeds the set cutting parameter ε threshold (ε is 200 in this embodiment), the merging is stopped to obtain the final patch segmentation result. Further, based on the above patch segmentation results, the average soil erosion modulus A of each patch is calculated. According to the "Soil Erosion Classification and Grading Standard" (SL190-2007), A<500 is a slight patch (no soil and water loss), 500≤A<2500 is a mild soil and water loss patch, 2500≤A<5000 is a moderate soil and water loss patch, 5000≤A<8000 is a strong soil and water loss patch, 8000≤A<15000 is an extremely strong soil and water loss patch, and A≥15000 is a severe soil and water loss patch. The cultivated land, garden land, forest land and grassland patches with an average erosion modulus value of A ≥ 500 are determined as soil and water loss patches, and patches without soil and water loss with A < 500 are discarded.

[0092] Step S2 specifically includes:

[0093] S21. Obtain slope raster data and vegetation coverage raster data of the area to be measured based on the remote sensing monitoring data of soil and water loss, and obtain desertification intensity raster data based on the results of desertification monitoring in karst areas;

[0094] S22, serially numbering each soil and water loss map spot;

[0095] S23. Spatially overlay slope raster data and vegetation coverage raster data, use professional geographic information software, distinguish statistical tools with tables, and obtain the corresponding average slope, average vegetation coverage and average rocky desertification intensity table data for each soil erosion map according to the number field;

[0096] S24. Using the number field as a link, link the map attribute information counted in the form of a table back to the soil erosion map vector data to obtain the soil erosion map results with average slope, average vegetation coverage and desertification intensity level information.

[0097] Specifically, in this embodiment, the slope raster data and vegetation coverage raster data of a certain area can be obtained through the results of remote sensing monitoring of soil and water loss in the industry, and the desertification intensity level raster data can be obtained through the results of monitoring desertification in karst areas, where the desertification intensity level can be divided into no desertification (level 0), mild desertification (level 1), moderate desertification (level 2), and severe desertification (level 3). Each soil and water loss map is numbered continuously, and the slope raster data and vegetation coverage raster data of a certain area are spatially superimposed. The table partition statistics tool in the professional geographic information software ArcGIS is used, and the statistical type selects the mean value Mean, and the average slope and average vegetation coverage inside each map are calculated in turn according to the number field; the desertification intensity level raster data is spatially superimposed, and the statistical type selects the majority Majority, and the desertification intensity level value that appears most in each soil and water loss map is determined in turn according to the number field and used as the desertification intensity level of the map. The above statistical results are all tabular data. Using the number field as a link, the map attribute information counted in tabular form is linked back to the soil erosion map vector data to obtain soil erosion map results with average slope, average vegetation coverage and desertification intensity level information.

[0098] In step S3, the calculation formulas of the evaluation indicators L, Q and H of the severity of soil erosion in the small watershed are:

[0099] L = B i / B s , formula (3),

[0100] Q = G i / G s , formula (4),

[0101] H= J i / J s , formula (5),

[0102] Among them, B i is the area of ​​soil erosion in each small watershed, B s is the total area of ​​soil erosion patches in the area to be tested, G i is the area of ​​moderate and severe soil erosion in each small watershed, G s is the total area of ​​moderate or above soil erosion in the area to be measured, J i is the area of ​​rocky desertification and soil erosion in each small watershed, J s It is the total area of ​​rocky desertification and soil erosion in the area to be measured.

[0103] Specifically, in this embodiment, the area tabulation tool in the professional geographic information software ArcGIS is used to calculate the soil erosion patch area B, the moderate and above soil erosion patch area G, and the rocky desertification soil erosion patch area J within the boundary of each small watershed. Among them, the moderate and above soil erosion patch refers to the patch with an average erosion modulus A ≥ 2500 within the patch; the rocky desertification soil erosion patch refers to the patch with a rocky desertification intensity level of 1 or above within the patch.

[0104] The calculation formula of the comprehensive index S of the suitability for management of each small watershed in step S4 is:

[0105] S=E e (S g +S y )+ E v (S l +S c ), formula (6),

[0106] , formula (7),

[0107] E v =1-E e , formula (8),

[0108] In the formula, S g , S y , S l and S c are the proportions of the areas of soil and water loss control patches of cultivated land, gardens, forests and grasslands in the small watershed to the total areas of soil and water loss control in the measured area, %; cultivated land and gardens belong to the types of land for implementing engineering measures, forests and grasslands belong to the types of land for implementing enclosure measures or biological measures, E e is the proportion of the area under engineering measures to the total area under test, %; j is the type of engineering measures, which include high-standard farmland, horizontal terraces, slope-to-terrace conversion, ridges and horizontal terraces; N j is the area of ​​control by different types of engineering measures; E is the total area of ​​comprehensive control of soil erosion in the tested area, E v It is the ratio of the sum of the areas under closure and biological measures to the total area under test, %. The types of biological measures include: tending, forest economy, transformation of inefficient forests and artificial replanting measures.

[0109] The conditions for easy implementation of treatment measures for the cultivated land, garden land, forest land and grassland in step S4 are:

[0110] The average slope of the cultivated land patches corresponding to the patches suitable for soil and water loss control is less than 25°. Such patches are easy to construct high-standard farmland, horizontal terraces or slope-to-terrace engineering measures;

[0111] The average slope of the garden plot corresponding to the plot suitable for soil and water loss control is less than 25°, and such plots are easy to implement ridge or horizontal terrace land preparation measures;

[0112] Forest land suitable for water and soil erosion control includes: forest land with an average vegetation coverage greater than or equal to 75%, such land is easy to implement closure measures; forest land with an average vegetation coverage less than 75% and an average slope less than 25°, such land is easy to implement tending, forest economy or low-efficiency forest transformation measures;

[0113] Grassland patches suitable for soil and water loss control include: grassland patches with an average vegetation coverage greater than or equal to 60%, and such patches are easy to implement enclosure measures; grassland patches with an average vegetation coverage less than 60% and an average slope less than 25°, and such patches are easy to implement tending or artificial replanting measures.

[0114] Specifically, because the implementation of control measures in moderate and above rocky desertification areas is difficult, requires high investment and has low effectiveness, the suitable control areas are all selected from light and below rocky desertification soil erosion patches.

[0115] The calculation formula for the soil and water loss control priority P of small watersheds in step S5 is:

[0116] P = M' + S', formula (9),

[0117] After normalizing P, we get the data P', where the larger the P' value is, the higher the priority of small watershed management is.

[0118] In steps S3 to S5, the formula for normalization processing is:

[0119] Z=(XX min ) / ( X max -X min ), formula (10),

[0120] Among them, X is the data before normalization, including comprehensive indicators M, S and priority P, Z is the corresponding data after normalization, including M', S' and P', X max and X min are the maximum and minimum values ​​of the corresponding data samples before normalization.

[0121] Specifically, for example, when the comprehensive index M of the severity of soil and water loss in a small watershed, the comprehensive index S of the suitability for management, and the priority of soil and water loss management P are normalized, according to formula (10), we have: M' = (M - M min ) / (M max -M min ), S' = (S - S min ) / ( S max -Smin ),

[0122] P'=(PP min ) / ( P max -P min ).

[0123] After step S2, the method further includes: verifying the soil and water loss image segmentation result according to the relative error evaluation method:

[0124] Q=|K a -K b | / K b 100%, formula (11),

[0125] Among them, Q is the relative error of area, K a K is the soil erosion area calculated based on the soil erosion map; b It is the area of ​​soil erosion calculated based on remote sensing monitoring results.

[0126] Specifically, in order to verify the accuracy of the segmentation results of soil erosion patches, the areas of soil erosion patches of each level were compared with the results of soil erosion dynamic monitoring in the industry. The results of soil erosion dynamic monitoring include the total area of ​​soil erosion in each category and the statistical data of soil erosion areas by erosion intensity level. Relative error analysis was used to compare the differences between the areas of soil erosion patches of each erosion intensity level and the areas of soil erosion remote sensing monitoring.

[0127] Specifically, taking a certain area as an example, based on the remote sensing monitoring of soil and water loss spatial data, the soil and water loss spots of key erosion land types are identified, and the priority of small watershed management is quantitatively evaluated, so as to explain in detail the implementation process of the small watershed soil and water loss management priority evaluation method based on soil and water loss spot identification of the present invention.

[0128] Step 1: Combine land use vector data and soil erosion raster data, based on the principle of multi-scale segmentation method in step S1, use Ecognition software, set the cutting scale ε to 200, and the shape information corresponding weight W s and the compactness weight w cmpct The values ​​are 0.1 and 0.5 respectively. Based on the first segmented objects, merge the first segmented objects within the land use vector boundary, calculate the overall heterogeneity value F1, and judge whether F1 is less than ε: If so, continue to merge based on the objects generated by the first merger, calculate the overall heterogeneity value F2, and judge whether F2 is less than ε again: If so, continue to merge until the objects generated based on the n-1 mergers have an overall heterogeneity value F nWhen the value is greater than or equal to ε, the merging stops and the final patch object results of different land use types are obtained. The patches of cultivated land, garden land, forest land and grassland with an average erosion modulus value of A ≥ 500 are determined as soil erosion patches. The areas of soil erosion patches with different erosion intensities in various types are counted.

[0129] Based on the statistical data of soil and water loss monitoring results in the industry, the annual soil and water loss areas of cultivated land, gardens, forests and grasslands in a certain region are 274.70 km2, 9.01 km2, 117.75 km2 and 35.32 km2 respectively. The soil and water loss areas of different erosion intensity levels are shown in Table 1. In order to verify the accuracy of the segmentation results of the soil and water loss patches, the relative error is calculated according to formula (11), and the differences between the soil and water loss patch areas of each erosion intensity level of cultivated land, gardens, forests and grasslands and the soil and water loss remote sensing monitoring areas are compared. The results are shown in Table 1.

[0130] The total relative errors between the soil erosion area of ​​cultivated land, garden land, forest land and grassland and remote sensing monitoring were -0.32%, 2.66%, 0.02% and 2.35% respectively, all below 5%. From the perspective of erosion intensity level, the relative errors of each erosion intensity level were also below 20%. Due to the small base number of high-intensity erosion area, the relative errors of some land types were high. The overall soil erosion area was in good agreement with the remote sensing monitoring results.

[0131] Table 1 Evaluation of soil and water loss map accuracy for key erosion land types in a certain area

[0132]

[0133] Step 2: Obtain slope raster data and vegetation coverage raster data of a certain area based on the results of remote sensing monitoring of soil erosion in the industry, and obtain rocky desertification intensity grade raster data based on the results of rocky desertification monitoring in karst areas in the industry. Use professional geographic information software ArcGIS to obtain soil erosion map results with average slope, average vegetation coverage and rocky desertification grade intensity information.

[0134] Step 3: Spatially superimpose the soil erosion map obtained in step 2 with the small watershed range vector data. First, calculate the evaluation indicators L, Q and H of the severity of soil erosion in the small watershed according to formulas (3) to (5), calculate the comprehensive index M of the severity of soil erosion in each small watershed according to formula (1), and then obtain the data M' after normalization according to formula (10). The larger the M' value in the small watershed, the higher the severity of soil erosion. Sort the calculated results M' from large to small, and the numbers of the top 10 small watersheds are: 30, 12, 27, 44, 57, 18, 9, 50, 52, and 99.

[0135] Step 4: According to the soil erosion map conditions of cultivated land, gardens, woodlands and grasslands where control measures are easy to implement, select the soil erosion map spots of cultivated land, gardens, woodlands and grasslands that are suitable for control, and count the total area of ​​the selected soil erosion map spots that are suitable for control, including cultivated land, gardens, woodlands and grasslands; then use the table partition display tool in the professional geographic information software ArcGIS to count the area of ​​soil erosion map spots that are suitable for control in each small watershed according to the small watershed number; finally, calculate the proportion of the area of ​​soil erosion map spots that are suitable for control of cultivated land, gardens, woodlands and grasslands in each small watershed to the total area of ​​soil erosion that is suitable for control in a certain area, and get the corresponding S g , S y , S l and S c ;

[0136] According to the soil and water conservation data of a certain region in recent years, the proportions of the conservation areas of engineering measures, wind and rain measures and biological measures are calculated according to formulas (7) and (8) to be 68% and 32% respectively. Then E e and E v The values ​​are 0.68 and 0.32 respectively; then the above values ​​are substituted into formula (6) to calculate the comprehensive index S of the suitability for management of each small watershed, and then the data S' is obtained after normalization according to formula (10). The larger the S' value in the small watershed, the higher the suitability for management. The calculated results S' are sorted from large to small, and the numbers of the top 10 small watersheds are: 18, 30, 12, 15, 52, 94, 19, 47, 53, and 27.

[0137] Step 5: Based on the analysis results of the comprehensive index M of the severity of soil and water loss in the small watershed and the comprehensive index S of the suitability for management, the priority P of soil and water loss management in the small watershed is calculated according to formula (9), and then the data P' is obtained after normalization according to formula (10). The larger the P' value in the small watershed, the higher the priority of soil and water loss management. The calculated results P' are sorted, and the numbers of the top 10 small watersheds in priority management are: 30, 18, 12, 27, 52, 44, 19, 94, 83, and 9.

[0138] A region can select small watersheds for priority management based on the phased soil and water conservation rate targets and the management priority of small watersheds obtained according to the above analysis results, and give priority to comprehensive soil and water loss management work based on the spatial distribution locations of soil and water loss patches that are suitable for management in small watersheds.

[0139] In summary, the present invention can start from the key objects of soil and water loss control, take small watersheds as evaluation units to quantitatively evaluate the priority of soil and water loss control, highlight the spatial distribution of key areas of priority soil and water loss control, and accurately identify key erosion land types suitable for control. It can provide technical support and decision-making basis for scientifically, meticulously and efficiently carrying out precise soil and water loss control work in each tested area.

[0140] The above embodiments are preferred implementations of the present invention and are not intended to limit the present invention. Any other changes or other equivalent replacement methods that do not deviate from the technical solutions of the present invention are included in the protection scope of the present invention.

Claims

1. A method for evaluating the priority of soil and water loss control in small watersheds based on soil and water loss pattern recognition, characterized in that: Includes steps: S1. Based on the soil erosion modulus raster data and land use vector data, the key land loss categories in the area to be measured are obtained based on the multi-scale segmentation algorithm. The key land loss categories in the area to be measured include soil erosion patches of cultivated land, gardens, forests and grasslands; S2, spatially superimposing slope raster data, vegetation coverage raster data and rocky desertification intensity grade raster data, calculating the average slope, average vegetation coverage and rocky desertification intensity grade information of each soil erosion map, and obtaining the first superimposed soil erosion map; S3, spatially superimpose the vector boundary data of the small watershed range to obtain the second superimposed soil erosion map with the small watershed as the evaluation unit, calculate the comprehensive index M of the severity of soil erosion in each small watershed based on the second superimposed soil erosion map, and obtain the data M' after normalization, M'∈[0,1] and dimensionless: M i = aL i + bQ i + cH i , Equation (1), Among them, i is the number of the small watershed, and the evaluation indicators of the severity of soil erosion in the small watershed include L, Q and H. The evaluation index L represents the proportion of the area of ​​soil erosion spots in the small watershed to the total area of ​​soil erosion spots in the tested area. The evaluation index Q represents the proportion of moderate and above soil erosion spots in the small watershed to the total area of ​​moderate and above soil erosion spots in the tested area. The evaluation index H represents the proportion of the area of ​​rocky desertification soil erosion spots in the small watershed to the total area of ​​rocky desertification soil erosion spots in the tested area. a, b and c represent the weight coefficients of the evaluation indicators L, Q and H respectively. S4. According to the conditions for implementing governance measures on cultivated land, gardens, forests and grasslands, calculate the comprehensive index S of the governance suitability of each small watershed, and obtain the data S' after normalization, S'∈[0,1] and dimensionless; S5. Calculate the priority P of soil and water loss control in the small watershed according to the normalized data M' of the comprehensive index M of the severity of soil and water loss in each small watershed and the normalized data S' of the comprehensive index S of the suitability for control, and obtain the data P' after normalization, P'∈[0,1] and dimensionless; In step S3, the calculation formulas of the evaluation indicators L, Q and H of the severity of soil erosion in the small watershed are: L = B i / B s , formula (3), Q = G i / G s , formula (4), H= J i / J s , formula (5), Among them, B i is the area of ​​soil erosion in each small watershed, B s is the total area of ​​soil erosion patches in the area to be tested, G i is the area of ​​moderate and severe soil erosion in each small watershed, G s is the total area of ​​moderate or above soil erosion in the area to be measured, J i is the area of ​​rocky desertification and soil erosion in each small watershed, J s The total area of ​​rocky desertification and soil erosion in the area to be measured; The calculation formula of the comprehensive index S of the suitability for management of each small watershed in step S4 is: S=E e (S g +S y )+ E v (S l +S c ), formula (6), , formula (7), E v =1-E e , Equation (8), In the formula, S g , S y , S l and S c are the proportions of the areas of soil and water loss control patches of cultivated land, gardens, forests and grasslands in the small watershed to the total areas of soil and water loss control in the measured area, %; cultivated land and gardens belong to the types of land for implementing engineering measures, forests and grasslands belong to the types of land for implementing enclosure measures or biological measures, E e is the ratio of the area treated by engineering measures to the total treated area of ​​the area to be measured, %; j N is the type of engineering measure, which includes high-standard farmland, horizontal terraces, slope-to-terrace conversion, ridges and horizontal steps; j is the area of ​​control by different types of engineering measures; E is the total area of ​​comprehensive control of soil erosion in the tested area, E v It is the ratio of the sum of the areas under closure and biological measures to the total area under test, %. The types of biological measures include: tending, forest economy, transformation of inefficient forests and artificial replanting measures.

2. The method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss pattern recognition according to claim 1 is characterized in that: Step S1 specifically includes segmenting and clustering the soil erosion modulus raster data within the land use vector boundary by using two parameters, namely, segmentation scale and heterogeneity factor, and merging adjacent homogeneous pixels of the soil erosion modulus raster data to obtain patch objects of different land use types. The specific method is as follows: S11, superimposing land use vector data and soil erosion modulus raster data to obtain an image object; S12, setting the cutting scale ε; S13, merge image pixels and calculate the overall heterogeneity value F of the merged object: F = W c ·h c + W s ·h s , Equation (2), Among them, W c is the weight corresponding to the spectral information, W s is the weight corresponding to the shape information, W c +W s =1; h c is the spectral heterogeneity factor, h s is the shape heterogeneity factor, h c + h s =1,h s According to the compactness factor h cmpct and the smoothness factor h smooth Two parameters are used to represent h cmpct +h smooth =1; S14. Compare whether the overall heterogeneity value F is greater than the cutting scale ε: If not, the adjacent image pixels are determined to be homogeneous pixels and continue to be merged, returning to step S13; If yes, stop merging and get the segmentation result of soil erosion patch; S15. Calculate the average soil erosion modulus A of each patch according to the segmentation result, and determine the patches of cultivated land, garden land, forest land and grassland with an average soil erosion modulus A ≥ 500 as the final soil and water loss patches.

3. The method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss pattern recognition according to claim 1 is characterized in that: Step S2 specifically includes: S21. Obtain slope raster data and vegetation coverage raster data of the area to be measured based on the remote sensing monitoring data of soil and water loss, and obtain desertification intensity raster data based on the results of desertification monitoring in karst areas; S22, serially numbering each soil and water loss map spot; S23. Spatially overlay slope raster data and vegetation coverage raster data, use professional geographic information software, distinguish statistical tools with tables, and obtain the corresponding average slope, average vegetation coverage and average rocky desertification intensity table data for each soil erosion map according to the number field; S24. Using the number field as a link, link the map attribute information counted in the form of a table back to the soil erosion map vector data to obtain the soil erosion map results with average slope, average vegetation coverage and desertification intensity level information.

4. The method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss pattern recognition according to claim 1 is characterized in that: The conditions for easy implementation of treatment measures for the cultivated land, garden land, forest land and grassland in step S4 are: The average slope of the cultivated land patches corresponding to the patches suitable for soil and water loss control is less than 25°. Such patches are easy to construct high-standard farmland, horizontal terraces or slope-to-terrace engineering measures; The average slope of the garden plot corresponding to the plot suitable for soil and water loss control is less than 25°, and such plots are easy to implement ridge or horizontal terrace land preparation measures; Forest land suitable for water and soil erosion control includes: forest land with an average vegetation coverage greater than or equal to 75%, such land is easy to implement closure measures; forest land with an average vegetation coverage less than 75% and an average slope less than 25°, such land is easy to implement tending, forest economy or low-efficiency forest transformation measures; Grassland patches suitable for soil and water loss control include: grassland patches with an average vegetation coverage greater than or equal to 60%, and such patches are easy to implement enclosure measures; grassland patches with an average vegetation coverage less than 60% and an average slope less than 25°, and such patches are easy to implement tending or artificial replanting measures.

5. The method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss pattern recognition according to claim 1 is characterized in that: The calculation formula for the soil and water loss control priority P of small watersheds in step S5 is: P = M' + S', formula (9), After normalizing P, we get the data P', where the larger the P' value is, the higher the priority of small watershed management is.

6. The method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss pattern recognition according to claim 1 is characterized in that: In steps S3 to S5, the formula for normalization processing is: Z=(XX min ) / ( X max -X min ), formula (10), Among them, X is the data before normalization, including comprehensive indicators M, S and priority P, Z is the corresponding data after normalization, including M', S' and P', X max and X min are the maximum and minimum values ​​of the corresponding data samples before normalization.

7. The method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss pattern recognition according to claim 1 is characterized in that: After step S2, the method further includes: verifying the soil and water loss image segmentation result according to the relative error evaluation method: Q=|K a -K b | / K b 100%, formula (11), Among them, Q is the relative error of area, K a is the soil erosion area calculated based on the soil erosion map, K b It is the area of ​​soil erosion calculated based on remote sensing monitoring results.

8. The method for evaluating the priority of soil and water loss control in a small watershed based on soil and water loss pattern recognition according to claim 1 is characterized in that: In step S3, the weight coefficients a, b and c are set to 0.5, 0.3 and 0.2 respectively.

Images