Nonlinear evolution pathway and characteristic evaluation method of regional soil organic carbon and its application

Abstract

The present invention belongs to the field of soil science and technology and provides a method and application for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon. The method first calculates environmental variable data and soil organic carbon data for each pixel in the region based on multi-source remote sensing image data. Then, a self-organizing map network is used to classify the state of soil organic carbon. The nonlinear evolution path of the soil organic carbon state type is then identified in the constructed state space. Finally, nonlinear characteristics such as hysteresis, resilience, and mutation thresholds are determined based on the nonlinear evolution path of soil organic carbon. The method uses soil organic carbon data from different spatial and temporal locations to reconstruct its nonlinear evolution path, addressing the lack of historical soil organic carbon data. The method also rapidly and accurately quantifies the nonlinear characteristics of the soil organic carbon evolution path, such as hysteresis, resilience, and mutation thresholds, to achieve trend prediction and situation warning for soil organic carbon, thereby serving soil organic carbon management.

Description

Technical Field

[0001] The present invention belongs to the field of soil science and technology, and in particular relates to a nonlinear evolution path and characteristic evaluation method of regional soil organic carbon. Background Art

[0002] Given the crucial role of soil organic carbon in mitigating global climate change and ensuring food security, numerous initiatives and targets have been launched globally to increase soil organic carbon. However, as Earth enters the Anthropocene, the unpredictable, nonlinear changes in soil organic carbon under human land use and management make accurate early warning of soil organic carbon trends difficult. Therefore, identifying the complex evolutionary pathways of soil organic carbon and quantifying its nonlinear characteristics are pressing challenges in soil organic carbon management.

[0003] In ecology, steady-state transitions refer to large-scale, sustained, and abrupt changes in ecosystem structure and function, resulting in the rapid reorganization of an ecosystem from one relatively stable state to another. Two conceptual models have been developed to address this phenomenon: the S-shaped model and the ball-and-cup model. The S-shaped model describes the bistability and steady-state transitions of an ecosystem under changing external environmental conditions within a two-dimensional plane, while the ball-and-cup model illustrates the steady-state transition process under external disturbances. Therefore, conceptual models can be used to determine the range of multiple stable states in an ecosystem and the critical points at which steady-state transitions occur, enabling the analysis and prediction of the nonlinear evolutionary characteristics of the ecosystem. These conceptual models provide an important foundation for the assessment of the nonlinear evolutionary pathways and characteristics of soil organic carbon.

[0004] Although the S-shaped and ball-cup conceptual models can effectively describe the nonlinear evolutionary characteristics of ecosystems under changes in the external environment, these characteristics cannot be quantitatively expressed in the conceptual models, especially the nonlinear evolutionary characteristics of soil organic carbon under human land use and management.

[0005] Furthermore, because soil organic carbon data often relies on soil profiles, it is difficult to obtain long-term continuous measured data and describe the evolution path of soil organic carbon. Therefore, the present invention provides a method and application for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon. Summary of the Invention

[0006] The purpose of the present invention is to overcome the existing defects and provide a regional soil organic carbon nonlinear evolution path and characteristic evaluation method and application, using soil organic carbon data in different spaces at the same time to reconstruct its nonlinear evolution path, and quickly and accurately quantify the nonlinear characteristics of the soil organic carbon evolution path such as hysteresis, toughness and mutation threshold, so as to realize soil organic carbon trend prediction and situation warning, thereby serving the improvement and management of soil organic carbon.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] The first object of the present invention is to provide a method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon, comprising:

[0009] Acquiring multi-source remote sensing image data of a target area, wherein each pixel in the multi-source remote sensing image data represents surface reflectance data, surface temperature data, and elevation data at a corresponding location;

[0010] Using a machine learning model, based on the surface reflectance data, surface temperature data, and elevation data of each pixel in the multi-source remote sensing image data, the soil organic carbon content and its environmental variable data of each pixel are calculated;

[0011] Using the self-organizing map network model, the soil organic carbon content of each pixel and environmental variable data were used as input vectors to classify all pixels in the area into several different soil organic carbon status types.

[0012] A two-dimensional state space was constructed based on the average cover function index and average soil organic carbon content of each soil organic carbon state type. A quadratic function was used to fit the degradation path curve and the recovery path curve in the constructed state space to identify the nonlinear evolution path of soil organic carbon.

[0013] Based on the nonlinear evolution path of soil organic carbon, the mutation threshold, hysteresis and resilience of the soil organic carbon evolution path are determined.

[0014] Furthermore, the step of calculating the environmental variable data includes:

[0015] (1) Based on the surface reflectance data of each pixel in the remote sensing image data, spectral mixture decomposition is used to obtain the endmember abundance values of all endmembers of each pixel in the area. The categories of the endmembers include vegetation, sand, salt and dark matter;

[0016] All end-member abundance values of each pixel are input into the pre-trained cover type discrimination model to obtain the cover type of the corresponding pixel;

[0017] (2) Calculate the cover function index according to the following formula:

[0018] ;

[0019] in, 、 and are the end-member abundance values of sand, salt, and dark matter in spring, respectively; is the end-member abundance value of vegetation in summer; LCI is the cover function index, ranging from 0 to 1;

[0020] (3) Calculate the vegetation production rate index of the corresponding pixel based on the surface temperature data of each pixel;

[0021] (4) Based on the elevation data of each pixel, calculate the terrain humidity index of the corresponding pixel.

[0022] Furthermore, the step of calculating the soil organic carbon content includes:

[0023] A number of pixels are selected from the remote sensing image data as samples, and the soil organic carbon content of each pixel in the sample is measured using laboratory analysis methods;

[0024] All pixels in the sample were randomly divided into training and validation sets according to the proportion. The measured soil organic carbon content and environmental variable data of each pixel in the training set were input into the random forest model. The grid search method was used to determine the optimal parameters of the random forest model for the number of trees, maximum depth, and maximum number of features based on the results of 5-fold cross-validation.

[0025] The determined optimal parameters were used as the setting parameters of the random forest model to obtain the optimal random forest prediction model for soil organic carbon content in the region;

[0026] The environmental variable data corresponding to each pixel is input into the optimal random forest prediction model and run several times. The mean of the results of the several runs is determined as the soil organic carbon content of the corresponding pixel. The mean map of the predicted soil organic carbon content is output, and the standard deviation of the results of the multiple runs is determined as the uncertainty value of the soil organic carbon content.

[0027] The measured values of soil organic carbon content and environmental variable data of each pixel in the validation set were input into the optimal random forest prediction model to obtain the predicted values of soil organic carbon content of each pixel in the validation set. Based on the measured and predicted values of each pixel in the validation set, the accuracy of the mean map drawing results was evaluated using the coefficient of determination and root mean square error.

[0028] Furthermore, the self-organizing map network model is used to classify all pixels in the region into several different soil organic carbon status types, specifically including:

[0029] The scale of the self-organizing map network was set, and the cover type, cover function index, vegetation production rate, terrain moisture index and soil organic carbon content of all pixels in the region were input into the self-organizing map network model to obtain several different soil organic carbon state types;

[0030] In setting the scale of the self-organizing map network, first, an initial scale value of the self-organizing map network is set according to an empirical formula, and the initial scale value is five times the square root of the total number of regional pixels; then, a number of quantitative values close to the initial scale value are selected to obtain a set of quantitative values to be determined; and for any quantitative value in the set of quantitative values to be determined, a quantitative error and a structural error are calculated respectively, and the scale of the self-organizing map model is determined according to the calculation results.

[0031] Furthermore, the degradation path curve and the recovery path curve are fitted using quadratic functions in the constructed state space to identify the nonlinear evolution path of soil organic carbon, specifically including:

[0032] The average cover function index and average soil organic carbon content of all pixels in each soil organic carbon state type were used as fitting sample points, and quadratic functions were used to fit the degradation path curve and the recovery path curve respectively.

[0033] According to the S-type conceptual model, the degradation path curve and the recovery path curve are connected by a dotted line to form a complete S-type evolution path.

[0034] Furthermore, the nonlinear evolution path of soil organic carbon is used to determine the mutation threshold, hysteresis and toughness of the soil organic carbon evolution path, specifically including:

[0035] In the nonlinear evolution path of soil organic carbon, the portion with the same cover function index in the degradation path curve and the recovery path curve is defined as an unstable state, the portion with a cover function index less than the unstable state in the degradation path curve is defined as a good stable state, and the portion with a cover function index greater than the unstable state in the recovery path curve is defined as a poor stable state. The good mutation threshold and the poor mutation threshold are determined based on the separation value of the cover function index between the two stable states and the unstable state.

[0036] The hysteresis of soil organic carbon evolution pathway was determined based on the range between two mutation thresholds;

[0037] The resilience of the soil organic carbon evolution pathway was assessed based on the degree of change in the cover function index required to convert any soil organic carbon state type in an unstable state to another achievable stable state soil organic carbon state type.

[0038] Furthermore, the calculation method of the good mutation threshold includes:

[0039] ;

[0040] Among them, S1 is the set of pixels corresponding to the soil organic carbon state type with the largest average cover function index in the good stable state, LCI(x) is the cover function index corresponding to pixel x, and LCI1 is the pixel with the largest cover function index in the good stable state, that is, the largest LCI(x);

[0041] ;

[0042] Among them, S2 is the set of pixels corresponding to the soil organic carbon state type with the smallest average cover function index in the unstable state on the recovery path curve, LCI(x) is the cover function index corresponding to pixel x, and LCI2 is the pixel with the smallest cover function index in the unstable state, that is, the smallest LCI(x);

[0043] ;

[0044] Among them, LCIa is the good mutation threshold, μ1 and μ2 are the average values of the cover function index of the pixels corresponding to S1 and S2, σ1 and σ2 are the standard deviations of the cover function index of the pixels corresponding to S1 and S2, n∈(0.1, 0.2, 0.3, …, 2), and the value of n should meet the following conditions:

[0045] .

[0046] Furthermore, the calculation method of the mutation threshold of the difference includes:

[0047] ;

[0048] Among them, S3 is the set of pixels corresponding to the soil organic carbon state type with the largest average cover function index in the unstable state on the degradation path, LCI(x) is the cover function index corresponding to pixel x, and LCI3 is the pixel with the largest cover function index in the unstable state, that is, the largest LCI(x);

[0049] ;

[0050] Among them, S4 is the set of pixels corresponding to the soil organic carbon state type with the smallest average cover function index in the poor stable state, LCI(x) is the cover function index corresponding to pixel x, and LCI4 is the pixel with the smallest cover function index in the poor stable state, that is, the smallest LCI(x).

[0051] ;

[0052] Where LCIb is the mutation threshold of the difference, μ3 and μ4 are the average values of the cover function index of the pixels corresponding to state types S3 and S4, σ3 and σ4 are the standard deviations of the cover function index of the pixels corresponding to state types S3 and S4, n∈(0.1, 0.2, 0.3, …, 2), and the value of n should meet the following conditions:

[0053] .

[0054] Furthermore, the resilience of the soil organic carbon evolution path is assessed based on the degree of change in the cover function index required for the conversion of any soil organic carbon state type in an unstable state to another soil organic carbon state type in an achievable stable state, specifically including:

[0055] All soil organic carbon state types in the good or poor stable state to which any soil organic carbon state type in the unstable state can recover or degenerate are identified as candidate states;

[0056] Among the candidate states, screening is performed based on the principle of consistent cover type, and a set of soil organic carbon state types with the same cover type as the current soil organic carbon state type is screened out;

[0057] According to the principle of the closest topographic wetness index, the soil organic carbon state type in the achievable stable state corresponding to the current soil organic carbon state type is determined in the soil organic carbon state type set;

[0058] The absolute value of the difference between the current soil organic carbon state type and the achievable soil organic carbon state type's cover function index is calculated, which is the resilience of the current soil organic carbon state type to recover or degenerate to a good or poor stable state.

[0059] Another object of the present invention is to provide an electronic device comprising a processor and a memory storing a computer program, wherein when the processor executes the computer program, the nonlinear evolution path and characteristic evaluation method of regional soil organic carbon provided by the first object of the present invention is realized.

[0060] Another object of the present invention is to provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the regional soil organic carbon nonlinear evolution path and characteristic evaluation method provided by the first object of the present invention.

[0061] In combination with the above technical solutions, the present invention has the following beneficial effects compared with the prior art:

[0062] Based on the principle of exchanging space for time, the present invention uses soil organic carbon data from different spaces at a certain time to reconstruct the state evolution path of soil organic carbon in the same space when constructing the state space. This effectively solves the problem of being unable to construct the evolution path due to the lack of long-term measured data on soil organic carbon (especially historical data).

[0063] Based on limited field-measured sample data and multi-source remote sensing data, this invention enables rapid and highly accurate mapping of soil organic carbon and its environmental variables. This allows for accurate assessment of soil organic carbon and its dynamic changes, providing fundamental data support for carbon neutrality. Furthermore, this invention can accurately identify the nonlinear evolutionary pathways of soil organic carbon under land use and management, and quantify its nonlinear characteristics, such as mutation thresholds, hysteresis, and resilience, effectively addressing the current lack of trend prediction and early warning methods for soil organic carbon management. According to the evolution path of soil organic carbon, the trend of soil organic carbon can be predicted, and the mutation threshold can further provide early warning information of soil organic carbon, identify early warning areas and avoid the huge restoration costs caused by mutations. The resilience assessment results determine the priority for soil organic carbon management and provide decision-making support for the management of soil organic carbon in my country. In addition, the data source used in the present invention is mainly remote sensing data. On the one hand, the spatial and temporal resolution of remote sensing data is high, which can greatly improve the accuracy and update frequency of monitoring results. At the same time, the remote sensing data used in the present invention are all open and free data products, which can greatly reduce the manpower and material costs consumed by field sampling. On the other hand, remote sensing data can cover areas that are difficult to reach including traditional field surveys, thereby achieving full coverage of the region.

[0064] Based on limited field-measured sample data and multi-source remote sensing data, this method accurately identifies the nonlinear evolutionary pathways of soil organic carbon under land use and management, and quantifies its nonlinear characteristics, such as mutation thresholds, hysteresis, and resilience, to enable trend prediction and early warning of soil organic carbon, thus supporting soil organic carbon management. Furthermore, the model constructed in this invention exhibits good stability. After evaluating the evolutionary pathways and their characteristics, the status of soil organic carbon can be assessed solely based on multi-source remote sensing data, significantly reducing the cost of field sampling of soil organic carbon. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0066] Figure 1 This is a flow chart of a method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon provided by an embodiment of the present invention;

[0067] Figure 2This is a schematic diagram of end member extraction based on geometric vertices provided by an embodiment of the present invention

[0068] Figure 3 Schematic diagram of a self-organizing map network model provided by an embodiment of the present invention;

[0069] Figure 4 This is a schematic diagram of the soil organic carbon evolution path reconstruction method provided by an embodiment of the present invention;

[0070] Figure 5 is a schematic diagram of a state space provided by an embodiment of the present invention;

[0071] Figure 6 is a schematic diagram of the S conceptual model provided by an embodiment of the present invention;

[0072] Figure 7 This is a schematic diagram of a state evolution path provided by an embodiment of the present invention;

[0073] Figure 8 This is a schematic diagram of state type identification provided by an embodiment of the present invention;

[0074] Figure 9 2 is a schematic diagram of state nonlinear characteristic evaluation provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0075] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0076] Example 1:

[0077] like Figure 1 FIG. 1 is an embodiment of the method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon provided by the present invention, which specifically includes the following steps:

[0078] S1: Acquire multi-source remote sensing image data of a target area, wherein each pixel in the multi-source remote sensing image data represents surface reflectance data, surface temperature data, and elevation data at a corresponding position;

[0079] S2: Using a machine learning model, based on the surface reflectance data, surface temperature data, and elevation data of each pixel in the multi-source remote sensing image data, calculate the soil organic carbon content and its environmental variable data of each pixel, wherein the environmental variable data includes cover type, cover function index, vegetation production rate, and terrain moisture index;

[0080] S3: Using the self-organizing map network model, the soil organic carbon content of each pixel and environmental variable data are used as input vectors to classify all pixels in the area into several different soil organic carbon status types;

[0081] S4: Construct a two-dimensional state space based on the average cover function index and average soil organic carbon content of each soil organic carbon state type. Use quadratic functions to fit the degradation path curve and the recovery path curve in the constructed state space to identify the nonlinear evolution path of soil organic carbon;

[0082] S5: Based on the nonlinear evolution path of soil organic carbon, determine the mutation threshold, hysteresis and resilience of the soil organic carbon evolution path.

[0083] Preferably, the step of calculating the environmental variable data in the embodiment of the present invention includes:

[0084] (1) Based on the surface reflectance data of each pixel in the remote sensing image data, spectral mixture decomposition is used to obtain the endmember abundance values of all endmembers of each pixel in the area. The categories of the endmembers include vegetation, sand, salt and dark matter; all the endmember abundance values of each pixel are input into the pre-trained cover type discrimination model to obtain the cover type of the corresponding pixel;

[0085] Specifically, according to the land cover classification system (such as forest land, grassland, cultivated land, construction land and water types), about 1000 samples were selected for each land cover category using high spatial resolution images of Google Earth, and divided into training samples and validation samples at a ratio of 75% and 25%;

[0086] The training samples and their corresponding endmember abundance values and all endmember abundance value images obtained by spectral mixture decomposition are input into the random forest model for cover type classification;

[0087] The accuracy evaluation is performed based on the coverage type of the verification sample and the corresponding classification results. The evaluation indicators include the overall accuracy p e And the Kappa coefficient k, calculated as follows:

[0088]

[0089] ;

[0090] in, It is the sum of the number of correctly classified samples of each land cover category divided by the total number of samples, that is, the overall classification accuracy. 、 … is the number of true samples of each land cover category, and 、 … is the number of samples for each land cover category in the prediction results, and the total number of samples is n.

[0091] It should be noted that spectral mixture decomposition mainly includes three steps: principal component analysis, endmember extraction based on geometric vertices, and calculation of the endmember abundance value of each pixel in the study area.

[0092] First, based on local knowledge within the study area, we determined the months with the highest vegetation cover (e.g., August) and the lowest surface cover (e.g., May). We then performed principal component analysis (PCA) on the Landsat surface reflectance data for these months using ENVI software to obtain transformed PCA images. The type and number of endmembers were determined based on existing regional research. In arid regions, monitoring typically involves four endmembers: vegetation (GV), sand (SL), salt (SA), and dark matter (DA). The vegetation (GV) endmember was extracted in August, while the sand (SL), salt (SA), and dark matter (DA) endmembers were extracted in May.

[0093] Then, if Figure 2 As shown in the figure, a two-dimensional scatter plot extraction method is used to extract the endmember spectral curve. Based on the mathematical mechanism of the linear mixed model, the scatter plot of the principal component should be triangular in distribution. The three vertices of the triangle can be regarded as pure endmembers, and the interior of the triangle is a mixed pixel composed of linear combinations of pure endmembers. In ENVI software, a scatter plot is constructed using the principal component image of the extraction month corresponding to the endmember. 200-400 pixels are selected as endmember pixels at the vertices of the principal component scatter plot, and the average reflectance of the original image at the location of the endmember pixel is used as the endmember spectrum.

[0094] Finally, after obtaining the endmember spectra in the region, the endmember spectra and regional reflectance images were used to perform linear spectral mixing decomposition in ENVI software to obtain the endmember abundance values of each endmember in the study area.

[0095] Linear spectral mixture decomposition is the main method for abundance estimation, and its expression is:

[0096] ;

[0097] Among them, i is the remote sensing image band, is the reflectivity of band i, m is the number of end members, for each end member j (1≤j≤m), is the reflectivity of end member j in band i, is the abundance value of end member j, that is, the percentage of end member in pixel area, In practice, in order to make the estimated abundance value have physical meaning, the fully constrained least squares method is generally used to obtain the abundance value of each end member in each pixel and obtain the corresponding abundance value image.

[0098] .

[0099] (2) Calculate the cover function index according to the following formula:

[0100] ;

[0101] in, 、 and are the end-member abundance values of sand, salt, and dark matter in spring, respectively; It is the end-member abundance value of vegetation in summer; LCI is the cover function index, which ranges from 0 to 1.

[0102] (3) The vegetation production rate index of the corresponding pixel is calculated based on the surface temperature data of each pixel. The calculation formula is as follows:

[0103] ;

[0104] Among them, VPI represents the vegetation production rate index, is a function of the reference surface temperature curve, is a function of the pixel surface temperature curve, and These are the start and end of the main growing season, respectively;

[0105] (4) According to the elevation data of each pixel, the terrain humidity index of the corresponding pixel is calculated. The calculation formula is as follows:

[0106] ;

[0107] Where TWI stands for Terrain Wetness Index, It is the upstream area on the unit contour line through which surface water flows. For the slope.

[0108] Preferably, the step of calculating the soil organic carbon content in the embodiment of the present invention includes:

[0109] (1) Select a certain number of pixels from the remote sensing image data as samples and use laboratory analysis methods to measure the soil organic carbon content of each pixel in the sample;

[0110] (2) All pixels in the sample are randomly divided into a training data set and a validation data set at a ratio of 75% and 25%. In the training data set, the grid search method is used to determine the optimal parameter combination of the number of trees, maximum depth and maximum number of features of the random forest model based on the results of 5-fold cross-validation. The determined optimal parameter combination is used as the setting parameters of the random forest model. The specific operations are as follows:

[0111] 1) Based on experience, the number of trees in the random forest model is preset to 100 to 1000, with an interval of 50; the maximum depth is 4 to 10, with an interval of 1; the maximum number of features is 2 to 5, with an interval of 1;

[0112] 2) Grid search method, which combines all possible values of the three parameters within a preset range;

[0113] 3) Then perform 5-fold cross validation for each combination of parameter values: divide the training data set into 5 samples, use each sample as a validation sample, and the remaining four samples as training samples, and calculate the average determination coefficient R of the 5 times 2 and root mean square error RMSE, and the highest value is selected as the optimal model parameter.

[0114]

[0115] ;

[0116] in, and are the predicted and measured values of soil organic carbon content in pixel i, is the average value of the measured soil organic carbon content in all pixels, and n is the number of pixels.

[0117] Using the random forest model of the "rpart" package in R language, the measured values of soil organic carbon content in each pixel in the training set and environmental variable data are input. The optimal model parameters are selected as the setting parameters of the random forest model, and the optimal random forest prediction model of soil organic carbon content in each pixel in the study area can be obtained.

[0118] (3) Next, the environmental variable dataset was input into the optimal random forest prediction model, and the mean of the results of running the optimal random forest prediction model 100 times was used as the final prediction result. The soil organic carbon mean map was output and the determination coefficient R was used. 2 The root mean square error (RMSE) was used to evaluate the accuracy of the mapping results, and the standard deviation of the 100 prediction values was used to evaluate the uncertainty of the model.

[0119] Preferably, the self-organizing map network model is used in the embodiment of the present invention to divide all pixels in the region into several different soil organic carbon state types, specifically including:

[0120] The scale of the self-organizing map network was set, and the cover type, cover function index, vegetation production rate, terrain moisture index and soil organic carbon content of all pixels in the area were input into the self-organizing map network model to obtain several different soil organic carbon state types.

[0121] In setting the scale of the self-organizing map network, first, an initial scale value of the self-organizing map network is set according to an empirical formula, and the initial scale value is five times the square root of the total number of regional pixels; then, a number of quantitative values close to the initial scale value are selected to obtain a set of quantitative values to be determined; and for any quantitative value in the set of quantitative values to be determined, a quantitative error and a structural error are calculated respectively, and the scale of the self-organizing map model is determined according to the calculation results.

[0122] Specifically, such as Figure 3 As shown. In the self-organizing map network, first according to Vesanto's empirical formula, use Estimate the initial size of the self-organizing map network:

[0123] ;

[0124] Where M is the size of the self-organizing map network, and S is the number of samples;

[0125] Assuming the number of samples is 253, the initial scale is 80. Several scale values are then selected to the left and right of the initial scale value to obtain the set of undetermined scale values. Since the self-organizing map network is a two-dimensional network, based on experience, three scale combinations are set to the left and right of the initial scale of 80. Therefore, the scales are preset as 10×7, 9×8, 11×7, 10×8, 11×8, 10×9, and 12×8. Finally, for each scale value in the set of undetermined scale values, the quantitative error (QE) and structural error (TE) are calculated for each scale. The number of scales (i.e., the number of soil organic carbon state types) in the self-organizing map model is determined based on the calculated results.

[0126] ;

[0127] Among them, Num is the total number of samples, is the i-th input sample, It is a sample The corresponding weight vector of the best matching neuron, is the Euclidean distance between the two;

[0128] ;

[0129] Among them, Num is the total number of samples, and The samples are The first neighboring neuron and the second neighboring neuron, when and When not adjacent, Returns 1 if yes, otherwise returns 0.

[0130] In the embodiment of the present invention, in constructing a two-dimensional state space based on the average cover function index and average soil organic carbon content of each soil organic carbon state type, the construction of the state space is mainly based on the method of exchanging space for time. It is assumed that there are enough soil organic carbon state types in the space and the evolution time of the soil organic carbon state types is long enough. The set of soil organic carbon state types at different spatial positions at the same time and the set of soil organic carbon state types at the same spatial position at different times are equivalent. Figure 4 As shown in Figure 2, this principle is like reconstructing a collection of plants at different growth stages at the same time into a collection of different growth stages of the same plant. Therefore, the state evolution path of the soil organic carbon state type at different times in the same space can be reconstructed using the soil organic carbon state types at different times in the same space.

[0131] According to the determined soil organic carbon state type, the two-dimensional state space of soil organic carbon is constructed with the average cover function index of all pixels in each soil organic carbon state type as the X-axis and the average soil organic carbon content of all pixels in each soil organic carbon state type as the Y-axis. Figure 5 As shown, the colored points represent the corresponding positions of each soil organic carbon state type in the state space.

[0132] Preferably, in the embodiment of the present invention, the specific steps of fitting the degradation path curve and the recovery path curve respectively using a quadratic function in the constructed state space to identify the nonlinear evolution path of soil organic carbon include:

[0133] Based on the S-shaped conceptual model, quadratic functions were used to fit the curves of the degradation and recovery paths in the state space. The quadratic function was fitted using the polyfit function in the Python numpy library. The average cover functional index and average soil organic carbon content of all pixels in each soil organic carbon state type obtained from the self-organizing map network were used as fitting sample points. The function form was set as follows:

[0134] ;

[0135] Where y is the soil organic carbon content, x is the cover function index, and a, b, and c are coefficients. The polyfit function can use the least squares method to solve the coefficients.

[0136] It should be noted that when external environmental conditions change, the system state of the land will respond accordingly and present a bistable state. Figure 6The solid line in the figure represents the steady state, while green and red represent the best and worst stable states under the corresponding environmental conditions, respectively. The dashed line represents an unstable non-equilibrium state, and orange represents multiple unstable states under the same environmental conditions. The upper half of the S-shaped curve represents the degradation path. As external environmental conditions gradually degrade to a critical level, the system will suddenly change from one stable state to another through a threshold, a process known as a steady-state transition. The points where this transition occurs are called critical points (P1 and P2). However, this positive steady-state transition path is not completely reversible; the land system exhibits significant hysteresis: even if external environmental conditions return to their pre-transition levels, the land system cannot recover to its previous state.

[0137] like Figure 7 As shown, the quadratic function curve fitted to the green points in the upper half of the state space is defined as the degradation path. As the cover function index increases (the worse the natural environment and human management), the state of soil organic carbon moves to the right along the green solid line and degrades (i.e., the soil organic carbon content gradually decreases). And as the cover function index becomes larger and larger (the natural environment and human management become worse and worse), the degradation of the organic carbon state (soil organic carbon content) caused by the increase in the unit cover function index gradually increases. Until a certain threshold is reached and a mutation occurs, i.e. Figure 7 The path is the green dashed line.

[0138] The same approach was used to identify soil organic carbon recovery pathways. Figure 7 As shown, the quadratic function curve fitted to the orange points in the lower half of the state space is defined as the recovery path. As the cover function index decreases (the better the natural environment and human management), the state of soil organic carbon moves to the left along the orange solid line and recovers (i.e., the soil organic carbon content gradually increases). As the cover function index becomes smaller and smaller (the better the natural environment and human management), the recovery of the organic carbon state (soil organic carbon content) caused by the reduction of the unit cover function index gradually increases, until it reaches a certain threshold and a mutation occurs, i.e. Figure 7 The path is the orange dashed line.

[0139] like Figure 8 As shown in Figure 2, based on the S-type conceptual model, the degradation path curve and the recovery path curve are connected by a dotted line to form a complete S-type evolution path.

[0140] Preferably, in an embodiment of the present invention, based on the nonlinear evolution path of soil organic carbon, the specific steps of determining the mutation threshold, hysteresis and toughness of the soil organic carbon evolution path include:

[0141] In the nonlinear evolution path of soil organic carbon, the portion with the same cover function index in the degradation path curve and the recovery path curve is defined as an unstable state, the portion with a cover function index less than the unstable state in the degradation path curve is defined as a good stable state, and the portion with a cover function index greater than the unstable state in the recovery path curve is defined as a poor stable state. The good mutation threshold and the poor mutation threshold are determined based on the separation value of the cover function index between the two stable states and the unstable state.

[0142] The hysteresis of soil organic carbon evolution path was determined based on the range between two mutation thresholds;

[0143] The resilience of the soil organic carbon evolution pathway was assessed based on the degree of change in the cover function index required to convert any soil organic carbon state type in an unstable state to another achievable stable state soil organic carbon state type.

[0144] Specifically, such as Figure 8 As shown, in the S-shaped pathway, the soil organic carbon state types located along the orange path within the folded interval are defined as unstable. Within this interval, the same land cover index can correspond to multiple soil organic carbon state types with different soil organic carbon contents. When the natural environment or human management undergoes slight changes, the soil organic carbon state types will shift along the pathway. The states along the dotted paths are non-equilibrium states. These states exist only transiently at the moment of observation and will eventually evolve to the equilibrium state along the solid path. Therefore, these non-equilibrium states are not included in the evolutionary pathway. On the left side of the folded interval of the S-shaped pathway, the soil organic carbon state types along the green path represent a good stable state, while on the right side of the folded interval, the soil organic carbon state types along the red path represent a poor stable state. Compared to unstable states, these stable soil organic carbon state types can withstand minor changes in the natural environment or human management through their inherent resilience and remain stable.

[0145] The evaluation of nonlinear characteristics includes quantitative characterization of mutation threshold, hysteresis and toughness. Figure 9 As shown in the figure, the mutation threshold means that as the cover function index (natural environment or human management) changes, the state of soil organic carbon will gradually degrade or recover to a critical level, and mutate to another stable state. The cover function index value corresponding to this mutation point is called the mutation threshold. The mutation threshold includes good mutation threshold ( Figure 9 LCIa) and the difference mutation threshold ( Figure 9 These thresholds are determined by the coverage function index split between the two stable and unstable states.

[0146] A good calculation method for the mutation threshold includes:

[0147] ;

[0148] Among them, S1 is the set of pixels corresponding to the soil organic carbon state type with the largest average cover function index in a good stable state, LCI(x) is the cover function index corresponding to pixel x, and LCI1 is the largest LCI(x);

[0149] ;

[0150] Among them, S2 is the set of pixels corresponding to the soil organic carbon state type with the smallest average cover function index in the unstable state on the recovery path curve, LCI(x) is the cover function index corresponding to pixel x, and LCI2 is the smallest LCI(x);

[0151] ;

[0152] Among them, LCIa is the good mutation threshold, μ1 and μ2 are the average values of the cover function index of the pixels corresponding to S1 and S2, σ1 and σ2 are the standard deviations of the cover function index of the pixels corresponding to S1 and S2, n∈(0.1, 0.2, 0.3, …, 2), and the value of n should meet the following conditions:

[0153] .

[0154] The calculation method of the difference mutation threshold includes:

[0155] ;

[0156] Among them, S3 is the set of pixels corresponding to the soil organic carbon state type with the largest average cover function index in the unstable state on the degradation path, LCI(x) is the cover function index corresponding to pixel x, and LCI3 is the largest LCI(x);

[0157] ;

[0158] Among them, S4 is the set of pixels corresponding to the soil organic carbon state type with the smallest average cover function index in the poor stable state, LCI(x) is the cover function index corresponding to pixel x, and LCI4 is the smallest LCI(x).

[0159] ;

[0160] Where LCIb is the mutation threshold of the difference, μ3 and μ4 are the average values of the cover function index of the pixels corresponding to state types S3 and S4, σ3 and σ4 are the standard deviations of the cover function index of the pixels corresponding to state types S3 and S4, n∈(0.1, 0.2, 0.3, …, 2), and the value of n should meet the following conditions:

[0161] .

[0162] Hysteresis means that the degradation path of soil organic carbon state is irreversible. When the cover function index (natural environment or human management) returns to the level before the mutation, the state of soil organic carbon cannot return to its original state. The level of hysteresis depends on the range between the two mutation thresholds, i.e. Figure 9 The absolute value of the difference between LCIa and LCIb in the soil organic carbon index. The larger the range, the stronger the hysteresis, and the more difficult it is to recover to the previous state after a sudden change in soil organic carbon.

[0163] Resilience refers to the degree of change in the cover function index required for the current state of soil organic carbon to transform into another achievable stable state. The achievable state corresponding to a certain state of soil organic carbon should not only be affected by the evolutionary path, but also by the influence of its cover type and terrain moisture index. Figure 9 As shown in the figure, taking the calculation of soil organic carbon state type B to restore to the good stable state soil organic carbon state type A1 as an example: (1) First, all the good stable states that soil organic carbon state type B may restore to are determined as candidate states; (2) Among the candidate states, the soil organic carbon state type set with the same cover type as soil organic carbon state type B is screened out according to the principle of consistent cover type; (3) In the screened soil organic carbon state type set, the soil organic carbon state type A1 in the achievable stable state corresponding to soil organic carbon state type B is determined according to the principle of closest terrain moisture index; (4) The absolute value of the difference between the cover function index of soil organic carbon state type B and soil organic carbon state type A1 is calculated, which is the resilience of soil organic carbon state type B to restore to the good stable state. The calculation method for calculating the resilience of the state degenerating to the poor stable state C1 is the same as the above method.

[0164] Example 2:

[0165] An embodiment of the present invention provides an electronic device, including a processor and a memory storing a computer program. When the processor executes the computer program, the regional soil organic carbon nonlinear evolution path and characteristic evaluation method provided in Example 1 of the present invention is implemented.

[0166] Example 3:

[0167] Another object of the present invention is to provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the regional soil organic carbon nonlinear evolution path and characteristic evaluation method provided in Example 1 of the present invention.

[0168] It should be understood that, although the various steps in the flow charts of the various embodiments of the present invention are shown in sequence according to the instructions of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless otherwise specified in the present invention, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the various embodiments may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0169] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When executed, the program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media used in the various embodiments provided herein may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0170] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon, characterized in that: The method comprises: Acquiring multi-source remote sensing image data of a target area, wherein each pixel in the multi-source remote sensing image data represents surface reflectance data, surface temperature data, and elevation data at a corresponding location; Using a machine learning model, based on the surface reflectance data, surface temperature data, and elevation data of each pixel in the multi-source remote sensing image data, the soil organic carbon content and its environmental variable data of each pixel are calculated; Using the self-organizing map network model, the soil organic carbon content of each pixel and environmental variable data were used as input vectors to classify all pixels in the area into several different soil organic carbon status types. A two-dimensional state space was constructed based on the average cover function index and average soil organic carbon content of each soil organic carbon state type. A quadratic function was used to fit the degradation path curve and the recovery path curve in the constructed state space to identify the nonlinear evolution path of soil organic carbon. Based on the nonlinear evolution path of soil organic carbon, the mutation threshold, hysteresis and resilience of the soil organic carbon evolution path are determined. Specifically, in the nonlinear evolution path of soil organic carbon, the parts with the same cover function index in the degradation path curve and the recovery path curve are defined as unstable states, the parts with a cover function index less than the unstable state in the degradation path curve are defined as good stable states, and the parts with a cover function index greater than the unstable state in the recovery path curve are defined as poor stable states. The good mutation threshold and the poor mutation threshold are determined according to the segmentation value of the cover function index between the two stable states and the unstable state; the hysteresis of the soil organic carbon evolution path is determined according to the range between the two mutation thresholds; the resilience of the soil organic carbon evolution path is evaluated according to the degree of change in the cover function index required for any soil organic carbon state type in the unstable state to be converted to another achievable soil organic carbon state type in the stable state.

2. The method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon according to claim 1, characterized in that: The step of calculating the environmental variable data includes: (1) Based on the surface reflectance data of each pixel in the remote sensing image data, spectral mixture decomposition is used to obtain the endmember abundance values of all endmembers of each pixel in the area. The categories of the endmembers include vegetation, sand, salt and dark matter; All end-member abundance values of each pixel are input into the pre-trained cover type discrimination model to obtain the cover type of the corresponding pixel; (2) Calculate the cover function index according to the following formula: ； in, 、 and are the end-member abundance values of sand, salt, and dark matter in spring, respectively; is the end-member abundance value of vegetation in summer; LCI is the cover function index, ranging from 0 to 1; (3) Calculate the vegetation production rate index of the corresponding pixel based on the surface temperature data of each pixel; (4) Based on the elevation data of each pixel, calculate the terrain humidity index of the corresponding pixel.

3. The method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon according to claim 1, wherein: The calculation steps of the soil organic carbon content include: A number of pixels are selected from the remote sensing image data as samples, and the soil organic carbon content of each pixel in the sample is measured using laboratory analysis methods; All pixels in the sample were randomly divided into training and validation sets according to the proportion. The measured soil organic carbon content and environmental variable data of each pixel in the training set were input into the random forest model. The grid search method was used to determine the optimal parameters of the random forest model for the number of trees, maximum depth, and maximum number of features based on the results of 5-fold cross-validation. The determined optimal parameters were used as the setting parameters of the random forest model to obtain the optimal random forest prediction model for soil organic carbon content in the region; The environmental variable data corresponding to each pixel is input into the optimal random forest prediction model and run several times. The mean of the results of the several runs is determined as the soil organic carbon content of the corresponding pixel. The mean map of the predicted soil organic carbon content is output, and the standard deviation of the results of the multiple runs is determined as the uncertainty value of the soil organic carbon content. The measured values of soil organic carbon content and environmental variable data of each pixel in the validation set were input into the optimal random forest prediction model to obtain the predicted values of soil organic carbon content of each pixel in the validation set. Based on the measured and predicted values of each pixel in the validation set, the accuracy of the mean map drawing results was evaluated using the coefficient of determination and root mean square error.

4. The method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon according to claim 1, wherein: The self-organizing map network model is used to classify all pixels in the region into several different soil organic carbon status types, including: The scale of the self-organizing map network was set, and the cover type, cover function index, vegetation production rate, terrain moisture index and soil organic carbon content of all pixels in the region were input into the self-organizing map network model to obtain several different soil organic carbon state types; In setting the scale of the self-organizing map network, first, an initial scale value of the self-organizing map network is set according to an empirical formula, and the initial scale value is five times the square root of the total number of regional pixels; then, several quantitative values close to the initial scale value are selected to obtain a set of quantitative values to be determined; and for any quantitative value in the set of quantitative values to be determined, a quantitative error and a structural error are respectively calculated, and the scale of the self-organizing map model is determined according to the calculation results.

5. The method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon according to claim 1, wherein: The method of fitting the degradation path curve and the recovery path curve using a quadratic function in the constructed state space to identify the nonlinear evolution path of soil organic carbon specifically includes: The average cover function index and average soil organic carbon content of all pixels in each soil organic carbon state type were used as fitting sample points, and quadratic functions were used to fit the degradation path curve and the recovery path curve respectively. According to the S-type conceptual model, the degradation path curve and the recovery path curve are connected by a dotted line to form a complete S-type evolution path.

6. The method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon according to claim 1, wherein: The calculation method of the good mutation threshold includes: ； Among them, S1 is the set of pixels corresponding to the soil organic carbon state type with the largest average cover function index in a good stable state, LCI(x) is the cover function index corresponding to pixel x, and LCI1 is the largest LCI(x); ； Among them, S2 is the set of pixels corresponding to the soil organic carbon state type with the smallest average cover function index in the unstable state on the recovery path curve, LCI(x) is the cover function index corresponding to pixel x, and LCI2 is the smallest LCI(x); ； Among them, LCIa is the good mutation threshold, μ1 and μ2 are the average values of the cover function index of the pixels corresponding to S1 and S2, σ1 and σ2 are the standard deviations of the cover function index of the pixels corresponding to S1 and S2, n∈(0.1, 0.2, 0.3, …, 2), and the value of n should meet the following conditions: 。 7. The method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon according to claim 1, wherein: The calculation method of the mutation threshold of the difference includes: ； Among them, S3 is the set of pixels corresponding to the soil organic carbon state type with the largest average cover function index in the unstable state on the degradation path, LCI(x) is the cover function index corresponding to pixel x, and LCI3 is the pixel with the largest cover function index in the unstable state, that is, the largest LCI(x); ； Among them, S4 is the set of pixels corresponding to the soil organic carbon state type with the smallest average cover function index in the poor stable state, LCI(x) is the cover function index corresponding to pixel x, and LCI4 is the pixel with the smallest cover function index in the poor stable state, that is, the smallest LCI(x); ； Where LCIb is the mutation threshold of the difference, μ3 and μ4 are the average values of the cover function index of the pixels corresponding to state types S3 and S4, σ3 and σ4 are the standard deviations of the cover function index of the pixels corresponding to state types S3 and S4, n∈(0.1, 0.2, 0.3, …, 2), and the value of n should meet the following conditions: 。 8. The method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon according to claim 1, wherein: The resilience of the soil organic carbon evolution path is assessed based on the degree of change in the cover function index required for the conversion of any soil organic carbon state type in an unstable state to another soil organic carbon state type in an achievable stable state, specifically including: All soil organic carbon state types in the good or poor stable state to which any soil organic carbon state type in the unstable state can recover or degenerate are identified as candidate states; Among the candidate states, screening is performed based on the principle of consistent cover type, and a set of soil organic carbon state types with the same cover type as the current soil organic carbon state type is screened out; Determining, from the soil organic carbon state type set, a soil organic carbon state type in an achievable stable state corresponding to the current soil organic carbon state type; The absolute value of the difference between the current soil organic carbon state type and the achievable soil organic carbon state type's cover function index is calculated, which is the resilience of the current soil organic carbon state type to recover or degenerate to a good or poor stable state.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for evaluating the nonlinear evolution path and characteristics of regional soil organic carbon as described in any one of claims 1 to 8 is implemented.

Images