A network analysis-based cultivated soil health evaluation method and system

Abstract

This invention discloses a method and system for evaluating arable land soil health based on network analysis, relating to the field of arable land soil health evaluation technology. The method includes acquiring multi-source data of arable land soil in a target area; constructing corresponding soil health evaluation indicators based on preset soil health functions, wherein the preset soil health functions include at least one of the following: water regulation function, function for maintaining plant and animal habitats and life activities, pollutant filtration and buffering function, nutrient cycling function, and physical stability and structural support function. This invention constructs corresponding soil health evaluation indicators guided by preset soil health functions, builds an indicator association network by calculating partial correlation coefficients, and determines the indicator weights of indicator nodes through the strength centrality of the indicator nodes. Objective weighting based on node strength centrality effectively analyzes the multidimensional nonlinear relationships between soil indicators, overcoming the subjectivity and linear limitations of traditional weight determination methods.

Description

Technical Field

[0001] This invention relates to the field of farmland soil health assessment technology, and in particular to a method and system for farmland soil health assessment based on network analysis. Background Technology

[0002] Arable land soil is the core carrier for ensuring stable grain production and maintaining regional terrestrial ecological balance. The health status of arable land soil directly affects the sustainability of agricultural production and regional ecological security. At present, the superposition of high-intensity farming, continuous input of industrial and agricultural pollutants, and frequent changes in land use patterns has led to multiple problems in arable land, such as shallow topsoil, compacted soil, nutrient imbalance, and degradation of microbial communities. Therefore, health assessment of arable land soil has become a key research direction in the field of agricultural resources and environment.

[0003] Currently, the main methods for assessing arable land soil health include the Cornell Assessment of Soil Health (CASH), the Highly Sensitive Soil Health Test (HSHT), the Modified Soil Quality Rating (M-SQR), and my country's arable land quality assessment methods. The weights of the indicators in these assessment methods are mainly determined through subjective weighting methods, such as the Analytic Hierarchy Process (AHP) and the Delphi method, or linear weighting methods, such as the entropy weight method.

[0004] However, the subjective weighting method relies on the prior knowledge of experts, and the weighting results are easily affected by subjective factors. The weights given by different evaluators vary greatly, and the method lacks objectivity and repeatability. The linear weighting method only assigns weights based on the data fluctuation characteristics of each indicator, which does not reflect the non-linear relationship between soil indicators. It severs the overall structural characteristics of the evaluation indicator system, and the weight allocation cannot reflect the actual influence and pivotal position of each indicator in the evaluation network. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for evaluating arable land soil health based on network analysis, which solves the problems of subjective weighting methods lacking objectivity and repeatability, and linear weighting methods failing to reflect the nonlinear correlation between soil indicators.

[0006] The technical means employed in this invention are as follows: In a first aspect, the present invention provides a method for evaluating the health of arable land soil based on network analysis, comprising: Acquire multi-source data on arable land soil in the target area; Based on preset soil health functions, corresponding soil health evaluation indicators are constructed. The preset soil health functions include at least one of the following: water regulation function, function of maintaining plant and animal habitat and life activities, pollutant filtration and buffering function, nutrient cycling function, and physical stability and structural support function. The partial correlation coefficients between the soil health evaluation indicators are calculated, and an indicator association network is constructed based on the partial correlation coefficients. Determine the edge weights between indicator nodes in the indicator association network, calculate the strength centrality of the indicator nodes based on the edge weights, and determine the indicator weights of the indicator nodes based on the strength centrality. The multi-source data is standardized to obtain standardized data; Based on the index weights and the standardized data, the soil health index of the multi-source data sampling points is calculated. Based on the soil health index, the natural breakpoint method is used to classify the soil health of cultivated land, and the soil health evaluation result is output based on the classification result of cultivated land soil health.

[0007] Furthermore, the soil health evaluation indicators corresponding to the water regulation function include at least one of topsoil texture, soil bulk density, and irrigation capacity; The soil health evaluation indicators corresponding to the maintenance of plant and animal habitats and life activities include at least one of soil pH, soil respiration, fungal Chao1 index and fungal Shannon index. The soil health evaluation indicators corresponding to the pollutant filtration and buffering functions include at least one of organic matter content, cation exchange capacity, and Nemerow comprehensive pollution index. The soil health evaluation indicators corresponding to the nutrient cycling function include at least one of soil organic carbon density, total nitrogen, total phosphorus, total potassium, available phosphorus, and available potassium. The soil health evaluation indicators corresponding to the physical stability and structural support function include at least one of the topsoil thickness and texture configuration.

[0008] Furthermore, the formula for calculating the intensity centrality is:

[0009] in, Indicates the strength centrality of index node i; This represents the edge weight between index node i and index node j; This represents the summation of edge weights among all index nodes j connected to index node i.

[0010] Furthermore, the edge weight is the absolute value of the partial correlation coefficient.

[0011] Furthermore, the index weight is the ratio of the intensity centrality of the corresponding index node to the sum of the intensity centralities of all index nodes.

[0012] Furthermore, the standardization process for the multi-source data includes performing various types of standardization processing on the multi-source data. The various types of standardization processing include at least one of the following: membership function standardization based on upper and lower thresholds, standardization threshold determination according to preset standards, and hierarchical assignment using the natural breakpoint method.

[0013] Furthermore, the formula for calculating the soil health index is as follows:

[0014] in, This represents the soil health index; n represents the total number of indicator nodes. (NA) represents the index weight of the i-th index node obtained based on network analysis; This represents the i-th standardized data value.

[0015] Furthermore, the method of classifying arable land soil health using the natural breakpoint method includes: The health of arable land soil was classified into healthy, sub-healthy, medium, fragile and unhealthy levels using the natural breakpoint method.

[0016] Secondly, the present invention also provides a farmland soil health assessment system based on network analysis, comprising: The multi-source data acquisition module is used to acquire multi-source data of farmland soil in the target area; The indicator network construction module is used to construct corresponding soil health evaluation indicators based on preset soil health functions. The preset soil health functions include at least one of the following: water regulation function, function of maintaining plant and animal habitat and life activities, pollutant filtration and buffering function, nutrient cycling function, and physical stability and structural support function; calculate the partial correlation coefficient between the soil health evaluation indicators, and construct an indicator association network based on the partial correlation coefficient. The weight calculation module is used to determine the edge weights between index nodes in the index association network, calculate the strength centrality of the index nodes based on the edge weights, and determine the index weights of the index nodes based on the strength centrality. The data standardization module is used to standardize the multi-source data to obtain standardized data. The index calculation module is used to calculate the soil health index of multi-source data sampling points based on the index weights and the standardized data. The grading output module is used to grade the soil health of cultivated land based on the soil health index using the natural breakpoint method, and output the soil health evaluation result based on the grading result of cultivated land soil health.

[0017] Compared with the prior art, the present invention has the following advantages: This invention constructs corresponding soil health evaluation indicators based on preset soil health functions, builds an indicator association network by calculating partial correlation coefficients, and determines the indicator weights of indicator nodes by strength centrality. Objective weighting based on node strength centrality can effectively analyze the multidimensional nonlinear associations between soil indicators, avoid the bias caused by subjective scoring, and overcome the subjectivity and linearity limitations of traditional weight determination methods. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a method for evaluating arable land soil health based on network analysis, as described in an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of the index association network in an embodiment of the present invention. Detailed Implementation

[0020] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0021] It should be noted that the terms "comprising" and "having" and any variations thereof in this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.

[0022] The embodiments of the present invention will now be described with reference to the accompanying drawings.

[0023] Please see Figure 1This invention provides a method for evaluating the health of arable land based on network analysis, comprising the following steps: Step 101: Obtain multi-source data of cultivated land soil in the target area.

[0024] Within the target area of ​​the study, sampling points were determined using a random stratification method based on factors such as soil type and land use patterns. At each sampling point, topsoil samples (0-20 cm depth) were collected, and their latitude and longitude coordinates and land use information were recorded. After air-drying and sieving, the collected topsoil samples were used to determine various evaluation indicators according to national or industry standards. This data serves as the multi-source data for the cultivated land soil in the target area, forming the basis for calculating the soil health index. Background indicators at the regional scale (such as topsoil texture and texture configuration) in the evaluation indicators can be obtained from the national soil census dataset and regional land use change survey results.

[0025] Step 102: Construct corresponding soil health evaluation indicators based on preset soil health functions. The preset soil health functions include at least one of the following: water regulation function, maintenance of plant and animal habitat and life activities function, pollutant filtration and buffering function, nutrient cycling function, and physical stability and structural support function; calculate the partial correlation coefficients between soil health evaluation indicators, and construct an indicator association network based on the partial correlation coefficients.

[0026] The average connection strength, the proportion of positively correlated edges, and the proportion of negatively correlated edges in the indicator association network are used to characterize the synergistic or antagonistic relationships among various dimensions of soil health indicators.

[0027] In some embodiments, the soil health evaluation indicators corresponding to the water regulation function include at least one of topsoil texture, soil bulk density, and irrigation capacity; the soil health evaluation indicators corresponding to the function of maintaining plant and animal habitats and life activities include at least one of soil pH, soil respiration, fungal Chao1 index, and fungal Shannon index; the soil health evaluation indicators corresponding to the pollutant filtration and buffering function include at least one of organic matter content, cation exchange capacity, and Nemerow comprehensive pollution index; the soil health evaluation indicators corresponding to the nutrient cycling function include at least one of soil organic carbon density, total nitrogen, total phosphorus, total potassium, available phosphorus, and available potassium; and the soil health evaluation indicators corresponding to the physical stability and structural support function include at least one of topsoil thickness and texture configuration.

[0028] The formula for calculating the Nemerow Comprehensive Pollution Index is:

[0029] in, This indicates the Nemerow Comprehensive Pollution Index; Indicates the pollution index of each individual pollutant. The average value; Indicates the pollution index of all individual pollutants. The square of the maximum value in.

[0030] Single pollutant pollution index The calculation formula is:

[0031] in, This represents the measured value of the pollutants; This represents the evaluation standard value.

[0032] By setting five categories of soil health functions and configuring corresponding soil health evaluation indicators, the partial correlation coefficients between the evaluation indicators are calculated, and each soil health evaluation indicator is used as a node to construct an indicator association network. This can distinguish the attribution relationship of each evaluation indicator under different soil function dimensions, and present the strength of the association between each soil health evaluation indicator. This facilitates the analysis of the degree of mutual influence of indicators within each soil health function and between different soil health functions through the indicator association network.

[0033] Step 103: Determine the edge weights between indicator nodes in the indicator association network, calculate the strength centrality of the indicator nodes based on the edge weights, and determine the indicator weights of the indicator nodes based on the strength centrality.

[0034] In some embodiments, the edge weights are the absolute values ​​of the partial correlation coefficients. By setting the absolute values ​​of the partial correlation coefficients as the edge weights of the indicator association network, the interference caused by the positive and negative correlation directions on the determination of the strength of the indicator association is eliminated, and the magnitude of the association effect between each indicator node is uniformly quantified.

[0035] In some embodiments, the formula for calculating intensity centrality is:

[0036] in, Indicates the strength centrality of index node i; This represents the edge weight between index node i and index node j; This represents the summation of edge weights among all index nodes j connected to index node i.

[0037] In some embodiments, the index weight is the ratio of the intensity centrality of the corresponding index node to the sum of the intensity centralities of all index nodes, and is determined based on the contribution of the index node. By using the proportion of the intensity centrality of a single index node in the sum of the intensity centralities of all index nodes as the index weight, the contribution of different index nodes in the index association network is quantified, and the weight values ​​of each soil health assessment index are assigned based on the association relationships between index nodes.

[0038] Step 104: Standardize the multi-source data to obtain standardized data. By standardizing the multi-source data to uniformly standardize it to the range of [0.1, 1.0], the interference caused by the differences in the dimensions and numerical ranges of different soil health evaluation indicators on the calculation of soil health index can be eliminated, and the impact on the overall assessment results can be reduced.

[0039] In some embodiments, standardization processing of multi-source data includes: performing multi-type standardization processing on multi-source data; the multi-type standardization processing includes at least one of membership function standardization processing based on upper and lower thresholds, standardization threshold determination processing according to preset standards, and hierarchical assignment processing using the natural breakpoint method.

[0040] Specifically, indicators with optimal ranges (such as soil pH, soil bulk density, and topsoil thickness) are standardized using S-shaped or inverse S-shaped membership functions based on upper and lower thresholds. For nutrient indicators (such as organic matter content, total nitrogen, and total phosphorus), the standardization critical values ​​can be determined with reference to the "National Second Soil Survey Nutrient Grading Standards." For indicators without clear reference standards (such as soil respiration and soil organic carbon density), the natural breakpoint method is used for grading and assigning values. By matching corresponding standardization methods to different types of soil health assessment indicators, the data distribution characteristics and judgment criteria of various soil health assessment indicators are adapted to improve the matching degree between the standardized values ​​of multi-source data and the actual performance of soil health.

[0041] Step 105: Calculate the soil health index for multi-source data sampling points based on indicator weights and standardized data. Calculating the soil health index for multi-source data sampling points based on indicator weights and standardized data allows for a comprehensive assessment of the overall impact on soil health at the sampling points, providing a direct representation of the soil health level at each sampling point.

[0042] In some embodiments, the soil health index is calculated using the following formula:

[0043] in, This represents the soil health index; n represents the total number of indicator nodes. (NA) represents the index weight of the i-th index node obtained based on network analysis; This represents the i-th standardized data value.

[0044] Step 106: Based on the soil health index, the soil health of cultivated land is classified using the natural breakpoint method, and the soil health evaluation result is output based on the soil health classification results.

[0045] By classifying arable land soil health into multiple levels, a soil health grading result is obtained. By matching the spatial coordinates of each sampling point with the soil health grading result, a spatial distribution map of health is constructed as the soil health assessment result, and this map is output. This method can intuitively distinguish the differences in soil health levels among different plots of arable land, presenting the spatial distribution characteristics of arable land soil health status within the region.

[0046] In some embodiments, classifying arable land soil health using the natural discontinuity method includes classifying arable land soil health into healthy, sub-healthy, medium, vulnerable, and unhealthy levels using the natural discontinuity method.

[0047] This invention also provides a network analysis-based system for evaluating farmland soil health, comprising: a multi-source data acquisition module for acquiring multi-source data of farmland soil in a target area; an indicator network construction module for constructing corresponding soil health evaluation indicators based on preset soil health functions, including at least one of water regulation, maintenance of plant and animal habitats and life activities, pollutant filtration and buffering, nutrient cycling, and physical stability and structural support; calculating partial correlation coefficients between soil health evaluation indicators and constructing an indicator association network based on the partial correlation coefficients; a weight calculation module for determining the edge weights between indicator nodes in the indicator association network, calculating the strength centrality of indicator nodes based on the edge weights, and determining the indicator weights of indicator nodes based on the strength centrality; a data standardization module for standardizing the multi-source data to obtain standardized data; an index calculation module for calculating the soil health index of multi-source data sampling points based on the indicator weights and standardized data; and a classification output module for classifying farmland soil health based on the soil health index using the natural breakpoint method and outputting soil health evaluation results based on the farmland soil health classification results.

[0048] The same or similar parts among the various embodiments in this specification can be referred to mutually, and will not be repeated here.

[0049] This invention constructs corresponding soil health evaluation indicators based on preset soil health functions, builds an indicator association network by calculating partial correlation coefficients, and determines the indicator weights of indicator nodes based on the strength centrality of the nodes. This objective weighting based on node strength centrality effectively analyzes the multidimensional nonlinear relationships between soil indicators, avoids biases caused by subjective scoring, and overcomes the subjectivity and linearity limitations of traditional weight determination methods. Furthermore, the output health spatial distribution map can be directly used to guide regional farmland soil zoning management, precision fertilization, and soil improvement, demonstrating strong practical application value.

[0050] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0051] Please see Figure 2 , Figure 2 The symbols in the table are: PT: topsoil texture; BD: bulk density; IC: irrigation capacity; pH: soil acidity / alkalinity; SR: soil respiration; FC: fungal Chao1 index; FS: fungal Shannon index; SOM: organic matter content; CEC: cation exchange capacity; PN: Nemerow integrated pollution index; SOCD: soil organic carbon density; N: total nitrogen; P: total phosphorus; K: total potassium; AP: available phosphorus; AK: available potassium; PLT: topsoil thickness; TC: texture configuration.

[0052] As a specific example: Taking Tieling County, Liaoning Province as an example area, 100 surface soil samples of cultivated land were collected. The sampling method was as follows: stratified random sampling was adopted, and the number of sampling points was allocated according to the proportion of cultivated land area of ​​each land use grade to the total cultivated land area of ​​the county. The corresponding number of sampling points were randomly generated within each land use grade using a random point generation tool. GPS was used for precise positioning, and 2-3 sub-samples were collected within a 20m radius around each sampling point. After being mixed evenly, the samples were reduced to 1kg using the quartering method.

[0053] Soil health assessment indicators, including soil bulk density, soil physicochemical properties (soil pH, organic matter content, total nitrogen, total phosphorus, total potassium, available phosphorus, available potassium, cation exchange capacity), heavy metal content (As, Cd, Cr, Cu, Ni, Pb, Zn, Hg), fungal community diversity (Chao1 index, Shannon index), irrigation capacity, topsoil texture and texture configuration, soil respiration, and soil organic carbon density, were measured.

[0054] Partial correlation coefficients were calculated among the indicators, and an indicator association network was constructed based on these coefficients. In this embodiment, 31 significant non-zero edges were identified, with a network sparsity of 79.739%. There were 20 positively correlated edges (64.516%) and 11 negatively correlated edges (35.484%), indicating an overall co-evolutionary trend among the soil indicators. The average network connectivity strength was 0.257. Specifically, the partial correlation coefficient between organic matter content (SOM) and total nitrogen (N) was 0.613, between total phosphorus (P) and available phosphorus (AP) was 0.44, between the fungal Chao1 index (FC) and the fungal Shannon index (FS) was 0.361, and between irrigation capacity (IC) and soil texture (TC) was -0.36.

[0055] The intensity centrality of each indicator node was calculated: In this embodiment, the top six indicators in terms of intensity centrality are, in order: soil organic matter content (SOM), total phosphorus (P), available phosphorus (AP), available potassium (AK), soil organic carbon density (SOCD), and fungal Chao1 index (FC). Among them, soil organic matter content has the highest intensity centrality and is the key node connecting other indicators. The weight of each indicator is determined according to the contribution of node intensity, with indicators having higher intensity centrality having greater weight.

[0056] Multi-source data of cultivated soil in the target area were standardized to the range of [0.1, 1.0]. pH was standardized using a peak-shaped membership function; bulk density was standardized using a peak-shaped membership function; topsoil thickness was standardized using a peak-shaped membership function; organic matter content was standardized according to the nutrient grading standards of the Second National Soil Survey: >40g / kg was grade 1, 30-40g / kg was grade 0.85, 20-30g / kg was grade 0.6, 10-20g / kg was grade 0.4, 6-10g / kg was grade 0.2, and <6g / kg was grade 0.1; the Chao1 index and Shannon index of fungi were graded and assigned values ​​using the natural breakpoint method; the soil health index of each sampling point was calculated according to the formula. The evaluation results were divided into five levels using the natural breakpoint method: healthy, sub-healthy, moderate, fragile, and unhealthy.

[0057] In the application in Tieling County of this embodiment, the area proportions of each level are as follows: fragile 32.021%, moderate 24.746%, sub-healthy 18.938%, unhealthy 18.487%, and healthy 5.808%. Soil health generally shows a spatial distribution pattern of high in the southeast and low in the northwest, with high-value areas concentrated in Liqianhu Township and low-value areas mainly distributed in the northern part of Tieling County.

[0058] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for assessing arable land soil health based on network analysis, characterized in that, include: Acquire multi-source data on arable land soil in the target area; Based on preset soil health functions, corresponding soil health evaluation indicators are constructed. The preset soil health functions include at least one of the following: water regulation function, function of maintaining plant and animal habitat and life activities, pollutant filtration and buffering function, nutrient cycling function, and physical stability and structural support function. Calculate the partial correlation coefficients among the soil health assessment indicators, and construct an indicator association network based on the partial correlation coefficients; Determine the edge weights between indicator nodes in the indicator association network, calculate the strength centrality of the indicator nodes based on the edge weights, and determine the indicator weights of the indicator nodes based on the strength centrality. The multi-source data is standardized to obtain standardized data; Based on the index weights and the standardized data, the soil health index of the multi-source data sampling points is calculated. Based on the soil health index, the natural breakpoint method is used to classify the soil health of cultivated land, and the soil health evaluation result is output based on the classification result of cultivated land soil health.

2. The method for evaluating arable land soil health based on network analysis according to claim 1, characterized in that, The soil health evaluation indicators corresponding to the water regulation function include at least one of topsoil texture, soil bulk density and irrigation capacity. The soil health evaluation indicators corresponding to the maintenance of plant and animal habitats and life activities include at least one of soil pH, soil respiration, fungal Chao1 index and fungal Shannon index. The soil health evaluation indicators corresponding to the pollutant filtration and buffering functions include at least one of organic matter content, cation exchange capacity, and Nemerow comprehensive pollution index. The soil health evaluation indicators corresponding to the nutrient cycling function include at least one of soil organic carbon density, total nitrogen, total phosphorus, total potassium, available phosphorus, and available potassium. The soil health evaluation indicators corresponding to the physical stability and structural support function include at least one of the topsoil thickness and texture configuration.

3. The method for evaluating arable land soil health based on network analysis according to claim 1, characterized in that, The formula for calculating the intensity centrality is: in, Indicates the strength centrality of index node i; This represents the edge weight between index node i and index node j; This represents the summation of edge weights among all index nodes j connected to index node i.

4. The method for evaluating arable land soil health based on network analysis according to claim 1, characterized in that, The edge weights are the absolute values ​​of the partial correlation coefficients.

5. The method for evaluating arable land soil health based on network analysis according to claim 1, characterized in that, The index weight is the ratio of the strength centrality of the corresponding index node to the sum of the strength centralities of all index nodes.

6. The method for evaluating arable land soil health based on network analysis according to claim 1, characterized in that, Standardization processing of the multi-source data includes: performing various types of standardization processing on the multi-source data; The various types of standardization processing include at least one of the following: membership function standardization based on upper and lower thresholds, standardization threshold determination according to preset standards, and hierarchical assignment using the natural breakpoint method.

7. The method for evaluating arable land soil health based on network analysis according to claim 1, characterized in that, The formula for calculating the soil health index is as follows: in, This represents the soil health index; n represents the total number of indicator nodes. (NA) represents the index weight of the i-th index node obtained based on network analysis; This represents the i-th standardized data value.

8. The method for evaluating arable land soil health based on network analysis according to claim 1, characterized in that, The method of classifying arable land soil health using the natural breakpoint method includes: The health of arable land soil was classified into healthy, sub-healthy, medium, fragile and unhealthy levels using the natural breakpoint method.

9. A farmland soil health assessment system based on network analysis, characterized in that, include: The multi-source data acquisition module is used to acquire multi-source data on arable soil in the target area; The indicator network construction module is used to construct corresponding soil health evaluation indicators based on preset soil health functions. The preset soil health functions include at least one of the following: water regulation function, function of maintaining plant and animal habitat and life activities, pollutant filtration and buffering function, nutrient cycling function, and physical stability and structural support function. Calculate the partial correlation coefficients among the soil health assessment indicators, and construct an indicator association network based on the partial correlation coefficients; The weight calculation module is used to determine the edge weights between index nodes in the index association network, calculate the strength centrality of the index nodes based on the edge weights, and determine the index weights of the index nodes based on the strength centrality. The data standardization module is used to standardize the multi-source data to obtain standardized data. The index calculation module is used to calculate the soil health index of multi-source data sampling points based on the index weights and the standardized data. The grading output module is used to grade the soil health of cultivated land based on the soil health index using the natural breakpoint method, and output the soil health evaluation result based on the grading result of cultivated land soil health.

Images