Agricultural soil pollution prediction method

Abstract

The invention relates to the technical field of soil pollution component analysis, and discloses an agricultural soil pollution prediction method. The method comprises the following steps: carrying out soil physicochemical property detection through layered gridding sampling, determining the heavy metal content, the organic matter content and the enzyme activity, and obtaining a pollution characteristic data set; monitoring diffusion characteristics of pollutants by adopting an electrochemical sensing detection array, and generating a migration rule model; a multi-parameter monitoring system is used for detecting the pH value, the oxidation-reduction potential and the ion activity, and a dynamic change sequence is formed; establishing a colloid interface action response matrix, and constructing a pollutant accumulation model; determining the occurrence form and the ion concentration of the pollutants, and generating an interaction map; and finally, constructing an early warning system, and outputting a prevention and control scheme. According to the invention, through cooperative application of multiple analysis and test means, and in combination with interface action mechanism research of pollutants in soil, whole-process intelligent monitoring of soil pollution prediction is realized.

Description

Technical Field

[0001] The present application relates to the field of soil pollution component analysis, and in particular to a method for predicting agricultural soil pollution. Background Art

[0002] Agricultural soil pollution monitoring and prediction is an important research topic in the field of environmental protection. At present, the commonly used soil pollution monitoring methods mainly include physical and chemical analysis methods such as atomic absorption spectroscopy, inductively coupled plasma mass spectrometry, and online monitoring systems based on electrochemical sensors. These methods have made certain progress in the determination of pollutant content. At the same time, a variety of mathematical models have been developed in the field of soil pollution prediction, such as regression models based on statistical analysis and prediction models based on machine learning. In the study of pollutant migration and transformation, researchers have conducted in-depth research on the occurrence forms and environmental behaviors of pollutants in soil through chromatography-mass spectrometry, synchrotron radiation technology and other means.

[0003] However, the relevant technologies have the following shortcomings: existing monitoring methods are often limited to the detection of a single pollutant or a certain type of pollutant, and it is difficult to systematically evaluate the interactions between multiple pollutants and their comprehensive impact on the soil environment; conventional prediction models mainly rely on historical data for statistical analysis, and lack an in-depth understanding of the migration and transformation mechanisms of pollutants at the microscopic scale; traditional sampling and analysis methods mostly use laboratory offline detection, which makes it difficult to achieve real-time monitoring of the dynamic changes of pollutants; in addition, existing technologies often lack sufficient consideration of regional differences in the formulation of soil pollution prevention and control plans. Summary of the invention

[0004] The present application provides a method for predicting agricultural soil pollution, which is used to establish a technical system for predicting agricultural soil pollution that can systematically characterize the interactions between multiple pollutants, dynamically monitor the migration and transformation of pollutants, and output targeted prevention and control plans. The present invention realizes intelligent monitoring of the entire process of soil pollution prediction through the coordinated application of multiple analytical and testing methods, combined with the study of the interface mechanism of pollutants in the soil.

[0005] In a first aspect, the present application provides a method for predicting agricultural soil pollution, which comprises: performing multi-point physical and chemical property detection on soil samples in agricultural areas through layered grid sampling, performing heavy metal content determination, organic matter content analysis and enzyme activity test on soil samples, and combining ion chromatography analysis and atomic absorption spectroscopy detection to obtain a soil pollution characteristic data set; Based on the soil pollution characteristic data set, an electrochemical sensor detection array is used to dynamically monitor the diffusion characteristics of pollutants in soil samples, and a pollutant migration law characteristic model is generated by combining dielectric property analysis and fluorescence spectrum detection; Based on the characteristic model of pollutant migration law, a multi-parameter online monitoring system is used to perform real-time detection of soil pH value, redox potential and ion activity, and a dynamic change characteristic sequence of soil pollutants is formed through multi-channel signal acquisition and data correction; According to the dynamic change characteristic sequence of soil pollutants, a response relationship matrix between pollutant concentration and soil colloid interface interaction is established, and a pollutant accumulation model considering the interface adsorption effect is constructed by combining capillary electrophoresis analysis and mass spectrometry detection; Using the pollutant accumulation model, the occurrence forms of pollutants in the soil are determined by synchronous thermal analysis and X-ray diffraction methods, and the changes in key ion concentrations are detected using a selective ion electrode array to generate a multi-pollutant interaction characteristic map; According to the characteristic map of multi-pollutant interactions, a pollutant content early warning system is constructed based on thermogravimetric analysis and colorimetry, and combined with spectral analysis methods, a regional soil pollution prevention and control plan is output.

[0006] In the technical solution provided by this application, a comprehensive assessment of the pollution status of farmland soil is achieved through layered grid sampling and multi-point physical and chemical property detection, ensuring the representativeness of sampling and the accuracy of data. The diffusion characteristics of pollutants in soil samples are dynamically monitored by electrochemical sensor detection array, and the migration law of pollutants in soil is effectively captured by combining dielectric property analysis and fluorescence spectrum detection. The soil pH value, redox potential and ion activity are detected in real time by using a multi-parameter online monitoring system, and the dynamic change characteristics of soil pollutants are accurately reflected through multi-channel signal acquisition and data correction. A response relationship matrix between pollutant concentration and soil colloid interface is established, and combined with capillary electrophoresis analysis and mass spectrometry detection, the influence mechanism of interfacial adsorption effect on pollutant accumulation is deeply revealed. The occurrence form of pollutants in soil is determined by synchronous thermal analysis and X-ray diffraction methods, and the change of key ion concentration is detected by selective ion electrode array, and the interaction characteristics between multiple pollutants are obtained. Based on thermogravimetric analysis and colorimetry, a pollutant content early warning system is constructed, and combined with spectral analysis methods, the accurate output of regional soil pollution prevention and control plans is realized, which improves the accuracy and efficiency of pollutant detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0008] Figure 1 This is a schematic diagram of an embodiment of the method for predicting agricultural soil pollution in the embodiments of the present application; Figure 2 Schematic diagram of the spectrum of the pollutant group in the embodiment of the present application. DETAILED DESCRIPTION

[0009] The embodiment of the present application provides a method for predicting agricultural soil pollution. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0010] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 , an embodiment of the agricultural soil pollution prediction method in the embodiment of the present application includes: Step S101: Perform multi-point physical and chemical property detection on soil samples in agricultural areas through stratified grid sampling, perform heavy metal content determination, organic matter content analysis and enzyme activity test on soil samples, and combine ion chromatography analysis and atomic absorption spectroscopy detection to obtain a soil pollution characteristic data set; Step S102: Based on the soil pollution characteristic data set, an electrochemical sensor detection array is used to dynamically monitor the diffusion characteristics of pollutants in the soil sample, and a pollutant migration law characteristic model is generated by combining dielectric property analysis and fluorescence spectrum detection; Step S103: Based on the characteristic model of pollutant migration law, a multi-parameter online monitoring system is used to perform real-time detection of soil pH value, redox potential and ion activity, and a dynamic change characteristic sequence of soil pollutants is formed through multi-channel signal acquisition and data correction; Step S104: According to the dynamic change characteristic sequence of soil pollutants, a response relationship matrix between pollutant concentration and soil colloid interface interaction is established, and a pollutant accumulation model considering the interface adsorption effect is constructed by combining capillary electrophoresis analysis and mass spectrometry detection; Step S105: using the pollutant accumulation model, determining the occurrence forms of pollutants in the soil by synchronous thermal analysis and X-ray diffraction methods, using a selective ion electrode array to detect changes in key ion concentrations, and generating a multi-pollutant interaction characteristic map; Step S106: Based on the characteristic spectrum of multi-pollutant interactions, a pollutant content early warning system is constructed based on thermogravimetric analysis and colorimetry, and a regional soil pollution prevention and control plan is output in combination with a spectral analysis method.

[0011] Specifically, the agricultural area was sampled in a stratified grid. The sampling was done by dividing the agricultural area into 10m x 10m detection units, and arranging sampling points at the center point, vertex and midpoint of the edge of each detection unit to form a sampling point array. Soil sampling was carried out at three depths (0-20cm, 20-40cm, 40-60cm) at each sampling point, and the soil samples were mixed and homogenized by the quartering method. The heavy metal content of the treated soil samples was determined, and the copper, zinc, lead and cadmium contents were determined by flame atomic absorption spectrophotometry, and the arsenic and mercury contents were determined by inductively coupled plasma mass spectrometry. The organic matter content was analyzed by potassium dichromate oxidation-external heating method for soil digestion, and then determined by ultraviolet spectrophotometer. The soil urease activity test was determined by alkaline hydrolysis. The sulfate, nitrate and chloride ion contents were determined by ion chromatography, and the sodium, potassium, calcium and magnesium ion contents were detected by atomic absorption spectroscopy.

[0012] On the basis of obtaining soil pollution characteristic data sets, an electrochemical sensor detection array is deployed for dynamic monitoring. The electrochemical sensor detection array consists of ion selective electrodes, conductivity electrodes and redox electrodes, and the concentration of heavy metal ions is continuously collected by potentiometric method. A four-electrode AC impedance spectrometer is used to measure the dielectric constant and dielectric loss of soil samples, and the polarization characteristics under the power frequency electric field are analyzed. Fluorescence spectrum detection uses a fluorescence spectrophotometer to scan the excitation-emission matrix, and the fluorescence characteristics of organic pollutants are studied by analyzing the peak position and peak intensity characteristics of the three-dimensional fluorescence spectrum. The obtained data is monitored online with multiple parameters, and a composite glass electrode is installed to determine the pH value, a platinum electrode is used to determine the redox potential, and an ion selective electrode is used to determine the ion activity. The electrode output signal is amplified and filtered by the signal conditioning circuit, and the measuring electrode is calibrated by the three-point calibration method and temperature compensation. The electrochemical workstation collects the potential signal, establishes the potential reference with the reference electrode, and uses digital filtering to eliminate external interference. The potential difference of the target ion is measured by the ion selective electrode, and the ion activity is calculated by the Nernst equation. The multi-channel data acquisition system uses wavelet denoising to reduce the noise of the signal.

[0013] After obtaining the characteristic sequence of dynamic changes of soil pollutants, the soil colloids were ultrasonically dispersed and centrifuged, and the colloidal particle size distribution was determined by a laser particle size analyzer. The surface charge of the soil colloids was separated by capillary zone electrophoresis, and the migration time of each component was determined by a conductivity detector. The content of colloid-bound pollutants was determined by an inductively coupled plasma mass spectrometer, and the pollutants were qualitatively and quantitatively analyzed by tandem mass spectrometry. The isothermal adsorption equation was used to calculate the interface equilibrium constant and the maximum adsorption capacity, the scanning tunneling microscope was used to observe the microscopic morphology of the colloid surface, and the atomic force microscope was used to determine the surface adsorption force. In order to determine the occurrence form of pollutants in the soil, the soil samples were heated by a differential scanning calorimeter to monitor the heat flow during the temperature change process. The crystal phase structure of the soil samples was determined by an X-ray diffractometer, and the diffraction peaks were analyzed by the Bragg equation. The concentrations of potassium ions, sodium ions, and ammonium ions were measured by a selective ion electrode, and the ion concentrations were calculated by a potentiometric titration curve. The thermogravimetric analyzer was used to determine the change of sample mass with temperature, and the thermal decomposition products of pollutants were analyzed by combining thermogravimetric-mass spectrometry. X-ray photoelectron spectroscopy was used to determine the valence state of elements on the sample surface, and Auger electron spectroscopy was used to analyze the surface chemical composition.

[0014] Finally, a thermogravimetric-differential thermal analyzer was used to program the temperature of the soil samples, and the thermogravimetric and differential thermal curves were collected. A visible spectrophotometer was used to measure the absorbance of the heavy metal color development reaction, and the pollutant concentration was calculated in combination with the standard curve. An infrared spectrometer scanned the samples in the entire band, and the characteristic absorption peaks were identified and quantitatively analyzed. An ultraviolet-visible spectrometer was used to measure the absorption spectrum of the samples, and the second-order derivative spectrum was used to eliminate background interference. A fluorescence analyzer was used to measure the excitation-emission spectrum of the samples, and the spectral data was analyzed in combination with parallel factor analysis. A fiber optic spectrometer was used for in-situ detection, and the distribution characteristics of pollutants were established in combination with chemometric analysis.

[0015] For example: After the sampling area is divided into 10m x 10m grids, in the 0-20cm soil layer samples, the flame atomic absorption spectrophotometry measured the copper content to be 45.6mg / kg, the zinc content to be 132.8mg / kg, the lead content to be 68.3mg / kg, and the cadmium content to be 0.45mg / kg. The inductively coupled plasma mass spectrometry measured the arsenic content to be 12.4mg / kg and the mercury content to be 0.28mg / kg. Ion chromatography analysis showed that the sulfate content was 245mg / kg, the nitrate content was 168mg / kg, and the chloride content was 92mg / kg. Atomic absorption spectroscopy showed that the sodium ion content was 156mg / kg, the potassium ion content was 1850mg / kg, the calcium ion content was 3260mg / kg, and the magnesium ion content was 425mg / kg. The pH value was 6.8, the redox potential was 245mV, and the organic matter content was 2.8%. Through analysis of the established early warning system, the lead and cadmium content in the soil in the area is close to the critical value, and targeted prevention and control measures need to be taken, including adjusting the planting structure and improving the physical and chemical properties of the soil.

[0016] In the embodiment of the present application, a comprehensive assessment of the pollution status of farmland soil is achieved through layered grid sampling and multi-point physical and chemical property detection, ensuring the representativeness of sampling and the accuracy of data. The diffusion characteristics of pollutants in soil samples are dynamically monitored by electrochemical sensing detection array, and the migration law of pollutants in soil is effectively captured by combining dielectric property analysis and fluorescence spectrum detection. The soil pH value, redox potential and ion activity are detected in real time by using a multi-parameter online monitoring system, and the dynamic change characteristics of soil pollutants are accurately reflected by multi-channel signal acquisition and data correction. A response relationship matrix between pollutant concentration and soil colloid interface is established, and combined with capillary electrophoresis analysis and mass spectrometry detection, the influence mechanism of interfacial adsorption effect on pollutant accumulation is deeply revealed. The occurrence form of pollutants in soil is determined by synchronous thermal analysis and X-ray diffraction methods, and the change of key ion concentration is detected by selective ion electrode array, and the interaction characteristics between multiple pollutants are obtained. Based on thermogravimetric analysis and colorimetry, a pollutant content early warning system is constructed, and combined with spectral analysis methods, the accurate output of regional soil pollution prevention and control plans is realized, which improves the accuracy and efficiency of pollutant detection.

[0017] In a specific embodiment, the process of executing step S101 may specifically include the following steps: (1) The agricultural area is divided into multiple 10 m × 10 m detection units by presetting the sampling point spacing. The sampling point array is composed of the center point, vertex and edge midpoint of the detection unit to obtain the initial sampling point distribution data; (2) Based on the initial sampling point data, soil samples were taken at three depths of 0-20 cm, 20-40 cm, and 40-60 cm at each sampling point, and the soil samples were mixed and homogenized using the quartering method to obtain a stratified soil sample set; (3) The contents of copper, zinc, lead, and cadmium in the stratified soil samples were determined by flame atomic absorption spectrophotometry, and the contents of arsenic and mercury were determined by inductively coupled plasma mass spectrometry to obtain heavy metal content data; (4) The stratified soil sample set was digested by potassium dichromate oxidation-external heating method, the organic matter content was determined by ultraviolet spectrophotometer, and the soil urease activity was determined by alkaline hydrolysis method to obtain soil basic property data; (5) Based on the heavy metal content data and soil basic characteristics data, the sulfate, nitrate and chloride ion contents in the soil samples were quantitatively analyzed by ion chromatography, and the sodium, potassium, calcium and magnesium ions were detected by atomic absorption spectrometry to obtain the ion composition characteristic data; (6) The heavy metal content data, soil basic property data, and ion composition characteristic data were standardized and outliers were eliminated. The characteristic weights were determined by principal component analysis to obtain the soil pollution characteristic data set.

[0018] Specifically, an electrochemical sensor detection array is arranged at a predetermined sampling point. The array contains three core electrodes: ion-selective electrodes are used for the selective detection of specific heavy metal ions, conductivity electrodes measure the conductivity of soil solutions, and redox electrodes monitor the redox state of the soil. Potentiometric determination determines the concentration of the target ion by measuring the potential difference between the working electrode and the reference electrode. According to the principle of the Nernst equation, the magnitude of the potential difference is logarithmically related to the concentration of the target ion. By collecting data every 30 minutes and continuously monitoring for 24 hours, the dynamic changes in the concentration of heavy metal ions are recorded.

[0019] After obtaining the pollutant concentration time series data, a four-electrode AC impedance spectrometer is used to measure the dielectric properties of the soil samples. The four-electrode method is to apply an AC voltage through the two outer electrodes, measure the voltage drop through the two inner electrodes, and measure the current passing through the sample at the same time, so as to calculate the impedance value. The dielectric constant represents the ability of the material to store electrical energy, and the dielectric loss reflects the degree to which the material consumes electrical energy. The polarization characteristic analysis under the industrial frequency electric field includes measuring the dielectric parameters of the sample at different frequencies (usually 50Hz-1MHz) and recording the response characteristics of the polarization intensity with the change of the electric field. The fluorescence spectrum analysis of the soil sample is carried out using a fluorescence spectrophotometer, and a three-dimensional fluorescence spectrum is obtained by continuously changing the excitation wavelength and the emission wavelength. The process of excitation-emission matrix scanning is to record a complete emission spectrum at each excitation wavelength, and finally form a three-dimensional data matrix. The peak position information reflects the molecular structure characteristics of organic pollutants, and the peak intensity is related to the pollutant concentration. The fluorescence characteristic spectrum analysis focuses on the characteristic peaks of organic pollutants such as humus and aromatic compounds. During the data fusion process, the soil dielectric parameter set and the fluorescence characteristic spectrum of organic pollutants are extracted through covariance matrix decomposition. Covariance matrix decomposition first calculates the covariance matrix between the two sets of data, and then performs eigenvalue decomposition on the matrix to extract the main change characteristics. By analyzing the eigenvectors and eigenvalues, the main influencing factors in the migration process of pollutants are identified.

[0020] When using a scanning electron microscope to observe the surface morphology of soil particles, the sample needs to be pre-treated by drying, gold spraying, etc. The scanning electron microscope can provide microscopic morphological information of the sample surface with a resolution of up to nanometers. The energy spectrum analyzer determines the elemental composition and relative content of the sample surface by detecting the X-ray energy distribution. The distribution characteristics of pollutants on the surface of soil particles are quantitatively characterized. Finally, based on the obtained interface distribution characteristics, the anisotropic diffusion coefficient is calculated. Anisotropic diffusion means that the diffusion rate of pollutants in different directions of the soil is different, and the diffusion coefficients in the horizontal and vertical directions need to be calculated separately. The interfacial mass transfer resistance analysis takes into account the resistance of pollutants from the surface of soil particles to the solution phase, and combines the migration rate of pollutants to construct a complete migration law characteristic model.

[0021] For example: First, the cadmium ions in the soil solution were dynamically monitored for 24 hours by ion selective electrodes, and data was collected every 30 minutes. The measured cadmium ion concentration fluctuated between 0.08-0.15 mg / L. The four-electrode AC impedance spectrum measurement showed that the dielectric constant of the soil sample at a frequency of 1kHz was 12.5 and the dielectric loss was 0.85. Fluorescence spectrum analysis found that characteristic peaks appeared at an excitation wavelength of 280nm and an emission wavelength of 340nm, with a peak intensity of 2850 (relative units), indicating the presence of aromatic organic pollution. Through covariance matrix decomposition, the eigenvalue of the first principal component was 3.24, and the corresponding eigenvector showed that the migration of pollutants was mainly affected by the content of soil organic matter. Scanning electron microscopy observations showed that there were obvious needle-shaped crystals on the surface of soil particles, and energy spectrum analysis confirmed that they were cadmium compound forms. After calculation, the horizontal diffusion coefficient was 3.2×10^-6cm^2 / s, and the vertical diffusion coefficient was 1.8×10^-6 cm^2 / s, indicating that the pollutants had a stronger ability to migrate in the horizontal direction.

[0022] In a specific embodiment, the process of executing step S102 may specifically include the following steps: (1) Based on the soil pollution characteristic data set, an electrochemical sensor detection array consisting of ion selective electrodes, conductivity electrodes and redox electrodes was deployed at the sampling points to continuously collect the concentration of heavy metal ions through the potentiometric method to obtain the time series data of pollutant concentration; (2) Based on the pollutant concentration time series data, a four-electrode AC impedance spectrometer was used to measure the dielectric constant and dielectric loss of the soil samples. Combined with the polarization characteristics analysis under the power frequency electric field, the soil dielectric parameter set was obtained; (3) Scan the soil sample by excitation-emission matrix using a fluorescence spectrophotometer, and analyze the peak position and peak intensity of the three-dimensional fluorescence spectrum to form the fluorescence characteristic spectrum of organic pollutants; (4) Fusing the soil dielectric parameter set with the fluorescence characteristic spectrum of organic pollutants, extracting the pollutant migration characteristics through covariance matrix decomposition, and obtaining pollutant diffusion characteristic data; (5) Based on the pollutant diffusion characteristics data, the surface morphology of soil particles is observed using a scanning electron microscope, and the surface element distribution is measured using an energy spectrum analyzer to obtain the interface distribution characteristics of pollutants; (6) Based on the interfacial distribution characteristics of pollutants, a characteristic model of pollutant migration law is generated by calculating the anisotropic diffusion coefficient and analyzing the interfacial mass transfer resistance, combined with the migration rate of pollutants in the soil.

[0023] Specifically, based on the obtained soil pollution characteristic data set, an electrochemical sensor detection array is arranged at the sampling point. The array contains three types of electrodes: ion selective electrode, conductivity electrode and redox electrode. The ion selective electrode is a selective membrane electrode whose potential change only responds to specific ions. The conductivity electrode is used to measure the total concentration of ions in the soil, and the redox electrode detects the potential difference of the redox electrode in the soil. In practical applications, the potentiometric method determines the concentration of heavy metal ions by measuring the potential difference between the working electrode and the reference electrode based on the principle of the Nernst equation. The potential difference is linearly related to the logarithm of the ion concentration. When collecting data continuously, data is collected every 5 minutes for 48 hours continuously to record the dynamic change trend of the target heavy metal ion concentration. After obtaining the pollutant concentration time series data, the soil sample is measured using a four-electrode AC impedance spectrometer. The principle of the four-electrode method is to apply an AC excitation signal through two external electrodes and measure the response signal through two internal electrodes. The dielectric constant characterizes the ability of the material to store charge, and the dielectric loss reflects the degree to which the material dissipates electrical energy. Under the condition of industrial frequency electric field, an alternating electric field in the range of 50Hz-10kHz is applied to the soil sample, and the impedance value and phase angle at different frequencies are recorded to obtain polarization characteristic information. The measurement results form a parameter set including frequency, impedance modulus value, and phase angle.

[0024] The fluorescence spectrum analysis of soil organic pollutants uses three-dimensional fluorescence spectrum technology. The fluorescence spectrophotometer uses a cross-scan mode, with an excitation wavelength range of 200-400nm, an emission wavelength range of 250-500nm, and a wavelength interval of 5nm. Each sample obtains a three-dimensional data matrix containing information in three dimensions: excitation wavelength, emission wavelength, and fluorescence intensity. The presence and relative content of different types of organic pollutants are determined by analyzing the position and intensity of characteristic peaks.

[0025] The data fusion of the soil dielectric parameter set and the fluorescence characteristic spectrum of organic pollutants adopts the covariance matrix decomposition method. First, the two sets of data are standardized to eliminate the dimension effect. Then the covariance matrix between the two sets of data is calculated, and the matrix is ​​subjected to eigenvalue decomposition to obtain eigenvalues ​​and eigenvectors. The size of the eigenvalue indicates the importance of the feature in the overall change, and the eigenvector reflects the correlation between the variables. By analyzing the eigenvectors corresponding to the main eigenvalues, the key features of pollutant migration are extracted. Before SEM observation, the soil samples need to be dried and gold-sprayed. The SEM uses the secondary electron signal generated by the interaction between the electron beam and the sample surface to obtain the morphological image of the sample surface with a resolution of 1-10nm. Energy spectrum analysis obtains the type and content information of the elements by detecting the characteristic X-rays generated by the elements on the sample surface excited by the electron beam. Through morphological observation and element distribution analysis, a database of the distribution characteristics of pollutants on the surface of soil particles is established.

[0026] Finally, the anisotropic diffusion coefficient is calculated based on the interface distribution characteristics. The diffusion rate of pollutants in the soil is different in the horizontal and vertical directions. This anisotropy originates from the layered structure characteristics of the soil. The interface mass transfer resistance analysis takes into account the resistance encountered by pollutants during the migration process at the solid-liquid interface, including the diffusion layer resistance and the surface adsorption resistance. By analyzing the diffusion coefficient and interface mass transfer coefficient of pollutants, combined with the measured migration rate data, a complete migration law characteristic model is constructed.

[0027] For example, the actual operation process is as follows: First, an electrochemical sensor detection array is deployed, and the copper ion selective electrode collects data every 5 minutes within 48 hours. It is recorded that the copper ion concentration fluctuates between 0.5-2.3 mg / L, and the fluctuation period is about 6 hours. The four-electrode AC impedance spectrum measurement shows that the dielectric constant of the soil sample is 18.3 and the dielectric loss is 0.92 at a frequency of 100 Hz, showing obvious polarization characteristics with increasing frequency. Three-dimensional fluorescence spectroscopy detected a characteristic peak with an intensity of 3200 at an excitation wavelength of 260 nm and an emission wavelength of 350 nm, indicating the presence of polycyclic aromatic hydrocarbons pollutants. After covariance matrix decomposition, the maximum eigenvalue of 4.56 was obtained, and the corresponding eigenvector showed that the migration of copper ions was mainly affected by soil pH and organic matter content. Scanning electron microscopy observations found blocky sediments on the surface of soil particles, and energy spectrum analysis confirmed that it was in the form of copper oxide, with a content of approximately 8.2% of the total surface elements. The horizontal diffusion coefficient was calculated , the vertical diffusion coefficient , interface mass transfer resistance coefficient .

[0028] In a specific embodiment, the process of executing step S103 may specifically include the following steps: (1) Based on the characteristic model of pollutant migration, composite glass electrodes, platinum electrodes and ion-selective electrodes are installed at the sampling points. The electrode output signals are amplified and filtered through the signal conditioning circuit to obtain the initial data of soil chemical parameters; (2) Using the initial data of soil chemical parameters, the measuring electrodes were calibrated at three points and temperature compensated, and the potential difference was measured in combination with the standard buffer solution to obtain the corrected pH value data set; (3) The potential signal measured by the platinum electrode is collected by an electrochemical workstation, a potential reference is established with a reference electrode, and digital filtering is used to eliminate external interference to form a redox potential data sequence; (4) Based on the pH value data set and the redox potential data sequence, the potential difference of the target ion is measured using an ion selective electrode, and the ion activity is calculated using the Nernst equation to obtain the ion activity distribution data; (5) Input the pH value data set, redox potential data sequence and ion activity distribution data into the multi-channel data acquisition system, use wavelet denoising to reduce the noise of the signal, and obtain the purified monitoring data; (6) Perform time series analysis and spatial interpolation processing on the purified monitoring data, and combine adaptive threshold filtering to eliminate outliers to form a dynamic change characteristic sequence of soil pollutants.

[0029] Specifically, based on the characteristic model of pollutant migration, three key electrodes are installed at the sampling point: composite glass electrode for pH determination, platinum electrode for redox potential measurement, and ion selective electrode for specific ion detection. The signal conditioning circuit processes the weak signal output by the electrode, including the preamplifier to amplify the microvolt signal to the millivolt level, and the bandpass filter to eliminate power frequency interference and high-frequency noise. The amplification factor is usually set to 1000 times, and the filter bandwidth is 0.1-100Hz to ensure signal quality.

[0030] The initial data of soil chemical parameters were processed by three-point calibration and temperature compensation. The three-point calibration used standard buffer solutions with pH values ​​of 4.00, 6.86 and 9.18, respectively. Each calibration point was measured three times, and the average value was taken to establish a calibration curve. Temperature compensation used platinum resistance thermometers. For every 1°C increase, the pH reading was corrected by -0.015 pH units. The potential difference was measured by standard buffer solution to obtain the calibrated pH value data set. When the electrochemical workstation collected the potential signal of the platinum electrode, the Ag / AgCl electrode was used as the reference electrode to establish the potential reference. The digital filter used a Butterworth low-pass filter with a cutoff frequency of 10Hz and an attenuation slope of -20dB / decade, which effectively eliminated external electromagnetic interference. The sampling frequency was 100Hz, and the recording time for each data point was 1 second to form a redox potential data sequence.

[0031] For the determination of ion activity, based on the pH data set and the redox potential data series, the potential difference of the target ion is measured using an ion selective electrode. The ion activity is calculated using the extended Nernst equation: ; Where: E is the measured electrode potential (V), is the standard electrode potential (V), is the activity coefficient correction factor (dimensionless), is the gas constant (8.314 J·mol⁻¹·K⁻¹), is the absolute temperature (K), is the ion valence (dimensionless), is the Faraday constant (96485C·mol⁻¹), is the activity of the target ion (mol·L⁻¹), is the soil solution ionic strength factor (dimensionless), is the ion interference correction factor (dimensionless).

[0032] Ion activity The specific calculation formula is: ; in is the temperature correction factor (J·mol⁻¹).

[0033] After the obtained pH data set, redox potential data sequence and ion activity distribution data were input into the multi-channel data acquisition system, the wavelet denoising technology was used for signal processing. The db4 wavelet basis function was selected, the number of decomposition layers was 4, and the wavelet coefficients were processed by the soft threshold method to reconstruct the purified monitoring data. The purified monitoring data were subjected to time series analysis and spatial interpolation processing. The time series analysis used the autocorrelation analysis method to identify the periodic characteristics of the data, and the spatial interpolation used the Kriging method for interpolation calculation. The adaptive threshold filter set the dynamic threshold window to ±3 times the standard deviation. Data points outside this range were marked as outliers and removed.

[0034] For example, in the pH value determination, the slope of the standard curve obtained by three-point calibration is -58.2mV / pH, and the theoretical Nernst slope at 25°C is -59.2mV / pH, with an error within 2%. The redox potential measurement shows that the original signal has 50Hz power frequency interference with an amplitude of about 2mV. After digital filtering, the interference is suppressed to below 0.1mV. In the determination of the activity of copper ions, the electrode potential E is 125mV, and the standard electrode potential is 260mV, the ion valence z is 2, and the temperature T is 298K. The calculated copper ion activity is After wavelet denoising, the signal-to-noise ratio was improved from the original 8.5dB to 15.3dB. Time series analysis showed that the data had an obvious 24-hour periodicity, and a complete pollutant concentration distribution map was formed after spatial interpolation.

[0035] In a specific embodiment, the process of executing step S104 may specifically include the following steps: (1) Based on the dynamic change characteristic sequence of soil pollutants, the soil colloids were ultrasonically dispersed and centrifuged, and the colloid particle size distribution was measured using a laser particle size analyzer to obtain the soil colloid characteristic parameters; (2) Based on the characteristic parameters of soil colloids, the surface charge of soil colloids is separated by capillary zone electrophoresis, and the migration time of each component is measured with a conductivity detector to obtain a colloid electrophoresis migration spectrum; (3) Based on the colloid electrophoretic migration spectrum, the content of colloid-bound pollutants is determined using an inductively coupled plasma mass spectrometer, and the pollutants are qualitatively and quantitatively analyzed by tandem mass spectrometry to form colloid-adsorbed pollutant data; (4) Input the colloid adsorption pollutant data into the isothermal adsorption equation, calculate the interface equilibrium constant and maximum adsorption capacity through Langmuir adsorption theory, and obtain the interface adsorption equilibrium parameters; (5) Using the interface adsorption equilibrium parameters, scanning tunneling microscopy was used to observe the microscopic morphology of the colloid surface, and the surface adsorption force was measured by atomic force microscopy to obtain the response relationship matrix; (6) Based on the response relationship matrix, the changes in the functional groups on the colloid surface were analyzed by Fourier transform infrared spectroscopy. Combined with the surface coordination theory, a pollutant accumulation model of the interface adsorption effect was constructed.

[0036] Specifically, the soil colloids extracted from the dynamic change characteristic sequence of soil pollutants were subjected to ultrasonic dispersion and centrifugal separation. During the ultrasonic dispersion process, an ultrasonic processor with a power of 500W, a frequency of 20kHz, and a processing time of 30 minutes was selected to ensure that the colloidal particles were fully dispersed. A high-speed centrifuge was used for centrifugal separation, the speed was set to 10,000rpm, the centrifugal time was 15 minutes, and particles with a particle size greater than 2μm were separated by sedimentation. The obtained colloidal suspension was measured for particle size distribution using a laser particle size analyzer, the laser wavelength was 633nm, the measurement range was 0.1-1000μm, the measurement was repeated 3 times and the average value was taken to obtain the particle size distribution curve and characteristic parameters of the soil colloid. According to the obtained soil colloid characteristic parameters, the capillary zone electrophoresis technology was used to separate the colloid surface charge. The capillary length was 60cm, the inner diameter was 75μm, the buffer solution was a phosphate buffer with a pH of 7.0, and the electric field strength was set to 300V / cm. During the electrophoresis process, the conductivity detector was used to monitor the migration of each component in real time, and the migration time and peak area of ​​different components were recorded. The response time of the conductivity detector was set to 0.1s and the sensitivity was , the detection data are processed by baseline correction and peak area integration to form a complete colloidal electrophoretic migration spectrum.

[0037] Subsequently, based on the separation effect of each component in the colloidal electrophoretic migration spectrum, the content of colloidal bound pollutants was determined using an inductively coupled plasma mass spectrometer. The plasma temperature was controlled at 6000-8000K, the radio frequency power was 1300W, and the sampling depth was 10mm. Tandem mass spectrometry analysis adopted the multiple reaction monitoring mode, the collision gas pressure was 2.5mTorr, and the residence time was set to 100ms. Quantitative analysis was performed by measuring the peak intensity ratio of the characteristic ion pairs, and the matrix effect was corrected by the standard addition method, and finally the accurate content data of the colloidal adsorbed pollutants were obtained. The colloidal adsorbed pollutant data were substituted into the Langmuir isothermal adsorption equation to calculate the interface equilibrium constant and the maximum adsorption capacity. The adsorption experiment was carried out in a constant temperature oscillator, the temperature was controlled at 25±0.5℃, the oscillation speed was 150rpm, and the equilibrium time was 24 hours. By changing the initial pollutant concentration (0.1-10mg / L), the distribution of pollutants in the solution and solid phase under equilibrium was determined, the isothermal adsorption curve was fitted, and the relevant parameters were calculated.

[0038] When using scanning tunneling microscopy to observe the microscopic morphology of the colloid surface, the sample preparation adopts ultra-thin coating technology, the scanning range is 500nm×500nm, the tunneling current is set to 2nA, and the bias voltage is 50mV. The atomic force microscope uses the contact mode to measure the surface adsorption force, the probe elastic constant is 0.1N / m, and the scanning rate is 1Hz. The surface adsorption characteristics are analyzed by force-distance curves, and the response relationship matrix is ​​constructed by combining the morphology and mechanical data. Finally, based on the response relationship matrix, the Fourier transform infrared spectrometer is used to characterize the functional groups on the colloid surface. The scanning wavenumber range is , with a resolution of , scanning times 32. By analyzing the position and intensity changes of characteristic absorption peaks, combined with surface coordination theory, the binding mechanism between pollutants and active sites on the colloid surface is analyzed.

[0039] For example: After the soil colloid was dispersed by ultrasound and separated by centrifugation, laser particle size analysis showed that the median particle size of the colloid was 0.8μm and the distribution range was 0.2-2μm. Capillary electrophoresis separation obtained three characteristic peaks, and the migration time of the main peak was 8.5min, corresponding to the negatively charged silicon aluminum oxide colloid. Inductively coupled plasma mass spectrometry analysis showed that cadmium mainly existed in a bound form with a content of 15.6mg / kg. The maximum adsorption capacity calculated by the Langmuir isotherm adsorption experiment was 42.5mg / g. Scanning tunneling microscopy observed that the surface of the colloid showed a flaky structure, and atomic force microscopy measured the surface adsorption force to be 2.8nN. Fourier transform infrared spectroscopy and The hydroxyl characteristic peak appears at , indicating that cadmium ions are mainly adsorbed through the coordination of surface hydroxyl groups. These data together constitute a complete model of pollutant accumulation due to interface adsorption effect.

[0040] In a specific embodiment, the process of executing step S105 may specifically include the following steps: (1) Based on the pollutant accumulation model, the soil samples were heated using a differential scanning calorimeter, and the heat flow during the temperature change process was continuously monitored to obtain the pollutant pyrolysis characteristic curve; (2) Based on the pyrolysis characteristic curve of pollutants, the crystal phase structure of soil samples is determined by X-ray diffractometer, and the diffraction peaks are analyzed in combination with the Bragg equation to obtain lattice parameter data; (3) Based on the lattice parameter data, the concentrations of potassium ions, sodium ions, and ammonium ions are potentiometrically measured using selective ion electrodes, and the ion concentrations are calculated using potentiometric titration curves to form key ion concentration data; (4) Based on the key ion concentration data, a thermogravimetric analyzer is used to measure the variation of sample mass with temperature, and the thermal decomposition products of pollutants are analyzed by thermogravimetric-mass spectrometry to obtain the composition data of the thermal decomposition products; (5) The composition data of the thermal decomposition products are used to determine the valence state of the elements on the sample surface by X-ray photoelectron spectroscopy, and the surface chemical composition is analyzed by Auger electron spectroscopy to obtain a surface chemical state distribution map; (6) Based on the surface chemical state distribution map, Raman spectrometry is used to characterize the molecular structure of pollutants, and nuclear magnetic resonance spectroscopy is used to analyze the structure of organic components to generate characteristic maps of multi-pollutant interactions.

[0041] Specifically, based on the pollutant accumulation model, the soil samples were first thermally analyzed using a differential scanning calorimeter. Differential scanning calorimetry is a method for studying the thermal behavior of a material by measuring the changes in the heat absorption or heat release of the sample during a programmed temperature increase. During the analysis, the soil sample was placed in a platinum crucible and heated from room temperature to 800°C at a rate of 10°C / min, with a nitrogen atmosphere protection and a gas flow rate of 50mL / min, while recording temperature and heat flow data. The heat flow data was collected in real time by a high-precision temperature difference sensor built into the calorimeter, with a sampling interval of 0.1s, and a DSC curve reflecting the pyrolysis characteristics of the pollutant was obtained. After analyzing the obtained pyrolysis characteristic curve, an X-ray diffractometer was used to determine the crystal phase structure of the soil sample. The X-ray diffraction analysis used a Cu target X-ray tube with an operating voltage of 40kV, a current of 30mA, a scanning step of 0.02°, a scanning speed of 4° / min, and a scanning range of 5-80°. The diffraction peaks were analyzed by the Bragg equation to determine the crystal plane spacing and unit cell parameters of the sample. The diffraction pattern was processed by background subtraction and peak fitting, and the phase qualitative analysis was performed using Jade software to obtain complete lattice parameter data.

[0042] According to the lattice parameter data, the concentration of key ions in the soil samples was determined. The selective ion electrode determination included three target ions: potassium ion, sodium ion and ammonium ion. A special ion selective electrode was selected for each ion, and a potential measurement system was established with a silver / silver chloride reference electrode. During the potentiometric titration process, an automatic potentiometric titrator was used to control the titration rate, and a potential value was recorded every 0.1 mL. The potential-volume curve was drawn, and the ion concentration was calculated by the inflection point of the curve. Using the key ion concentration data obtained, a thermogravimetric analyzer was further used to study the thermal decomposition behavior of the sample. The thermogravimetric analysis was carried out in a nitrogen atmosphere with a gas flow rate of 60 mL / min, a heating rate of 20℃ / min, and a temperature range of room temperature to 1000℃. The thermogravimetric-mass spectrometry technique connected the thermogravimetric analyzer to the mass spectrometer through a transmission line, and the transmission line temperature was maintained at 200℃ to prevent sample condensation. The mass spectrometer used an electron bombardment source with an ionization voltage of 70 eV and a scanning mass range of m / z 1-300 to obtain a real-time mass spectrum of the thermal decomposition products.

[0043] After importing the composition data of the thermal decomposition products, an X-ray photoelectron spectrometer was used to analyze the elemental valence state of the sample surface. The X-ray photoelectron spectroscopy analysis used a monochromatic Al Kα ray source with a power of 150W and a vacuum degree better than The binding energy information of surface elements is obtained by full spectrum scanning and narrow scanning, and the surface chemical composition is analyzed by combining Auger electron spectroscopy. Auger electron spectroscopy analysis uses a 5keV electron beam, and the detector works in a constant analysis energy mode.

[0044] Finally, Raman spectrometer and nuclear magnetic resonance spectrometer were used to characterize the molecular structure of the pollutants. Raman spectroscopy uses 532nm laser as the excitation light source, the laser power is 50mW, and the spectral resolution is , the scanning range is The nuclear magnetic resonance spectroscopy analysis was performed using a 600 MHz superconducting nuclear magnetic resonance spectrometer. The samples were dissolved in deuterated solvents and the ¹H and ¹³C spectra were collected.

[0045] For example: Through differential scanning calorimetry analysis, two exothermic peaks were found at 268°C and 425°C, corresponding to the thermal decomposition process of different types of PAHs. X-ray diffraction analysis showed the presence of mineral phases such as quartz and kaolinite in soil samples, and lattice parameter analysis determined that heavy metals mainly existed in the form of oxides. Selective ion electrode determination showed that the potassium ion concentration was 156 mg / L, the sodium ion concentration was 89 mg / L, and the ammonium ion concentration was 34 mg / L. Thermogravimetric-mass spectrometry analysis showed that a series of characteristic fragment ions were detected in the temperature range of 250-450°C, confirming the existence of PAHs. X-ray photoelectron spectroscopy and Auger electron spectroscopy analysis determined the valence information of heavy metals, and combined with Raman spectroscopy and nuclear magnetic resonance spectroscopy data, the interaction mechanism between pollutants was clarified, providing a scientific basis for pollution prevention and control.

[0046] In a specific embodiment, the process of executing step S106 may specifically include the following steps: (1) Based on the characteristic spectrum of multi-pollutant interactions, the soil samples were subjected to programmed temperature treatment using a thermogravimetric-differential thermal analyzer, and thermogravimetric and differential thermal curves were collected to obtain pollutant content change data; (2) Based on the pollutant content change data, the absorbance of the heavy metal color reaction is measured using a visible spectrophotometer, and the pollutant concentration is calculated in combination with the standard curve to obtain a quantitative analysis data set; (3) The quantitative analysis data set is used to scan the entire band of the sample through an infrared spectrometer, identify and quantitatively analyze the characteristic absorption peaks, and form spectral characteristic data; (4) Using the spectral characteristic data, the absorption spectrum of the sample is measured by UV-visible spectrometer, and the background interference is eliminated by combining the second-order derivative spectrum to obtain the spectrum of the pollutant components; (5) For the pollutant component spectrum, the excitation-emission spectrum of the sample is measured by a fluorescence analyzer, and the spectral data is analyzed by parallel factor analysis to obtain the pollutant concentration warning threshold; (6) Based on the pollutant concentration warning threshold, the fiber optic spectrometer is used to conduct in-situ detection of soil samples, and the pollutant distribution characteristics are established in combination with chemometric analysis to output regional soil pollution prevention and control plans.

[0047] Specifically, based on the characteristic spectrum of multi-pollutant interaction, the soil samples were analyzed using a thermogravimetric-differential thermal synchronous analyzer. The analyzer measures the sample mass change (thermogravimetric curve) and temperature change (differential thermal curve) through programmed temperature treatment. The temperature rise process uses precise temperature control, rising from room temperature to 1000°C at a rate of 10°C / min, with nitrogen atmosphere protection and a flow rate of 60mL / min. The sample mass loss and heat change are recorded in real time. The thermogravimetric curve reflects the mass change during the sample decomposition process, and the differential thermal curve shows the energy change in the endothermic or exothermic process. According to the obtained pollutant content change data, the heavy metal color determination is performed using a visible spectrophotometer. The color reaction uses a heavy metal-specific color developer, such as copper ions and sodium diethyldithiocarbamate to form a yellow complex, and the absorbance is measured at a wavelength of 460nm. The standard curve is drawn using a series of standard solutions with a concentration range of 0.1-5.0mg / L. The absorbance values ​​at each concentration point are measured, and the standard curve equation is obtained by least squares fitting to calculate the actual concentration of pollutants in the sample.

[0048] The quantitative analysis data set was input into the Fourier transform infrared spectrometer for full-band scanning analysis. The scanning wave number range was , with a resolution of , 32 cumulative scans were performed to improve the signal-to-noise ratio. The obtained infrared spectrum was baseline corrected and smoothed, and the characteristic absorption peaks were identified through the peak position database to determine the attribution of each functional group. The area integration method was used to quantitatively analyze the characteristic peaks and calculate the relative content of different functional groups. The obtained spectral characteristic data were further used for UV-visible spectral analysis. The UV-visible spectrometer was measured using a quartz cuvette with a scanning range of 190-800nm, a scanning speed of 300nm / min, and a slit width of 2nm. In order to eliminate matrix interference, the original spectrum was processed by the second-order derivative. The second-order derivative spectrum can effectively eliminate the influence of baseline drift and background absorption to obtain the pollutant component spectrum, such as Figure 2 As shown. It highlights the subtle differences in spectra and improves the accuracy of substance identification.

[0049] The obtained pollutant component spectrum was measured by excitation-emission spectrum using a fluorescence analyzer. The excitation wavelength range was 200-400nm, the emission wavelength range was 250-600nm, and the scanning interval was 5nm. The three-dimensional fluorescence data was analyzed by parallel factor analysis, which is a multidimensional data analysis method that can decompose complex spectral signals into independent components and give the relative contribution of each component. The pollutant concentration warning threshold was set according to the spectral characteristics and concentration distribution of each component.

[0050] Finally, the soil samples were tested in situ using a fiber optic spectrometer. The fiber optic spectrometer uses a reflection probe to directly contact the soil and collect reflection spectrum data in real time. Through chemometric methods such as partial least squares regression and principal component analysis, a quantitative relationship between spectral data and pollutant content was established to draw a spatial distribution map of pollutants.

[0051] For example, through thermogravimetric-differential thermal analysis, it was found that there was an obvious weight loss and exothermic peak around 350°C, corresponding to the decomposition of chromium oxide. After chromium reacted with the color developer diphenylcarbazide, the absorbance measured at 540nm was 0.625. The chromium content was 2.8mg / kg when compared with the standard curve. Infrared spectroscopy analysis The characteristic peak of chromate appeared at 372nm, and the relative content of hexavalent chromium was determined by peak area integration. The UV-visible spectrum showed a characteristic peak at 372nm after second-order derivative processing, confirming the presence of hexavalent chromium. Three-dimensional fluorescence spectroscopy combined with parallel factor analysis identified three main forms of chromium compounds and determined that an early warning would be triggered when the total chromium content exceeded 3.0mg / kg. Finally, through in-situ detection by fiber optic spectrometer, a pollution distribution map was drawn, showing that the chromium content in the eastern area of ​​the farmland was high and needed to be treated as a priority.

[0052] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0053] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for predicting agricultural soil pollution, characterized in that: The agricultural soil pollution prediction method comprises: Through stratified grid sampling, soil samples in agricultural areas are tested for physical and chemical properties at multiple points. Heavy metal content, organic matter content and enzyme activity tests are performed on soil samples. Combined with ion chromatography and atomic absorption spectroscopy, a soil pollution characteristic data set is obtained. Based on the soil pollution characteristic data set, an electrochemical sensor detection array is used to dynamically monitor the diffusion characteristics of pollutants in soil samples, and a pollutant migration law characteristic model is generated by combining dielectric property analysis and fluorescence spectrum detection; Based on the characteristic model of pollutant migration law, a multi-parameter online monitoring system is used to perform real-time detection of soil pH value, redox potential and ion activity, and a dynamic change characteristic sequence of soil pollutants is formed through multi-channel signal acquisition and data correction; According to the dynamic change characteristic sequence of soil pollutants, a response relationship matrix between pollutant concentration and soil colloid interface interaction is established, and a pollutant accumulation model considering the interface adsorption effect is constructed by combining capillary electrophoresis analysis and mass spectrometry detection; Using the pollutant accumulation model, the occurrence forms of pollutants in the soil are determined by synchronous thermal analysis and X-ray diffraction methods, and the changes in key ion concentrations are detected using a selective ion electrode array to generate a multi-pollutant interaction characteristic map; According to the characteristic map of multi-pollutant interactions, a pollutant content early warning system is constructed based on thermogravimetric analysis and colorimetry, and combined with spectral analysis methods, a regional soil pollution prevention and control plan is output.

2. The agricultural soil pollution prediction method according to claim 1, characterized in that: The soil samples in agricultural areas are sampled by stratified grid sampling to detect physical and chemical properties at multiple points, and heavy metal content, organic matter content and enzyme activity tests are performed on the soil samples. Combined with ion chromatography and atomic absorption spectroscopy, a soil pollution characteristic data set is obtained, including: The agricultural area is divided into multiple 10m×10m detection units by presetting the sampling point spacing, and the sampling point array is formed by the center point, vertex and edge midpoint of the detection unit to obtain the initial sampling point distribution data; According to the initial sampling point data, soil samples are taken at three depths of 0-20 cm, 20-40 cm, and 40-60 cm at each sampling point, and the soil samples are mixed and homogenized using the quartering method to obtain a stratified soil sample set; The contents of copper, zinc, lead and cadmium in the stratified soil samples were determined by flame atomic absorption spectrophotometry, and the contents of arsenic and mercury were determined by inductively coupled plasma mass spectrometry to obtain heavy metal content data; The stratified soil sample set is digested by potassium dichromate oxidation-external heating method, the organic matter content is determined by ultraviolet spectrophotometer, and the soil urease activity is determined by alkaline hydrolysis method to obtain soil basic characteristic data; Based on the heavy metal content data and soil basic characteristic data, the sulfate, nitrate and chloride ion contents in the soil samples are quantitatively analyzed by ion chromatography, and the sodium, potassium, calcium and magnesium ions are detected by atomic absorption spectrometry to obtain ion composition characteristic data; The heavy metal content data, soil basic property data and ion composition characteristic data are subjected to data standardization and outlier elimination, and the characteristic weights are determined by principal component analysis to obtain the soil pollution characteristic data set.

3. The agricultural soil pollution prediction method according to claim 1, characterized in that: Based on the soil pollution characteristic data set, the diffusion characteristics of pollutants in soil samples are dynamically monitored by using an electrochemical sensor detection array, and a pollutant migration law characteristic model is generated by combining dielectric property analysis and fluorescence spectrum detection, including: According to the soil pollution characteristic data set, an electrochemical sensor detection array consisting of ion selective electrodes, conductivity electrodes and redox electrodes is arranged at the sampling points, and the concentration of heavy metal ions is continuously collected by the potentiometric method to obtain the pollutant concentration time series data; Based on the pollutant concentration time series data, a four-electrode AC impedance spectrometer is used to measure the dielectric constant and dielectric loss of the soil sample, and a soil dielectric parameter set is obtained by combining the polarization characteristic analysis under the power frequency electric field; The soil sample is subjected to excitation-emission matrix scanning by a fluorescence spectrophotometer, and a fluorescence characteristic spectrum of organic pollutants is formed by combining peak position and peak intensity analysis of a three-dimensional fluorescence spectrum; The soil dielectric parameter set is fused with the fluorescence characteristic spectrum of the organic pollutant, and the pollutant migration characteristics are extracted by covariance matrix decomposition to obtain pollutant diffusion characteristic data; Based on the pollutant diffusion characteristic data, the surface morphology of soil particles is observed using a scanning electron microscope, and the surface element distribution is measured with an energy spectrum analyzer to obtain the pollutant interface distribution characteristics; Based on the interface distribution characteristics of the pollutants, through the calculation of anisotropic diffusion coefficients and analysis of interface mass transfer resistance, combined with the migration rate of pollutants in the soil, a characteristic model of the pollutant migration law is generated.

4. The agricultural soil pollution prediction method according to claim 1, characterized in that: According to the characteristic model of pollutant migration law, a multi-parameter online monitoring system is used to perform real-time detection of soil pH value, redox potential and ion activity, and a dynamic change characteristic sequence of soil pollutants is formed through multi-channel signal acquisition and data correction, including: Based on the characteristic model of pollutant migration law, a composite glass electrode, a platinum electrode and an ion selective electrode are installed at the sampling point, and the electrode output signal is amplified and filtered by a signal conditioning circuit to obtain the initial data of soil chemical parameters; Using the initial data of soil chemical parameters, three-point calibration and temperature compensation are performed on the measuring electrode, and a potential difference measurement is performed in combination with a standard buffer solution to obtain a corrected pH value data set; The potential signal measured by the platinum electrode is collected by an electrochemical workstation, a potential reference is established with a reference electrode, and digital filtering is used to eliminate external interference to form a redox potential data sequence; According to the pH value data group and the redox potential data sequence, the potential difference of the target ion is measured by using an ion selective electrode, and the ion activity is calculated by using the Nernst equation to obtain ion activity distribution data; Inputting the pH value data group, the redox potential data sequence and the ion activity distribution data into a multi-channel data acquisition system, performing a noise reduction process on the signal using wavelet denoising to obtain purified monitoring data; The purified monitoring data is subjected to time series analysis and spatial interpolation processing, and outliers are eliminated by combining adaptive threshold filtering to form a dynamic change characteristic sequence of soil pollutants.

5. The agricultural soil pollution prediction method according to claim 1, characterized in that: According to the soil pollutant dynamic change characteristic sequence, a response relationship matrix between pollutant concentration and soil colloid interface interaction is established, and a pollutant accumulation model considering the interface adsorption effect is constructed by combining capillary electrophoresis analysis and mass spectrometry detection, including: From the soil pollutant dynamic change characteristic sequence, the soil colloid is ultrasonically dispersed and centrifugally separated, and the colloid particle size distribution is measured by a laser particle size analyzer to obtain the soil colloid characteristic parameters; According to the soil colloid characteristic parameters, the surface charge of the soil colloid is separated by capillary zone electrophoresis, and the migration time of each component is measured in combination with a conductivity detector to obtain a colloid electrophoresis migration spectrum; According to the colloid electrophoretic migration spectrum, the content of colloid-bound pollutants is determined by using an inductively coupled plasma mass spectrometer, and the pollutants are qualitatively and quantitatively analyzed by tandem mass spectrometry to form colloid-adsorbed pollutant data; The colloid adsorption pollutant data is input into the isothermal adsorption equation, and the interface equilibrium constant and the maximum adsorption capacity are calculated by Langmuir adsorption theory to obtain the interface adsorption equilibrium parameters; Using the interface adsorption equilibrium parameters, a scanning tunneling microscope is used to observe the microscopic morphology of the colloid surface, and an atomic force microscope is used to measure the surface adsorption force to obtain a response relationship matrix; Based on the response relationship matrix, the changes of functional groups on the colloid surface are analyzed by Fourier transform infrared spectroscopy, and combined with surface coordination theory, a pollutant accumulation model of the interface adsorption effect is constructed.

6. The agricultural soil pollution prediction method according to claim 1, characterized in that: The pollutant accumulation model is used to determine the occurrence form of pollutants in soil by synchronous thermal analysis and X-ray diffraction methods, and a selective ion electrode array is used to detect changes in key ion concentrations to generate a multi-pollutant interaction characteristic map, including: Based on the pollutant accumulation model, the soil sample is heated by a differential scanning calorimeter, and the heat flow during the temperature change is continuously monitored to obtain the pollutant pyrolysis characteristic curve; For the pyrolysis characteristic curve of the pollutant, the crystal phase structure of the soil sample is measured by using an X-ray diffractometer, and the diffraction peak is analyzed in combination with the Bragg equation to obtain lattice parameter data; According to the lattice parameter data, the concentrations of potassium ions, sodium ions and ammonium ions are potentiometrically measured using a selective ion electrode, and the ion concentrations are calculated through a potentiometric titration curve to form key ion concentration data; According to the key ion concentration data, a thermogravimetric analyzer is used to measure the variation of sample mass with temperature, and the thermal decomposition products of the pollutants are analyzed by thermogravimetric-mass spectrometry to obtain the composition data of the thermal decomposition products; The composition data of the thermal decomposition products are used to determine the valence state of the elements on the surface of the sample by an X-ray photoelectron spectrometer, and the surface chemical composition is analyzed by an Auger electron spectroscopy to obtain a surface chemical state distribution map; Based on the surface chemical state distribution map, a Raman spectrometer is used to characterize the molecular structure of the pollutants, and a nuclear magnetic resonance spectrum is used to analyze the structure of the organic components to generate the multi-pollutant interaction characteristic map.

7. The agricultural soil pollution prediction method according to claim 1, characterized in that: According to the characteristic spectrum of multi-pollutant interaction, a pollutant content early warning system is constructed based on thermogravimetric analysis and colorimetry, and a regional soil pollution prevention and control plan is output in combination with a spectral analysis method, including: Based on the multi-pollutant interaction characteristic map, the soil sample is subjected to programmed temperature treatment by a thermogravimetric-differential thermal synchronous analyzer, and thermogravimetric curves and differential thermal curves are collected to obtain pollutant content change data; According to the pollutant content change data, the absorbance of the heavy metal color development reaction is measured using a visible spectrophotometer, and the pollutant concentration is calculated in combination with a standard curve to obtain a quantitative analysis data set; The quantitative analysis data set is used to perform full-band scanning of the sample through an infrared spectrometer, and characteristic absorption peaks are identified and quantitatively analyzed to form spectral characteristic data; Using the spectral characteristic data, an ultraviolet-visible spectrometer is used to measure the absorption spectrum of the sample, and the background interference is eliminated by combining the second-order derivative spectrum to obtain a spectrum diagram of the pollutant components; For the pollutant component spectrum, the excitation-emission spectrum of the sample is measured by a fluorescence analyzer, and the spectrum data is analyzed in combination with parallel factor analysis to obtain the pollutant concentration warning threshold; Based on the pollutant concentration warning threshold, the fiber optic spectrometer is used to perform in-situ detection of soil samples, and the pollutant distribution characteristics are established in conjunction with chemometric analysis to output the regional soil pollution prevention and control plan.

Images