Soil geology catalysis capability prediction method and related equipment

Abstract

The embodiment of the invention provides a soil geological catalytic capability prediction method and related equipment, and belongs to the technical field of machine learning. The method comprises the following steps: acquiring redox active component data of to-be-predicted soil; the oxidation-reduction active component data are input into the trained prediction model for catalytic capability prediction, the electron accepting capability index and the electron supply capability index of the to-be-predicted soil are obtained, and the electron accepting capability index and the electron supply capability index are used for representing the soil geological catalytic capability. According to the embodiment of the invention, the evaluation efficiency and accuracy of the soil geological catalytic capability can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of machine learning, and in particular to a soil geocatalytic capacity prediction method and related equipment. BACKGROUND

[0002] As the largest "biocatalytic reactor" in the critical zone of the earth, soil directly regulates the biogeochemical cycle through its redox catalytic capacity. Soil Electron Transfer Capacity (ETC) is a key micro index representing the redox catalytic behavior of soil, and has a profound impact on the migration, transformation and fate of pollutants such as heavy metals, as well as the emission of greenhouse gases such as carbon dioxide (CO2) and methane (CH4). Therefore, accurately assessing soil ETC is an important basis for scientifically predicting environmental risks, developing precise agricultural management measures, and implementing remediation strategies for contaminated land.

[0003] In related technologies, the quantitative determination of soil electron transfer capacity mainly relies on electrochemical methods and chemical probe methods. Although these methods can directly measure the electron accepting and supplying capacity of soil, they generally have problems such as high cost, complex operation, long time-consuming, and are difficult to meet the actual needs of large-scale, regional soil health surveys and rapid diagnosis. These limitations have become a technical bottleneck for current soil function evaluation and precise risk control. In addition, related technologies lack effective methods for predicting geocatalytic capacity.

[0004] In summary, the technical problems in related technologies need to be improved. SUMMARY

[0005] The main purpose of the embodiments of the present application is to provide a soil geocatalytic capacity prediction method and related equipment, which can improve the evaluation efficiency and accuracy of soil geocatalytic capacity.

[0006] To achieve the above-mentioned purpose, one aspect of the embodiments of the present application provides a soil geocatalytic capacity prediction method, which comprises: obtaining redox active component data of a soil to be predicted; inputting the redox active component data into a trained prediction model for catalytic capacity prediction to obtain an electron accepting capacity index and an electron supplying capacity index of the soil to be predicted, the electron accepting capacity index and the electron supplying capacity index being used to represent the soil geocatalytic capacity.

[0007] In some embodiments, the trained prediction model is obtained by the following steps: obtaining a soil data set, the soil data set comprising a training set and a validation set; inputting redox-active component data of the training set into a plurality of different machine learning models respectively to obtain predicted values of each of the machine learning models; calculating a performance index according to the predicted values and true values of the training set to obtain a first determination coefficient and a first root mean square error of each of the machine learning models; adjusting hyperparameters of the plurality of machine learning models respectively according to the first determination coefficient and the first root mean square error to obtain a plurality of machine learning models after tuning; inputting the validation set into the plurality of machine learning models after tuning respectively to perform performance verification to obtain a second determination coefficient and a second root mean square error; selecting from the plurality of machine learning models after tuning according to the second determination coefficient and the second root mean square error to determine the trained prediction model.

[0008] In some embodiments, the soil data set includes redox-active component data, electron acceptance index and electron supply index, and the soil data set is obtained by the following steps: measuring soil samples by a measuring instrument to obtain the redox-active component data; measuring soil samples by a mediated electrochemistry method to obtain the electron acceptance index and the electron supply index.

[0009] In some embodiments, the measuring soil samples by a mediated electrochemistry method to obtain the electron acceptance index and the electron supply index includes: obtaining current data of electrochemical reactions of the soil samples, the current data including oxidation current and reduction current; calculating according to electrochemical reaction time and the oxidation current to obtain the electron supply index; calculating according to electrochemical reaction time and the reduction current to obtain the electron acceptance index.

[0010] In some embodiments, before the redox-active component data of the training set is inputted into a plurality of different machine learning models respectively, the method further includes preprocessing the soil data set, and the preprocessing the soil data set includes: performing unit normalization processing on the soil data set to obtain a first data set; performing feature screening on the first data set to obtain a second data set; performing standardization processing on the second data set to obtain the soil data set after preprocessing.

[0011] In some embodiments, the method further includes: Feature importance analysis is performed on the trained prediction model to obtain the contribution of different components in the redox active component data to the electron accepting capacity index and the electron supply capacity index.

[0012] In some embodiments, the method further includes: Based on the electron accepting capacity index and the electron supply capacity index, a spatial distribution map of soil geological catalytic capacity is obtained by visualizing the data.

[0013] To achieve the above objectives, another aspect of this application provides a soil geological catalytic capacity prediction device, the device comprising: The acquisition module is used to acquire data on the redox active components of the soil to be predicted; The prediction module is used to input the redox active component data into a trained prediction model to predict catalytic capacity, and obtain the electron accepting capacity index and the electron supply capacity index of the soil to be predicted. The electron accepting capacity index and the electron supply capacity index are used to characterize the soil geological catalytic capacity.

[0014] To achieve the above objectives, another aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described above.

[0015] To achieve the above objectives, another aspect of the embodiments of this application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods described above.

[0016] To achieve the above objectives, another aspect of the embodiments of this application proposes a computer program product, including a computer program that, when executed by a processor, implements the aforementioned method.

[0017] The embodiments of this application include at least the following beneficial effects: This application provides a method, device, electronic device, storage medium, and program product for predicting soil geocatalytic capacity. This solution is based on the redox active component data of soil and uses a trained prediction model to predict the soil geocatalytic capacity, obtaining electron accepting capacity and electron supply capacity indicators that characterize the soil geocatalytic capacity. This achieves a leap from "measurement" to "prediction" of soil geocatalytic capacity. Prediction results can be obtained quickly based solely on easily accessible soil data, greatly improving the evaluation efficiency. It is suitable for large-scale applications. The trained prediction model can capture the complex nonlinear relationship between redox active component data and electron accepting capacity and electron supply capacity indicators, thereby improving prediction accuracy. Attached Figure Description

[0018] Figure 1 This is a flowchart of the soil geological catalytic capacity prediction method provided in the embodiments of this application; Figure 2 This is a complete implementation flowchart of the soil geological catalytic capacity prediction method provided in the embodiments of this application; Figure 3 This is a fitting graph of the true values ​​and predicted values ​​of the prediction model provided in the embodiments of this application on the training set and the validation set, specifically EAC and EDC. Figure 4 This is a SHAP importance analysis diagram of EAC and EDC provided in the embodiments of this application; Figure 5 This is a partial dependency graph of EAC provided in the embodiments of this application; Figure 6 This is a partial dependency graph of EDC provided in the embodiments of this application; Figure 7 This is a linear correlation analysis diagram of arsenic reduction capacity and geocatalytic capacity provided in the embodiments of this application; Figure 8 This is a linear correlation analysis diagram of methane emission potential and geocatalytic capacity provided in the embodiments of this application; Figure 9 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit it. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this application as detailed in the appended claims.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0021] Soil, as the largest "biocatalytic reactor" in the Earth's critical zone, directly regulates biogeochemical cycles through its redox catalytic capacity. Soil electron transport capacity (ETC) is a key microscopic indicator characterizing soil redox catalytic behavior, profoundly impacting the migration, transformation, and fate of pollutants such as heavy metals, as well as the emissions of greenhouse gases such as carbon dioxide (CO2) and methane (CH4). Therefore, accurate assessment of soil ETC is a crucial foundation for scientifically predicting environmental risks, developing precision agricultural management measures, and implementing remediation strategies for contaminated sites.

[0022] Currently, the quantitative determination of soil electron acceptor capacity (ETC) mainly relies on electrochemical methods (such as dielectric electrochemical reduction / oxidation) and chemical probe methods. Although these methods can directly measure the soil's electron acceptor and acceptor capacity, they generally suffer from high costs, complex operations, and long processing times, making it difficult to meet the practical needs of large-scale, regional soil health surveys and rapid diagnosis. These limitations have become the technical bottleneck for current soil function assessment and precise risk management.

[0023] Related studies have shown that redox active components (RACs) such as iron oxides, humus, and sulfides in soil are the main carriers of soil geocatalytic processes. However, there is currently a lack of effective methods to reliably predict geocatalytic capacity based on the content of these RACs. In addition, due to the complex interactions and nonlinear relationships among RACs, traditional statistical models (such as linear regression and principal component analysis) are difficult to capture and describe their inherent nonlinear and high-dimensional coupling relationships.

[0024] In view of this, this application provides a method and related equipment for predicting soil geocatalytic capacity. This method is based on the redox active component data of soil and uses a trained prediction model to predict the soil geocatalytic capacity, obtaining electron accepting capacity and electron supply capacity indicators that characterize the soil geocatalytic capacity. This achieves a leap from "measurement" to "prediction" of soil geocatalytic capacity. Prediction results can be obtained quickly based on easily accessible soil data, which greatly improves the evaluation efficiency and is suitable for large-scale applications. The trained prediction model can capture the complex nonlinear relationship between redox active component data and electron accepting capacity and electron supply capacity indicators, thereby improving the prediction accuracy.

[0025] The method for predicting soil geological catalytic capacity provided in this application relates to the field of machine learning technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle terminal, but is not limited to these. The server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network. The software can be an application that implements the method for predicting soil geological catalytic capacity, but is not limited to the above forms.

[0026] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0027] Figure 1 This is an optional flowchart of the soil geological catalytic capacity prediction method provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S101 to S102.

[0028] Step S101: Obtain data on the redox active components of the soil to be predicted; Step S102: Input the redox active component data into the trained prediction model to predict the catalytic capacity, and obtain the electron accepting capacity index and electron supply capacity index of the soil to be predicted. The electron accepting capacity index and electron supply capacity index are used to characterize the soil geological catalytic capacity.

[0029] In step S101 of some embodiments, the data on redox-active components (RACs) include, but are not limited to, the contents of amorphous iron (Feo), crystalline iron (Fec), free iron oxide (Fed), complexed iron (Fep) and clay iron (FeT-Fed), fulvic acid (FA), humic acid (HA), and humin (HM). RACs can be obtained by chemical extraction from the soil to be predicted. The data on redox-active components, as the main carrier of soil geocatalytic processes, can serve as the basic data for subsequent prediction of soil geocatalytic capacity.

[0030] In step S102 of some embodiments, the trained prediction model can be a machine learning model, GBDT (Gradient Boosting Decision Tree), which can capture the complex nonlinear relationship between RACs and soil electron transfer capacity (ETC), and has higher prediction accuracy than traditional linear models. Electron Accepting Capacity (EAC) and Electron Donating Capacity (EDC) indicators can jointly represent the soil's electron transfer capacity, i.e., its geocatalytic capacity. ETC can be expressed as the sum of EAC and EDC.

[0031] In some embodiments, the trained prediction model can be obtained through, but is not limited to, steps S201 to S206: Step S201: Obtain the soil dataset, which includes a training set and a validation set; Step S202: Input the redox active component data of the training set into multiple different machine learning models to obtain the predicted values ​​of each machine learning model. Step S203: Calculate performance metrics based on predicted values ​​and true values ​​in the training set to obtain the first coefficient of determination and the first root mean square error for each machine learning model; Step S204: Adjust the hyperparameters of multiple machine learning models according to the first coefficient of determination and the first root mean square error to obtain multiple optimized machine learning models. Step S205: Input the validation set into multiple optimized machine learning models respectively for performance verification, and obtain the second coefficient of determination and the second root mean square error. Step S206: Based on the second coefficient of determination and the second root mean square error, select from multiple optimized machine learning models to determine the trained prediction model.

[0032] In some embodiments, the soil dataset includes redox active component data, electron accepting capacity index, and electron supply capacity index, and the soil dataset is obtained through steps S210 to S220: Step S210: The soil sample is measured using a measuring instrument to obtain data on redox active components; Step S220: The soil sample is measured by a dielectric electrochemical method to obtain the electron accepting capacity index and the electron supply capacity index.

[0033] In steps S210 to S220 of some embodiments, multiple soil samples from different regions were collected, and the RAC content of each sample was extracted and determined using chemical extraction methods. This RAC content refers to the content of Feo, Fec, Fed, Fep, FeT-Fed, FA, HA, and HM in the soil. The measuring instrument can be an inductively coupled plasma atomic emission spectrometer (ICP-AES) and a total organic carbon analyzer. The ICP-AES measures the Fe concentration of Feo, Fec, Fed, Fep, and FeT-Fed, while the total organic carbon analyzer measures the carbon content of FA, HA, and HM.

[0034] EAC and EDC in soil were quantified separately by dielectric electrochemical method. The experiment was carried out in an anaerobic electrochemical workstation (CHI-1000C) under N2 atmosphere (purity >99.999%).

[0035] In some embodiments, step S220 may include, but is not limited to, steps S221 to S223: Step S221: Obtain current data of the electrochemical reaction of the soil sample, including oxidation current and reduction current; Step S222: Calculate the electron supply capacity index based on the electrochemical reaction time and oxidation current. Step S223: Calculate the electron accepting capacity index based on the electrochemical reaction time and reduction current.

[0036] In steps S221 to S223 of some embodiments, the anaerobic electrochemical workstation uses a three-electrode system to acquire current data of the electrochemical reaction of the soil sample. The current data includes oxidation current and reduction current that change with electrochemical reaction time. The electron accepting capacity index EAC and the electron supply capacity index EDC are obtained by formulas (1) and (2): (1) (2) in, and These are the baseline-corrected oxidation and reduction currents in MEO (Mediated Electrochemical Oxidation) and MER (Mediated Electrochemical Reduction), respectively; F (96485 [C / mol] e - ]) is the Faraday constant; m is the carbon content in the reaction or the amount of carbon added, in g; t is the reaction time, in s.

[0037] In some embodiments, before inputting the redox active component data of the training set into multiple different machine learning models, it is necessary to preprocess the soil dataset. The preprocessing of the soil dataset may include, but is not limited to, steps S210 to S230: Step S210: Perform unit normalization on the soil dataset to obtain the first dataset; Step S220: Perform feature filtering on the first dataset to obtain the second dataset; Step S230: Standardize the second dataset to obtain the preprocessed soil dataset.

[0038] In steps S210 to S230 of some embodiments, the training dataset is first normalized sequentially using a unified transformation principle; features are selected from the training dataset through characteristic correlation analysis, and Feo, Fec, Fep, FeT-Fed, FA, HA, and HM are retained as model input features; finally, the second dataset is standardized according to the needs of different machine learning models to obtain a preprocessed soil dataset, which is convenient for subsequent model training.

[0039] In step S202 of some embodiments, multiple machine learning models are developed, including but not limited to Random Forest (RF), Extreme Gradient Boosting (XGB), Gradient Boosting Decision Tree (GBDT), k-Nearest Neighbor (kNN), Ridge Regression (RR), Decision Tree Regressor (CART), Support Vector Regression (SVR), Extreme Random Tree Regressor (ETR), Light Gradient Boosting (LGB), and Linear Regression (LR). The training set is input into each model for training to obtain preliminary electron accepting capacity predictions and electron supply capacity predictions.

[0040] In step S203 of some embodiments, the first coefficient of determination is the proportion of the variance between the predicted value and the actual value to the total variance, and the first root mean square error is the mean (RMSE) of the mean square errors between the predicted value and the actual value. The predicted values ​​include predicted electron accepting capacity and predicted electron supply capacity, and the actual values ​​include the actual electron accepting capacity index and electron supply capacity index.

[0041] In step S204 of some embodiments, the first coefficient of determination and the first root mean square error can characterize the training effect of the model. By adjusting the hyperparameters of different machine learning models, the optimal parameter combination for the training effect of each model is determined, thereby obtaining multiple optimized machine learning models as the basis for subsequent screening.

[0042] In some embodiments, machine learning model optimization and selection includes the following steps: inputting the training set into each model for training; using five-fold cross-validation combined with different parameter combinations to evaluate the prediction performance of each model; obtaining the coefficient of determination (R²) of the training results; the coefficient of determination is the proportion of the variance between the predicted value and the true value to the total variance; obtaining the overall loss value of the training process; the overall loss value is the mean squared error (RMSE) between the predicted value and the true value; and selecting the optimal parameter combination for each model based on the coefficient of determination and the overall loss value.

[0043] In steps S205 to S206 of some embodiments, the validation set is input into multiple optimized machine learning models for performance verification, thereby obtaining their respective second coefficient of determination and second root mean square error. Based on the second coefficient of determination and second root mean square error of EAC and EDC, the machine learning model with the best performance is selected as the trained prediction model.

[0044] In some embodiments, the method for predicting soil geological catalytic capacity further includes, but is not limited to, step S103: Step S103: Perform feature importance analysis on the trained prediction model to obtain the contribution of different components in the redox active component data to the electron accepting capacity index and the electron supply capacity index.

[0045] In step S103 of some embodiments, to perform interpretability analysis on the trained prediction model, the trained prediction model is connected to a SHAP interpreter. The SHAP interpreter performs at least one of feature importance analysis, partial dependency graph analysis, and cross-dependency graph analysis on the prediction model to interpret the prediction results. In feature importance analysis, the contribution of different components in the redox active component data to the electron accepting capacity index and the electron supply capacity index is analyzed.

[0046] In some embodiments, the method for predicting the catalytic capacity of soil geochemistry further includes, but is not limited to, step S104: Step S104: Visualize the electron accepting capacity index and electron supply capacity index to obtain a spatial distribution map of soil geological catalytic capacity.

[0047] In step S104 of some embodiments, a spatial distribution map of soil geological catalytic capacity at the regional or national scale is drawn based on ArcGIS software, so that users can intuitively know the soil geological catalytic capacity of different regions.

[0048] In some embodiments, please refer to Figure 2 A method for predicting the catalytic capacity of soil geochemistry includes the following steps: Step 1: Obtain the training dataset, which includes soil redox active components (RACs) data as feature variables, and benchmark measurements of soil electron accepting capacity (EAC) and electron supply capacity (EDC) as target variables. Specifically, 56 soil samples were collected from different major paddy field producing areas. The content of RAC in each sample was extracted and determined by chemical extraction, that is, the content of Feo, Fec, Fed, Fep, FeT-Fed, FA, HA, and HM in the soil. The Fe concentration of Feo, Fec, Fed, Fep, and FeT-Fed was determined by inductively coupled plasma atomic emission spectrometry. The carbon content of FA, HA, and HM was determined by total organic carbon analyzer. EAC and EDC in the soil were quantified by dielectric electrochemical reduction and oxidation methods, respectively. The experiment was carried out in an anaerobic electrochemical workstation (CHI-1000C) under N2 atmosphere (purity > 99.999%) using a three-electrode system. EDC and EAC were obtained by integrating time-current curves and calculated according to equations (1) and (2). Using a unified transformation principle, the training dataset is sequentially normalized; feature selection is performed on the training data through characteristic correlation analysis, and finally Feo, Fec, Fep, FeT-Fed, FA, HA, and HM are retained as model input features.

[0049] Step 2: Build and train various machine learning regression models; Step 3: Use cross-validation and hyperparameter optimization methods to select the optimal parameters for each model; Specifically, during model training and optimization, the scikit-learn and XGBoost libraries are used to construct 10 regression models (RF, XGB, GBDT, kNN, RR, SVR, ETR, LGB, LR, and CART). The dataset obtained in step 1 is randomly divided into training and validation sets in an 8:2 ratio, with 80% of the data used as the training set and 20% as the validation set. The training set is then input into each machine learning model for training. Different parameter values ​​are used sequentially, and five-fold cross-validation is employed to predict the trained models. The predicted values ​​of EAC and EDC output by each algorithm are obtained. The evaluation metrics for the training results are obtained: coefficient of determination (R²) and root mean square error (RMSE). Based on the coefficient of determination and the overall loss value during the training process, the optimal parameter combination for each algorithm is selected.

[0050] In this embodiment, dozens of machine learning algorithms were developed, and the predictive performance of each algorithm was evaluated using the coefficient of determination (R²) and root mean square error (RMSE). Simultaneously, hyperparameter tuning was performed on each algorithm using five-fold cross-validation to improve the prediction accuracy of each algorithm for EAC and EDC.

[0051] Step 4: Evaluate the model's predictive performance using the validation dataset and determine the best model; Specifically, the validation dataset is input into the algorithm with the best parameters for result prediction. When selecting the best-performing algorithm, the validation dataset is input into the algorithm after hyperparameter tuning for result prediction. The R² and RMSE of the algorithm on the validation dataset are obtained. GBDT has the lowest R² of 0.80 (EAC) and 0.87 (EDC) on the validation set, and is selected as the best model with the lowest RMSE. Figure 3 As shown, Figure 3 This is a graph showing the fit between the true and predicted values ​​of EAC and EDC in GBDT on the training and validation sets.

[0052] The best-performing algorithm—the GBDT model—was selected from dozens of algorithms for subsequent extrapolation data prediction. Before making large-scale EAC / EDC predictions, the model was interpreted based on the best-performing algorithm for predicting EAC and EDC to explain how each RAC influences and determines soil EAC and EDC.

[0053] Specifically, performing SHAP analysis on the selected best-performing machine learning algorithm includes the following steps: Define the best machine learning algorithm as the model prediction module and connect it to the SHAP interpreter; The contribution of each RAC is calculated using the SHAP interpreter, such as Figure 4 As shown: FeO and HA make the most significant contributions to EAC and EDC; Partial dependency graphs, such as Figure 5 and Figure 6 As shown, when Fec > 20 mg / g and HA > 3 mg / g, the effect of promoting EAC is significant, while inhibiting the growth of soil EDC, indicating that the content of high crystalline iron and fulvic acid can improve the soil's electron accepting capacity.

[0054] Step 5: Obtain the extrapolation dataset at the regional or national scale; Specifically, when obtaining the national-scale dataset for model extrapolation, the following steps can be performed: Retrieve published literature related to RACs in paddy field soils; Integrate soil attribute parameters from the geographic database; A total of 196,642 RACs data were collected nationwide.

[0055] Step 6: Use the best model to extrapolate the dataset and make predictions, outputting the EAC and EDC results at the corresponding scales; To obtain the national-scale EAC and EDC results output after the best model performs predictions, the following steps can be performed: Nationwide RACs data were input into the GBDT model for prediction; Output the predicted EAC and EDC values ​​for soils nationwide.

[0056] Step 7: Draw a spatial distribution map of soil geological catalytic capacity at the regional or national scale using ArcGIS software.

[0057] In some embodiments, the soil geological catalytic capacity prediction method is applied to arsenic pollution risk prediction, as follows.

[0058] 1. Risk Model Construction: Thirty typical arsenic-contaminated topsoil samples (0–20 cm) from paddy fields were collected and an anaerobic microenvironment was established (25°C, dark, N2 protection) for 7 days. The concentrations of available arsenic (phosphate-extractable As(III) + As(V)) in the soil were measured before and after incubation, and the As(V) reduction rate η - As(V) = (As(V)) / (P) * (P) / P) was calculated. t0 -As(V) t7 ) / As(V) t0 ×100%; qPCR was used to quantify the copy number of the arsenic reduction functional gene arrA. Based on the national paddy field EAC and EDC distribution maps predicted by the GBDT model, a second-level meta-model (linear regression) was constructed to identify "high-risk areas for arsenic activation" (threshold: η-As(V)>60%). 2. Results: The results of the linear analysis are as follows Figure 7As shown, η-As(V) showed a linear negative correlation with soil EAC (r=-0.370) and a linear positive correlation with soil EDC (r=-0.269). The abundance of the arrA gene was significantly negatively correlated with EAC (r=-0.441) and positively correlated with EDC (r=0.311). These relationships confirm that when electron acceptors are sufficient, there is intense electron competition in the high EAC region, inhibiting As(V) reduction; conversely, increased EDC accelerates electron flow to arsenate, promoting the reduction of As(V) to the more mobile As(III). Therefore, a risk level can be determined. Low-risk area: EAC > 0.25 mmol / g soil and EDC < 0.1 mmol / g soil, η-As(V) ≈ 20-40%; High-risk area: EAC≤0.25mmol / gsoil and EDC≥0.1mmol / gsoil, η-As(V)≈60-85%.

[0059] Applying this threshold to 90m raster EAC and EDC layers of paddy fields nationwide can predict arsenic activation hotspots in paddy fields.

[0060] In some embodiments, the soil geological catalytic capacity prediction method is applied to arsenic pollution risk prediction, as follows.

[0061] 1. Risk Model Construction: Thirty typical paddy field topsoil samples (0–20 cm) were collected and an anaerobic microcosm system was established (25°C, dark, N2 protection) for 28 days. Headspace CH4 concentration was determined using gas chromatography-FID, and the cumulative emission per unit dry soil layer (μmol CH4 g) was calculated. -1 Simultaneously, qPCR was used to quantify the copy number of the methane-producing functional gene mcrA. Based on the national paddy field EAC and EDC distribution maps predicted by the GBDT model, a second-level meta-model (linear regression) was constructed to identify "high-risk areas for methane emissions" (threshold: CH4>15umol / gsoil). 2. Results: The results of the linear analysis are as follows Figure 8 As shown, CH4 showed no significant linear correlation with soil EAC, but a significant positive linear correlation with soil EDC (r=0.705). The abundance of the mcrA gene was significantly negatively correlated with EAC (r=-0.404) and positively correlated with EDC (r=0.311). These relationships confirm that when electron acceptors are abundant, soils with high EAC experience intense electron competition, inhibiting the methanogenesis pathway; while when EDC increases, electron donors are surplus, the CO2 / H2 reduction pathway is activated, mcrA expression is enhanced, and CH4 production increases accordingly. Based on this, the risk level can be determined as follows: Low-risk area: EDC < 0.15 mmol / g soil, CH4 < 15 μmol / g soil; High-risk areas: EDC ≥ 0.15 mmol / g soil, CH4 > 15 umol / g soil.

[0062] Applying this threshold to 90 m grid EAC and EDC layers of paddy fields nationwide can predict methane emission hotspots in paddy fields.

[0063] The embodiments of this application have at least the following beneficial effects: 1. The concept of "geocatalytic capacity" was first proposed, and the transformation from traditional electrochemical measurement to intelligent model prediction was realized; 2. For the first time, it has achieved a leap from "measurement" to "prediction" of soil geological catalytic capacity. It does not require expensive instruments and can obtain prediction results quickly and at low cost based solely on easily accessible soil data, which greatly improves the evaluation efficiency and is suitable for large-scale applications. 3. By utilizing machine learning algorithms to capture the complex nonlinear relationship between RACs and ETC, the prediction accuracy is significantly higher than that of traditional linear models; 4. By using SHAP analysis, we can reveal the impact mechanism of key soil components on ETC, thereby enhancing the scientific validity and interpretability of the model; 5. The prediction results can be directly used for environmental risk assessment (such as arsenic pollution and methane emissions), providing technical support for precision agriculture, pollution remediation, and climate change response.

[0064] This application embodiment also provides a soil geological catalytic capacity prediction device, which can implement the above method. The device includes: The acquisition module is used to acquire data on the redox active components of the soil to be predicted; The prediction module is used to input redox active component data into a trained prediction model to predict catalytic capacity, and obtain the electron accepting capacity index and electron supply capacity index of the soil to be predicted. The electron accepting capacity index and electron supply capacity index are used to characterize the soil geological catalytic capacity.

[0065] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0066] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0067] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0068] Please see Figure 9 , Figure 9 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 901 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 902 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 902 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 902 and is called and executed by the processor 901 using the methods described in the embodiments of this application. The input / output interface 903 is used to implement information input and output; The communication interface 904 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 905 transmits information between various components of the device (e.g., processor 901, memory 902, input / output interface 903, and communication interface 904); The processor 901, memory 902, input / output interface 903, and communication interface 904 are connected to each other within the device via bus 905.

[0069] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0070] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0071] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0072] It is understood that the content of the above method embodiments is applicable to the embodiments of this program product. The specific functions implemented by the embodiments of this program product are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0073] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0074] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0075] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0076] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0077] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0078] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0079] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0080] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0081] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0082] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0083] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0084] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for predicting the catalytic capacity of soil geology, characterized in that, The method includes the following steps: Obtain data on the redox active components of the soil to be predicted; The redox active component data are input into a trained prediction model to predict catalytic capacity, thereby obtaining the electron accepting capacity index and the electron supply capacity index of the soil to be predicted. The electron accepting capacity index and the electron supply capacity index are used to characterize the soil geological catalytic capacity.

2. The method according to claim 1, characterized in that, The trained prediction model is obtained through the following steps: Obtain a soil dataset, which includes a training set and a validation set; The redox active component data of the training set are input into multiple different machine learning models to obtain the predicted values ​​of each machine learning model. Calculate performance metrics based on the predicted values ​​and the true values ​​of the training set to obtain the first coefficient of determination and the first root mean square error of each machine learning model. Based on the first determination coefficient and the first root mean square error, the hyperparameters of the multiple machine learning models are adjusted to obtain multiple optimized machine learning models. The validation set is input into multiple optimized machine learning models to verify their performance, and the second coefficient of determination and the second root mean square error are obtained. Based on the second coefficient of determination and the second root mean square error, a selected, trained prediction model is determined from among multiple optimized machine learning models.

3. The method according to claim 2, characterized in that, The soil dataset includes redox active component data, electron accepting capacity indicators, and electron supply capacity indicators. The soil dataset is obtained through the following steps: The soil samples were measured using a measuring instrument to obtain the data of the redox active components; The electron accepting capacity index and the electron supply capacity index were obtained by measuring soil samples using a dielectric electrochemical method.

4. The method according to claim 2, characterized in that, The method of measuring soil samples using dielectric electrochemical methods to obtain the electron accepting capacity index and the electron supply capacity index includes: Acquire current data of the electrochemical reaction of the soil sample, the current data including oxidation current and reduction current; The electron supply capacity index is obtained by calculating based on the electrochemical reaction time and the oxidation current. The electron accepting capacity index is obtained by calculating based on the electrochemical reaction time and the reduction current.

5. The method according to claim 2, characterized in that, Before inputting the redox active component data of the training set into multiple different machine learning models, the method further includes preprocessing the soil dataset, which includes: The soil dataset is normalized to obtain the first dataset; The first dataset is subjected to feature filtering to obtain the second dataset; The second dataset is standardized to obtain the preprocessed soil dataset.

6. The method according to any one of claims 1 to 5, characterized in that, The method further includes: Feature importance analysis is performed on the trained prediction model to obtain the contribution of different components in the redox active component data to the electron accepting capacity index and the electron supply capacity index.

7. The method according to claim 1, characterized in that, The method further includes: Based on the electron accepting capacity index and the electron supply capacity index, a spatial distribution map of soil geological catalytic capacity is obtained by visualizing the data.

8. A soil geological catalytic capacity prediction device, characterized in that, The device includes: The acquisition module is used to acquire data on the redox active components of the soil to be predicted; The prediction module is used to input the redox active component data into a trained prediction model to predict catalytic capacity, and obtain the electron accepting capacity index and the electron supply capacity index of the soil to be predicted. The electron accepting capacity index and the electron supply capacity index are used to characterize the soil geological catalytic capacity.

9. An electronic device, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the method as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images