Method for predicting heavy metal pollution source in soil

Abstract

The invention discloses a method for predicting a heavy metal pollution source in soil, which comprises the following steps: collecting 0-20cm surface soil, and obtaining the heavy metal content of explained variable soil; respectively processing and quantifying the data of the explanatory variables of the pollution sources; the pollution sources comprise heavy industrial pollution sources, agricultural pollution sources, main road traffic pollution sources and natural pollution sources, and the natural pollution sources are from surface water, underground water, rainfall and dust fall; quantized data of all pollution sources serve as explanatory variables of agricultural land soil heavy metal content C and heavy metal to be added into a spatial measurement model, an inverse distance spatial weight matrix is adopted for analysis of heavy metal sources in small-size and regional-range point-shaped soil, and spatial dependency and heterogeneity inspection is conducted on the data. According to the method, heavy metals in soil in a certain area can be predicted according to the model, the prediction effect accords with the actual situation, and technical support is provided for local soil heavy metal pollution prevention and control, monitoring and early warning.

Description

Technical Field

[0001] The present invention belongs to the technical field of environmental detection, and in particular relates to a method for predicting sources of heavy metal pollution in soil. Background Art

[0002] Heavy metal pollution in soil is hidden, long-term, and irreversible. It often accumulates in the soil or transforms into highly toxic compounds and is enriched through the food chain, endangering human health. Cadmium and lead, in particular, can also cause endocrine disorders in the human body. They are defined as environmental hormone substances and have always been a type of pollution factor that has been the focus of attention in soil research. The main sources of heavy metal pollution in soil can be divided into two categories: natural sources and anthropogenic sources. Among them, anthropogenic sources are the core driving factors for the current increase in pollution. Anthropogenic sources are mainly divided into industrial sources, agricultural sources, and traffic pollution sources. Industrial sources mainly involve heavy metals released during the production process of heavy industry enterprises. Since heavy industry enterprises are currently subject to key supervision by the ecological and environmental protection departments, most of the wastewater from pollution sources is discharged through sewage treatment plants, and solid waste and hazardous waste are also strictly monitored and regulated. Therefore, the main way to affect the heavy metal content in the surrounding agricultural land soil is the accumulation and enrichment of heavy metals emitted by pollution sources in the soil through atmospheric deposition.

[0003] Current mainstream soil heavy metal source attribution techniques are based on two main approaches: source emission inventories and receptor models. The source emission inventory approach requires statistical analysis of pollution source emissions and emission factors to establish a pollution source inventory database. The receptor model, however, bypasses the need for detailed information on the composition of individual pollution sources and the transmission of emission factors, directly measuring the receptor environment. Therefore, the receptor model is the primary and most commonly used technique in current pollutant source attribution research.

[0004] At present, the most widely used receptor models at home and abroad are chemical mass balance models and multivariate statistical methods. They require various pollution source data to be included in the model calculation, such as quantitative data on pollution source emission factors (such as emission concentrations and emissions). At present, the problems that the receptor model method needs to solve are: first, data accessibility is low; second, historical pollution source data is generally missing, especially for small and medium-sized enterprises or informal emission sources; third, data uncertainty is high, and emission factors are affected by factors such as production processes and differences in time, making it difficult to accurately quantify. Therefore, emissions cannot accurately reflect the impact on the surrounding ecological environment; fourth, spatial heterogeneity is ignored. Traditional models often use global regression assumptions and cannot capture local differences in the spatial relationship between pollution sources and receptors (such as the impact of topography and hydrology on migration paths).

[0005] In recent years, geographic information systems (GIS) and spatial statistical methods (such as kriging interpolation and spatial autocorrelation analysis) have been introduced into soil pollution research. The main progress includes: spatial distribution visualization: identifying pollution hotspots through heat maps and spatial clustering; preliminary exploration of source-sink relationships: using spatial overlay analysis (such as buffer analysis and layer weight overlay) to qualitatively infer potential pollution sources. The current defects are: (1) Weak qualitative-quantitative conversion. There is a lack of systematic quantitative methods for the qualitative characteristics of pollution source attributes (such as industry type and production scale), resulting in a single explanatory variable in the spatial model (multi-dependent distance attenuation factor); (2) Insufficient dynamic coupling. Spatial econometric models (such as geographically weighted regression GWR and multi-scale models) have a low degree of coupling with pollution source attribute data, making it difficult to analyze the driving mechanism of pollution source characteristics on spatial differentiation; (3) Insufficient integration of multi-source data. There is a gap in the quantitative association between the spatiotemporal dynamic characteristics of pollution sources and environmental migration processes.

[0006] Based on the above analysis, soil heavy metal source tracing research faces three core problems: first, data bottleneck, the contradiction between the scarcity of quantitative data on pollution sources and the model input requirements; second, spatial heterogeneity, the contradiction between the spatial non-stationarity of the pollution process and the traditional homogenization modeling assumption; third, the contradiction between the fragmentation of qualitative attribute information and the structured input of the tracing model. Summary of the Invention

[0007] In order to overcome the shortcomings and deficiencies of the prior art, the present invention aims to provide a method for predicting the sources of heavy metal pollution in soil.

[0008] The present invention is achieved by providing a method for predicting the source of heavy metal pollution in soil, the method comprising the following steps:

[0009] S1. Collect 0-20 cm surface soil and obtain the heavy metal content of the explained variable soil;

[0010] S2. Processing and quantifying the data of the explanatory variables of each pollution source; the pollution sources include heavy industrial pollution sources, agricultural pollution sources, main road traffic pollution sources, and natural pollution sources, wherein the natural pollution sources come from surface water, groundwater, precipitation, and dustfall;

[0011] S3. The quantified data of each pollution source is used as the heavy metal content of agricultural soil C. 重金属 The explanatory variables were added to the spatial econometric model. The inverse distance spatial weight matrix was used to analyze the sources of heavy metals in small-scale, regional point soils. The spatial dependence and heterogeneity of the data were tested, and the spatial econometric model was constructed as follows:

[0012]

[0013] Among them, C 重金属—Concentration of heavy metals in soil at a certain monitoring point, mg / kg;

[0014] a—the impact coefficient of pollution sources from heavy industry enterprises on the heavy metal content in the soil at a certain point;

[0015] Pr i,年产量 —Annual output of the i-th heavy industry enterprise, tons;

[0016] M i,产污系数 —Pollution generation coefficient, i.e. the amount of heavy metal pollutants produced per ton of product by the i-th heavy industry enterprise, kg / ton, derived from the Pollution Generation and Emission Coefficient Manual;

[0017] η i,平均去除效率 —Average removal efficiency of the end-of-pipe treatment technology of the i-th heavy industry enterprise, %, from the Pollution Emission Coefficient Manual;

[0018] E i (0,1)—binary data, indicating whether the i-th heavy industry enterprise emits heavy metal pollutant i through atmospheric deposition, 0 represents no, and 1 represents yes.

[0019] IF i (0,1)—binary data, indicating whether the heavy metal pollutants discharged by the i-th heavy industry enterprise affect the accumulation of heavy metals in the soil at a certain point;

[0020] b—the influence coefficient of agricultural sources on the heavy metal content in soil at a certain location;

[0021] FP—fertilizer and pesticide use in the jurisdiction FP (county), derived from the city's statistical yearbook;

[0022] Area—agricultural land area. The subscript refers to the administrative unit corresponding to the area, square kilometers, derived from remote sensing interpretation vector data.

[0023] c—the influence coefficient of agricultural sources on the heavy metal content in the soil at a certain monitoring point;

[0024] D 主干道 —The distance between a monitoring point and the surrounding main traffic arterial roads within 0.5 km and with a width of ≥30 m, km, derived from the main road vector data of the jurisdiction;

[0025] width—the width of the main road, m, derived from the main road vector data of the jurisdiction;

[0026] P 地表水 —Surface water monitoring data around the monitoring point, where d is the surface water impact coefficient;

[0027] P 地下水 —Groundwater monitoring data around the monitoring point, where e is the groundwater impact coefficient;

[0028] P 降水 —Precipitation monitoring data of the area where the monitoring point belongs, g is the precipitation impact coefficient;

[0029] P 降尘 —Precipitation monitoring data of the area where the monitoring point belongs, h is the dustfall impact coefficient.

[0030] Preferably, step S1 includes the following specific steps:

[0031] The surface soil was dried, ground, and sieved. The soil passing the 2 mm sieve was used for the determination of moisture and pH value, the soil passing the 0.25 mm sieve was used for the determination of total organic carbon, and the soil passing the 0.15 mm sieve was used for the determination of heavy metal elements.

[0032] The measured soil heavy metal C heavy metals, total organic carbon, and pH values ​​were standardized to eliminate the dimension effect;

[0033] The qualitative indicators are assigned to quantitative indicators, wherein the qualitative indicators include the distribution area type SZ, the vegetation type VT, and the administrative division country.

[0034] Preferably, in step S2, the data processing and quantification of heavy industrial pollution sources includes the following steps:

[0035] (1) Relevant data on heavy industrial enterprises were collected from the National Pollutant Discharge Permit Management Information Platform, mainly including administrative divisions, operating time, industry type, longitude and latitude, and types of pollutant emissions. For atmospheric deposition pollution sources, according to the types of pollutant emissions obtained from the enterprises, if a certain heavy metal pollutant is present, it is assigned a value of 1, and if there is no heavy metal pollutant, it is assigned a value of 0; if there is smoke or particulate matter emission, it is assigned a value of 1, and if there is no smoke or particulate matter emission, it is assigned a value of 0, according to the following formula:

[0036] Whether a certain heavy metal is emitted (0,1) = particulate matter or smoke emission (0,1) * heavy metal element pollutants (0,1)

[0037] (2) According to the production volume of enterprise products in the second national pollution source census data, find the pollution coefficient and the average removal efficiency of the end-of-pipe treatment technology, and weight the heavy metal emissions to express the heavy metal emissions. 重金属 , the formula is as follows:

[0038] Emission 重金属 =Pr 年生产量 ×M 产污系数 ×(1-η 平均去除效率 )×E 是否排放 (0,1)

[0039] (3) According to the obtained latitude and longitude of soil heavy metal monitoring point information and heavy metal enterprise point information, convert it into radians. The conversion formula is: The Haversine formula is used to calculate the distance d between two points, as shown below:

[0040]

[0041] Where: φ1 and φ2 are the latitudes of the two points (in radians), φ2-φ1 is the difference in latitude, λ2-λ1 is the difference in longitude, and r is the radius of the Earth.

[0042] (4) Calculate the theoretical impact distance L of heavy industrial enterprises within the jurisdiction using the following formula:

[0043] Theoretical impact distance d 理论 (km) = basic range (industry benchmark value) + Σ (contribution value of each adjustment factor);

[0044] (5) Compare the calculated d with the impact distance L, and convert the result into binary data IF(0,1). If d is greater than L, IF is assigned a value of 0; if it is less than L, IF is assigned a value of 1. The obtained result is combined with the degree of heavy metal emission to obtain the impact factor P of the surrounding heavy industry enterprises on the heavy metals in the soil of a certain agricultural land. 工业源 , the formula is as follows:

[0045]

[0046] Where a is the coefficient of the impact of heavy industry enterprises on the heavy metal content in the soil, n is the total number of heavy industry enterprises around a certain point, and i represents the i-th heavy industry enterprise.

[0047] Preferably, in step S2, the quantitative formula for agricultural pollution sources is:

[0048]

[0049] Where b is the coefficient of the impact of agricultural sources on soil heavy metal content, FP is the amount of fertilizer and pesticide use in the jurisdiction (county), which is derived from the city's statistical yearbook; Area is the area of ​​agricultural land, and the subscripts refer to the administrative unit corresponding to the area, square kilometers, which is derived from remote sensing interpretation vector data.

[0050] Preferably, in step S2, the quantitative formula for the main road traffic pollution source is:

[0051]

[0052] Among them, D 主干道(km) is the distance between a monitoring point and the surrounding traffic arterial road within 0.5 km obtained from ArcMap, width (m) is the width of the arterial road, and c is the influence coefficient of the traffic source.

[0053] Preferably, in step S2, the quantitative formula of the natural pollution source is:

[0054] P 自然源 =d×P 地表水 +e×P 地下水 +f×P 降水 +g×P 降尘

[0055] Among them, surface water data, groundwater data, precipitation, and dustfall data are derived from routine monitoring data of environmental monitoring departments. If there are too many indicators, the indicators can be concentrated, dimensionally reduced, and collinearity reduced through factor analysis or principal component analysis. Dustfall data can be directly standardized, and then the processed data can be included in the model as other sources. After dimension reduction, surface water indicators can be concentrated into inorganic element indicators and eutrophication indicators.

[0056] Preferably, in step S3, the heavy industry pollution source factors, agricultural pollution source factors, and traffic pollution source factors are standardized, and then the natural pollution source data that have been subjected to dimensionality reduction processing are added to construct a spatial measurement model.

[0057] The present invention overcomes the shortcomings of the existing technology and provides a method for predicting the sources of heavy metal pollution in soil. The method proposed in the present invention is innovative at the following levels: first, at the data level, a conversion system is established to transform the qualitative attributes of pollution sources (industry type, spatial location, production years, and types of heavy metal emissions) into quantitative indicators (semantic coding, spatial weight, and time decay coefficient), breaking through the "quantitative data dependence" and integrating multi-source data to enhance the analysis dimension. The city-level statistical yearbook, national pollution source census data, regional vector data, and routine data from environmental monitoring departments are combined for use to break the data bottleneck and ensure research continuity; second, at the model level, the pollution source attributes and the spatial distribution of heavy metals are dynamically associated through a spatial metrology model to achieve "local mechanism explainability"; third, at the application level, a "low data threshold-high spatial accuracy" traceability framework is proposed, which is suitable for small-scale spatial models.

[0058] Compared with the shortcomings and deficiencies of the prior art, the present invention has the following beneficial effects:

[0059] (1) This invention solves the data bottleneck and uses readily available qualitative information (such as enterprise registration information and remote sensing images) to reconstruct pollution source characteristics, fill data gaps, and ensure research continuity.

[0060] (2) The present invention ensures that the model truly reflects the current situation, weakens the negative impact of the uncertainty of the original data (such as pollution source reporting errors) on the model through the data conversion mechanism, and improves the robustness of the traceability results.

[0061] (3) The present invention has high spatial resolution and can accurately capture the nonlinear spatial correlation between pollution source attributes and heavy metal distribution, thereby identifying local pollution contribution sources. In addition, the present invention supports innovative paths for refined management, by associating the current status of soil heavy metal enrichment with the spatial distribution of pollution sources through a model, thereby identifying high-risk areas.

[0062] (4) The present invention can predict the heavy metals in the soil of a certain area based on the model. The prediction effect is more in line with the actual situation, providing technical support for the prevention, control, monitoring and early warning of local soil heavy metal pollution. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is a schematic flow chart of the prediction method of the present invention;

[0064] Figure 2 This is a Moran index diagram of Application Example 1 of the present invention. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0066] Example

[0067] The present invention discloses a method for predicting the source of heavy metal pollution in soil. Figure 1 It is understood that the method comprises the following steps:

[0068] 1. Collection of explained variables: soil heavy metal content

[0069] Soil monitoring data of agricultural land for five years or more is obtained from the environmental monitoring department, or self-monitoring is carried out in accordance with the "Technical Specifications for Soil Environmental Monitoring" (HJ / T 166-2004) (hereinafter referred to as the Technical Specifications). Points are set up according to requirements, and soil from the surface layer (0 to 20 cm) is collected. The parameters recorded during sampling include point number, administrative division, year, point area type, vegetation type, longitude and latitude, etc.

[0070] After drying, the soil was ground and sieved according to the technical specifications. The 2mm sieve was used for moisture and pH value, the 0.25mm sieve was used for total organic carbon determination, and the 0.15mm sieve was used for heavy metal element determination. The determination was carried out according to the requirements of the technical specifications. 重金属The sampling of total organic carbon (TOC) and pH value was standardized to eliminate the dimension effect. Qualitative indicators such as the site type (SZ), vegetation type (VT), and administrative division (country) were assigned quantitative indicators.

[0071] 2. Data Processing and Quantification of Explanatory Variables

[0072] 1. Data processing and quantification of pollution sources in heavy industries

[0073] Relevant data on enterprises in heavy industries are collected from the National Pollutant Discharge Permit Management Information Platform, mainly including administrative divisions (based on county-level cities), operating time, industry type, longitude and latitude, and types of pollutant emissions. Currently, most wastewater from enterprises in heavy industries is connected to the municipal pipe network. Therefore, this technology mainly targets atmospheric deposition pollution sources. Based on the types of pollutant emissions obtained from the enterprises, if a certain heavy metal pollutant is present, the value is assigned to 1; if there is smoke or particulate matter emission, the value is assigned to 1. Whether a certain heavy metal is emitted is determined according to the following formula:

[0074] Whether a certain heavy metal is emitted (0,1) = particulate matter or smoke emission (0,1) * heavy metal element pollutants (0,1)

[0075] According to the production volume of enterprise products in the second national pollution source census data, the pollution coefficient and the average removal efficiency of the end-of-pipe treatment technology are found, and the heavy metal emissions are weighted to express the heavy metal emissions. 重金属 , the formula is as follows:

[0076] Emission 重金属 =Pr 年生产量 ×M 产污系数 ×(1-η 平均去除效率 )×E 是否排放 (0,1)

[0077] According to the obtained latitude and longitude of the soil heavy metal monitoring point information and the heavy metal enterprise point information, it is converted into radians. The conversion formula is: The Haversine formula is used to calculate the distance d between two points, as shown below:

[0078]

[0079] Where: φ1 and φ2 are the latitudes of the two points (in radians), φ2-φ1 is the latitude difference, λ2-λ1 is the longitude difference, and r is the radius of the Earth, which is approximately 6371 kilometers.

[0080] With reference to Appendix B "Influence Range of Key Soil Pollution Industry Enterprises" of DB41 / T1948-2020 "Technical Specifications for Survey of Soil Pollution Status in Agricultural Land", the theoretical influence distance L of heavy industry enterprises within the jurisdiction is calculated based on different production years, heavy metal emissions, industry type, enterprise scale, different types of pollution, and average precipitation over many years. The calculation formula is:

[0081] Theoretical impact distance d 理论 (km) = basic range (industry benchmark value) + Σ (contribution value of each adjustment factor).

[0082] The calculated d is compared with the impact distance L, and the result is converted into binary data IF(0,1). If d is greater than L, IF is assigned a value of 0, and if it is less than L, IF is assigned a value of 1. The obtained result is combined with the degree of heavy metal emission to obtain the impact factor P of the surrounding heavy industry enterprises on the heavy metals in the soil of a certain agricultural land. 工业源 , the formula is as follows:

[0083]

[0084] Where a is the coefficient of the impact of heavy industry enterprises on the heavy metal content in the soil, n is the total number of heavy industry enterprises around a certain point, and i represents the i-th heavy industry enterprise.

[0085] 2. Processing and quantification of agricultural source data

[0086] Agricultural sources mainly include the use of fertilizers and pesticides, livestock and poultry manure, and sewage discharge. For example, if a city's agricultural land is mainly used for planting, the fertilizer and pesticide use FP (county) in the jurisdiction is obtained from the city's statistical yearbook. Since township-level data is difficult to obtain, the agricultural land area within the township is counted based on the township administrative division vector data. 乡镇 If , the quantification of agricultural sources can be expressed as follows:

[0087]

[0088] 3. Data processing and quantification of traffic sources on main roads

[0089] Investigate the main roads (width ≥ 30m) through soil monitoring points and the main road vector data of a city, and obtain the distance D between a monitoring point and the surrounding traffic main roads within 0.5km from ArcMap. 主干道 (km), width of main road (m). The quantitative formula of traffic pollution source is as follows:

[0090]

[0091] 4. Data processing and quantification of surface water, groundwater, precipitation and dust sources

[0092] Through routine monitoring data from environmental monitoring departments, surface water data, groundwater data, precipitation data, and dustfall data are obtained. If there are many indicators, the indicators are concentrated and reduced in dimension to reduce the collinearity problem. Factor analysis or principal component analysis can be used for processing. Dustfall data can be directly standardized and the processed data can be included in the model as other sources. For example, precipitation indicators can be concentrated into biomass combustion indicators, crustal soil dust indicators, and other industrial and transportation source indicators after dimensionality reduction. Surface water indicators can be concentrated into inorganic element indicators, eutrophication indicators, etc. after dimensionality reduction. The formula is as follows:

[0093] P 自然源 =d×P 地表水 +e×P 地下水 +f×P 降水 +g×P 降尘

[0094] 3. Construction of spatial econometric model

[0095] Heavy metal content C in agricultural land soil from various sources 重金属 The explanatory variables are added to the spatial econometric model. The inverse distance spatial weight matrix is ​​suitable for the analysis of the sources of heavy metals in small-scale, regional point soils. The spatial dependence and heterogeneity of the data are tested and the spatial econometric model is constructed as follows:

[0096]

[0097] Note: In order to eliminate the influence of dimension, the industrial source factors, agricultural source factors, and transportation source factors are first standardized, and then the surface water, groundwater, and dustfall data that have been processed for dimensionality reduction are added to construct the spatial measurement model.

[0098] Application Implementation Case 1

[0099] Currently, the invented model is being used in a special project to trace the source of heavy metal pollution in soil in a certain area. It has greatly improved efficiency and reduced manpower, material and time costs. It has determined the main source and range of a certain metal element in a very short time. The model is shown below:

[0100]

[0101] from Figure 2The Moran Index plot shows that the third quadrant of soil lead concentrations has a relatively large number of samples, indicating positive spatial autocorrelation, and the samples all fall into the low-value range. The resulting model indicates that atmospheric deposition from agricultural, industrial, and transportation sources has a significant impact on soil lead levels. By coordinating intensified monitoring efforts, we were able to quickly identify and remediate pollution sources, achieving significant results in soil pollution regulation.

[0102] Application Implementation Case 2

[0103] 1. Use of economic indicators for pollution sources and agricultural sources for listed companies involved in heavy industries

[0104] If it is difficult to obtain the annual output of companies involved in heavy industries, the annual output value can be used instead. The annual output value data is suitable for studying the impact of listed companies on heavy metals in the surrounding agricultural land soil. In addition, if it is difficult to obtain fertilizer and pesticide data from agricultural sources, the agricultural economic data in the statistical yearbook is relatively complete, subdivided into agriculture, forestry, animal husbandry and fishery. The corresponding economic data can be used for modeling based on the land use type of the county. If the agricultural land in the area is mainly used for planting, the total grain output value (TOVG) in the statistical yearbook can be used. The following is an example:

[0105]

[0106] 2. Impact of pollution sources and relative values ​​of mean values ​​on soil heavy metals

[0107] The impact of some explanatory variables on soil heavy metal content requires data changes. For example, the absolute value of the statistic may have little direct impact, but the relative value has a greater impact on soil heavy metals. For example, the annual output of a company is converted into the ratio of annual output to the industry average output, and the use of fertilizers and pesticides is converted into the ratio to the five-year average. This is suitable for examining the impact of explanatory variables on soil heavy metals relative to a certain threshold. In this case, the model is as follows:

[0108]

[0109] 3. Impact of the change rate of each source on soil heavy metal content

[0110] If we want to systematically examine the impact of the changing values ​​of each source on the soil heavy metal content, we can take the logarithm of the data for modeling. The model is constructed as follows:

[0111]

[0112] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for predicting the source of heavy metal pollution in soil, characterized in that: The method comprises the following steps: S1. Collect 0-20 cm surface soil and obtain the heavy metal content of the explained variable soil; S2. Processing and quantifying the data of the explanatory variables of each pollution source; the pollution sources include heavy industrial pollution sources, agricultural pollution sources, main road traffic pollution sources, and natural pollution sources, wherein the natural pollution sources come from surface water, groundwater, precipitation, and dustfall; S3. The quantified data of each pollution source is used as the heavy metal content of agricultural soil C. 重金属 The explanatory variables were added to the spatial econometric model. The inverse distance spatial weight matrix was used to analyze the sources of heavy metals in small-scale, regional point soils. The spatial dependence and heterogeneity of the data were tested, and the spatial econometric model was constructed as follows: Among them, C 重金属 —Concentration of heavy metals in soil at a certain monitoring point, mg / kg; a—The influence coefficient of pollution sources from heavy industry enterprises on the heavy metal content in soil at a certain location; Pr i , 年产量 —Annual output of the i-th heavy industry enterprise, tons; M i , 产污系数 —Pollution coefficient, i.e. the amount of heavy metal pollutants produced per ton of product by the i-th heavy metal industry enterprise, kg / ton; η i,平均去除效率 —Average removal efficiency of the end-of-pipe treatment technology of the i-th heavy industry enterprise, %; E i (0,1)—binary data, indicating whether the i-th heavy industry enterprise emits heavy metal pollutant i through atmospheric deposition, 0 represents no, 1 represents yes; IF i (0,1)—binary data, indicating whether the heavy metal pollutants discharged by the i-th heavy industry enterprise affect the accumulation of heavy metals in the soil at a certain point; b—the influence coefficient of agricultural sources on the heavy metal content in soil at a certain location; FP—fertilizer and pesticide usage in the jurisdiction FP (county); Area—agricultural land area, the subscript refers to the administrative unit corresponding to the area, square kilometers; c—the influence coefficient of agricultural sources on the heavy metal content in the soil at a certain monitoring point; D 主干道 —The distance between a monitoring point and the surrounding main traffic artery within 0.5 km and with a width of ≥30 m, km; width—width of the main road, m; P 地表水 —Surface water monitoring data around the monitoring point, where d is the surface water impact coefficient; P 地下水 —Groundwater monitoring data around the monitoring point, where e is the groundwater impact coefficient; P 降水 —Precipitation monitoring data of the area where the monitoring point belongs, g is the precipitation impact coefficient; P 降尘 —Precipitation monitoring data of the area where the monitoring point belongs, h is the dustfall impact coefficient.

2. The prediction method according to claim 1, wherein: The step S1 includes the following specific steps: The surface soil was dried, ground, and sieved. The soil passing the 2 mm sieve was used for the determination of moisture and pH value, the soil passing the 0.25 mm sieve was used for the determination of total organic carbon, and the soil passing the 0.15 mm sieve was used for the determination of heavy metal elements. The measured soil heavy metal C heavy metals, total organic carbon, and pH values ​​were standardized to eliminate the dimension effect; The qualitative indicators are assigned to quantitative indicators, wherein the qualitative indicators include the distribution area type SZ, the vegetation type VT, and the administrative division country.

3. The prediction method according to claim 1, wherein: In step S2, the data processing and quantification of heavy industrial pollution sources includes the following steps: (1) Relevant data on heavy industrial enterprises were collected from the National Pollutant Discharge Permit Management Information Platform, mainly including administrative divisions, operating time, industry type, longitude and latitude, and types of pollutant emissions. For atmospheric deposition pollution sources, according to the types of pollutant emissions obtained from the enterprises, if a certain heavy metal pollutant is present, it is assigned a value of 1, and if there is no heavy metal pollutant, it is assigned a value of 0; if there is smoke or particulate matter emission, it is assigned a value of 1, and if there is no smoke or particulate matter emission, it is assigned a value of 0, according to the following formula: Whether a certain heavy metal is emitted (0,1) = particulate matter or smoke emission (0,1) * heavy metal element pollutants (0,1) (2) According to the production volume of enterprise products in the second national pollution source census data, find the pollution coefficient and the average removal efficiency of the end-of-pipe treatment technology, and weight the heavy metal emissions to express the heavy metal emissions. 重金属 , the formula is as follows: Emission 重金属 =Pr 年生产量 ×M 产污系数 ×(1-n) 平均去除效率 )×E 是否排放 (0.1) (3) According to the obtained latitude and longitude of soil heavy metal monitoring point information and heavy metal enterprise point information, convert it into radians. The conversion formula is: The Haversine formula is used to calculate the distance d between two points, as shown below: Where: φ1 and φ2 are the latitudes of the two points (in radians), φ2-φ1 is the difference in latitude, λ2-λ1 is the difference in longitude, and r is the radius of the Earth. (4) Calculate the theoretical impact distance L of heavy industrial enterprises within the jurisdiction using the following formula: Theoretical impact distance d 理论 (km) = basic range (industry benchmark value) + Σ (contribution value of each adjustment factor); (5) Compare the calculated d with the impact distance L, and convert the result into binary data IF(0,1). If d is greater than L, IF is assigned a value of 0; if it is less than L, IF is assigned a value of 1. The obtained result is combined with the degree of heavy metal emission to obtain the impact factor P of the surrounding heavy industry enterprises on the heavy metals in the soil of a certain agricultural land. 工业源 , the formula is as follows: Where a is the coefficient of the impact of heavy industry enterprises on the heavy metal content in the soil, n is the total number of heavy industry enterprises around a certain point, and i represents the i-th heavy industry enterprise.

4. The prediction method according to claim 1, wherein: In step S2, the quantitative formula for the agricultural pollution source is: Where b is the coefficient of the impact of agricultural sources on soil heavy metal content, FP is the amount of fertilizer and pesticide use in the jurisdiction (FP (county), which is derived from the city's statistical yearbook; Area is the area of ​​agricultural land, and the subscripts refer to the administrative unit corresponding to the area, square kilometers.

5. The prediction method according to claim 1, wherein: In step S2, the quantitative formula for the main road traffic pollution source is: Among them, D 主干道 (km) is the distance between a monitoring point and the surrounding traffic arterial road within 0.5 km obtained from ArcMap, width (m) is the width of the arterial road, and c is the influence coefficient of the traffic source.

6. The prediction method according to claim 1, wherein: In step S2, the quantification formula of the natural pollution source is: P 自然源 =dP 地表水 +eP 地下水 +fP 降水 +gP 降尘 Among them, surface water data, groundwater data, precipitation, and dustfall data are derived from routine monitoring data of environmental monitoring departments. If there are too many indicators, the indicators can be concentrated, dimensionally reduced, and collinearity reduced through factor analysis or principal component analysis. Dustfall data can be directly standardized, and then the processed data can be included in the model as other sources. After dimension reduction, surface water indicators can be concentrated into inorganic element indicators and eutrophication indicators.

7. The prediction method according to claim 1, wherein: In step S3, the pollution source factors related to heavy industry, agricultural pollution source factors, and traffic pollution source factors are standardized, and then the natural pollution source data that have been processed for dimensionality reduction are added to construct a spatial measurement model.

Images