Farmland residual film pollution comprehensive risk assessment and early warning method coupled with space-air-ground data

Abstract

The invention relates to the field of farmland residual film pollution risk assessment and early warning, and discloses a farmland residual film pollution comprehensive risk assessment and early warning method coupled with space-air-ground data. According to the method, spatial distribution of mulching films and residual films is extracted, and multi-dimensional data such as cultivated land quality grades, basic farmland ranges, administrative boundaries and agricultural statistics are integrated; after unifying the coordinate system and delimiting an evaluation unit, calculating the residual film rate, the retention rate and the land average production capacity; a farmland residual film pollution comprehensive risk assessment model including a residual film pollution index, a cultivated land quality grade, basic farmland attributes and average land production capacity is constructed, and a comprehensive risk index is calculated; and dividing risk levels based on a natural breakpoint method, analyzing and identifying spatial aggregation features in combination with cold and hot spots, and generating an early warning level spatial distribution diagram. According to the invention, through fusion of space-air-ground data, whole-process supervision from identification to evaluation to early warning is realized, evaluation scientificity, accuracy and efficiency are improved, and effective technical support is provided for agricultural non-point source pollution treatment.

Description

Technical Field

[0001] This invention relates to the field of risk assessment and early warning of agricultural residual film pollution, specifically to a comprehensive risk assessment and early warning method for agricultural residual film pollution that couples space-air-ground data, belonging to the field of agricultural environmental protection and remote sensing monitoring technology. Background Technology

[0002] Agricultural mulching technology is widely used globally due to its significant effects on moisture retention, warming, and weed control, playing a vital role in ensuring high and stable agricultural yields. However, if the mulch film is not promptly and thoroughly recycled after crop harvest, large amounts remain in the soil, creating "white pollution." Residual agricultural film damages soil structure, hinders the transport of water and nutrients, affects crop root growth, leading to reduced crop yields, and may pose a potential threat to the ecological environment and human health through the food chain.

[0003] Currently, the monitoring and assessment of agricultural plastic film pollution mainly relies on manual field surveys and sampling. This method is not only time-consuming and labor-intensive with limited coverage, but also highly subjective, making it difficult to achieve large-scale, rapid, and objective dynamic monitoring. With the development of remote sensing technology, some studies have attempted to use UAV remote sensing to identify plastic film residue. However, existing technologies only focus on the distribution of the plastic film itself and fail to conduct comprehensive risk assessments from the perspective of the entire agricultural ecosystem. The risk of plastic film pollution depends not only on the quantity of residue but also on various spatial multidimensional factors such as local climate, soil quality, farmland protection policies, and agricultural production levels. The lack of a systematic method that can integrate multi-source remote sensing information and spatial multidimensional data to scientifically, accurately, and efficiently assess the comprehensive risk of regional agricultural plastic film pollution and conduct spatial early warning has, to some extent, constrained the precise control of agricultural non-point source pollution and the formulation of decisions on the sustainable use of farmland resources.

[0004] Therefore, there is an urgent need for a method that integrates advanced Earth observation technology with spatial analysis models to overcome the shortcomings of existing technologies and achieve comprehensive assessment and graded early warning of the risk of residual plastic film pollution in large-scale farmland. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this application provides a comprehensive risk assessment and early warning method for farmland residual film pollution that couples space, air, and ground data. By comprehensively utilizing high-resolution satellite and UAV remote sensing imagery and multi-dimensional spatial data, it introduces multi-dimensional evaluation factors such as residual film pollution index, arable land quality grade, permanent basic farmland attributes, and average arable land productivity, constructing a comprehensive risk assessment model for farmland residual film pollution that integrates environmental, resource, and economic values. Relying on natural breakpoint classification and hot / cold spot analysis techniques, it achieves a combination of risk level classification and cluster type identification, generating a visualized spatial distribution map of farmland residual film pollution early warning levels. This invention realizes full-process monitoring from identification to assessment to early warning, providing direct and scientific basis for precise monitoring and decision-making regarding regional farmland residual film pollution risks. Furthermore, this method has a high degree of automation, requires minimal human intervention, and is easy to promote and apply on a large scale.

[0006] The integrated risk assessment and early warning method for farmland residual film pollution that couples space-air-ground data provided in this application adopts the following technical solution:

[0007] A comprehensive risk assessment and early warning method for farmland residual film pollution that couples space-air-ground data includes the following steps:

[0008] S1. Acquire high-resolution first remote sensing images of the key growth period of crop mulching in farmland of the study area, and extract the spatial distribution map of mulch film from the first remote sensing images.

[0009] S2. After the mulched crops were harvested and the land was tilled, a high-resolution second remote sensing image of the study area was acquired, and the spatial distribution map of the fields and the spatial distribution map of the residual film were extracted from the second remote sensing image.

[0010] S3, obtain the farmland quality grade map, permanent basic farmland delineation map, administrative unit boundary map, county farmland area, and county agricultural added value of the study area;

[0011] S4. Transform the spatial distribution map of plastic film in S1, the spatial distribution map of field plots and the spatial distribution map of residual plastic film in S2, and the cultivated land quality grade map, permanent basic farmland delineation map, and administrative management unit boundary map in S3 into the same coordinate system to obtain the results of the spatial distribution map of plastic film, the spatial distribution map of field plots, the spatial distribution map of residual plastic film, the results of the cultivated land quality grade, the results of the delineation map of permanent basic farmland, and the results of the boundary map of administrative management units.

[0012] S5, Delineate the assessment unit, which is a natural field, an administrative unit, or a regular grid;

[0013] S6, based on the evaluation unit in S5, calculate the residual film rate RLR and retention rate RRR;

[0014] ;

[0015] ;

[0016] S7. Construct a comprehensive risk assessment model for farmland residual film pollution that includes multiple evaluation factors, including residual film pollution index (RPI), arable land quality grade, basic farmland attributes, and arable land productivity per unit area.

[0017] S8. For each evaluation factor in S7, range standardization is performed to obtain standardized data for residual film pollution index (RPI), farmland quality grade, basic farmland attribute, and farmland productivity per unit area. The formula for range standardization is as follows:

[0018] If the evaluation indicator is positive, then:

[0019] ;

[0020] If the evaluation metric is reversed, then:

[0021] ;

[0022] In the formula, , The first The maximum and minimum values ​​of each evaluation indicator across all evaluation units. For the first The evaluation index is in the first The original values ​​of each evaluation unit, For the first The evaluation index is in the first Standardized data for each evaluation unit;

[0023] S9 assigns weights and determines the weights of the standardized data of residual film pollution index RPI, standardized data of cultivated land quality grade, standardized data of basic farmland attributes, and standardized data of cultivated land productivity per unit area in S8.

[0024] S10, based on the standardized data of residual film pollution index RPI, standardized data of cultivated land quality grade, standardized data of basic farmland attributes, and standardized data of cultivated land productivity per unit area in S8, as well as the weights corresponding to each standardized data determined in S9, the comprehensive index method is used to calculate the comprehensive risk index PFPI of plastic film pollution, as follows:

[0025] ;

[0026] In the formula, Indicates the first The comprehensive risk index of plastic film pollution for each evaluation unit; For the first The evaluation index is in the first Standardized data for each evaluation unit; For the first The weight of each evaluation indicator; The total number of evaluation indicators;

[0027] S11. The risk level of the PFPI obtained in S10 is divided by the natural segment method to generate a spatial distribution map of the comprehensive risk level of plastic film pollution.

[0028] S12, Spatial clustering analysis is performed on the PFPI obtained in S10. Based on the significance results, cold and hot spot spatial clustering types are divided, and a spatial distribution map of the spatial clustering types of comprehensive risk of plastic film pollution is generated.

[0029] S13. Based on the comprehensive analysis of the spatial distribution map of the comprehensive risk level of plastic film pollution in S11 and the spatial distribution map of the spatial clustering type of comprehensive risk of plastic film pollution in S12, the warning levels are classified, and a spatial distribution map of the warning level of residual plastic film pollution in farmland is generated.

[0030] Preferably, in step S7, the residual film fouling index RPI is calculated by a weighted summation method, a multiplicative synthesis method, or a polar coordinate method;

[0031] The weighted summation method is calculated as follows: ;

[0032] in and These are the weighting coefficients. ;

[0033] The calculation process of the multiplication composition method is as follows: ;

[0034] The polar coordinate method calculation process is as follows: .

[0035] Preferably, the spatial model of per capita productivity of cultivated land in step S7 is as follows:

[0036] ;

[0037] In the formula, Indicates the first The per capita productivity of arable land in each evaluation unit Representing the evaluation unit subordinate counties Agricultural added value, and Representing the evaluation units respectively He County The cultivated land area includes grain crops, oil crops, vegetables, and fruits.

[0038] Preferably, the spatial clustering analysis method in step S12 is the Getis-OrdGi* spatial hotspot analysis method, with the following formula:

[0039] ;

[0040] , ;

[0041] In the formula, Z is the score. The higher the Z score, the more clustered the high-value pixels are. The lower the Z score, the more clustered the low-value pixels are. When the Z score is close to 0, it means that there is no obvious spatial clustering of pixels. For the first GDI score for each evaluation unit; As an evaluation unit and The binary spatial weights between them are defined by distance rules, with 1 for adjacent spatial ranges and 0 for non-adjacent spatial ranges; The total number of evaluation units; This is the average value. Let Variance be the variance.

[0042] The beneficial effects of this invention are as follows:

[0043] 1. This invention breaks through the limitations of traditional methods that only focus on the amount of residual plastic film. It innovatively introduces multi-dimensional evaluation factors such as residual film pollution index, arable land quality grade, permanent basic farmland attributes, and average productivity of arable land, and constructs a comprehensive risk index for farmland residual film pollution. This expands the pollution problem into a comprehensive risk assessment that integrates environmental, resource, and economic value, providing objective and quantitative decision support for agricultural environmental management and significantly improving the pertinence and scientific nature of governance measures.

[0044] 2. This invention constructs a dual-indicator evaluation system that integrates "residual film rate" and "retention rate" to characterize the absolute degree of residual film pollution and the relative effectiveness of film recycling, respectively, and jointly quantifies the level of residual film and the treatment effect. This achieves an important leap from qualitative judgment to quantitative assessment and provides a reliable basis for accurately characterizing the degree of pollution.

[0045] 3. This invention generates a spatially clear spatial distribution map of residual film pollution risk warning through risk level classification and spatial clustering analysis. It can support multi-scale assessment units such as natural fields, administrative management units, and regular grids, meeting the early warning and decision-making needs of management entities at different levels.

[0046] 4. Based on the spatial distribution map of residual film pollution risk early warning, this invention realizes the priority division of governance by administrative management unit or specific natural field plot, clearly identifies key governance areas, promotes the efficient transformation of monitoring results into governance actions, provides direct basis for regional residual film pollution prevention and control, recycling policy formulation and ecological governance, and helps sustainable agricultural development. Attached Figure Description

[0047] Figure 1 This is a flowchart illustrating an embodiment of this application;

[0048] Figure 2 This is a schematic diagram of a high-resolution first remote sensing image (using satellite remote sensing image GF7) according to an embodiment of this application;

[0049] Figure 3 This is a schematic diagram of the boundary map of a county-level administrative unit according to an embodiment of this application;

[0050] Figure 4 This is a schematic diagram of land use data from an embodiment of this application;

[0051] Figure 5 This is a schematic diagram illustrating the added value of county-level agriculture in an embodiment of this application. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this invention clearer, the 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 merely illustrative and not intended to limit the invention.

[0053] This embodiment uses the risk and early warning of residual plastic film pollution in farmland in a certain area as an example. The data used includes Gaofen-7 remote sensing imagery (high-resolution satellite remote sensing imagery GF-7), images acquired by a DJI M300 RTK drone equipped with a Zenmuse H30T camera, and spatial multidimensional data such as farmland quality grade, permanent basic farmland area, county-level administrative unit boundaries, land use data, and county-level agricultural added value. The GF-7 satellite is equipped with a dual-line array stereo camera, which can acquire multispectral images (back-view) with a resolution of 3.2m and panchromatic stereo images (forward-view and back-view) with a resolution of 0.8m / 0.65m. After image fusion, multispectral data with a resolution of no less than 1m can be generated, providing a reliable basis for plastic film identification. At a flight altitude of 50 meters, the Zenmuse H30T's visible light camera has a spatial resolution of approximately 0.45cm, which can provide high-precision data support for monitoring residual plastic film in farmland. Land use data is sourced from the CLCD dataset, a 30m resolution land use dataset published by Wuhan University from 1985 to 2024. This dataset uses a 9-level classification system, effectively supporting the extraction and analysis of county-level cultivated land area. County-level agricultural added value is obtained from the *Hebei Rural Statistical Yearbook*. The combination of these multi-source remote sensing data and spatial multidimensional data comprehensively meets the data requirements for risk assessment and early warning of agricultural film residue pollution.

[0054] Reference Figure 1 This invention discloses a method for comprehensive risk assessment and early warning of farmland residual film pollution that couples space-air-ground data.

[0055] Among them, air-space-ground data includes multi-source remote sensing data and spatial multidimensional data;

[0056] Multi-source remote sensing data includes remote sensing data such as first-source remote sensing data and second-source remote sensing data;

[0057] Spatial multidimensional data includes data from dimensions such as cultivated land quality grade maps, permanent basic farmland delineation maps, administrative unit boundary maps, county cultivated land area, and county agricultural added value.

[0058] Specifically, the integrated risk assessment and early warning method for farmland residual film pollution that couples space-air-ground data includes the following steps:

[0059] S1. Acquire high-resolution first remote sensing images of the key growth period of crop mulching in farmland of the study area, and extract the spatial distribution map of mulch film from the first remote sensing images.

[0060] The first remote sensing image can be acquired by satellite or drone.

[0061] The aforementioned first remote sensing image may be an optical image, a SAR image, or other modal images or multimodal fusion images, etc. The first remote sensing image is not limited to one or more images that can achieve the function of extracting the spatial distribution map of the mulch film from the first remote sensing image.

[0062] Specifically, refer to Figure 2 In this embodiment, the high-resolution first remote sensing image is a satellite multispectral optical remote sensing image, specifically a GF-7 remote sensing image taken on December 18, 2024, during the period of exposed plastic film covering crops, and on February 25, 2025, during the period of vigorous growth of plastic film covering crops. The spatial resolution of the multispectral image (back view) is 3.2m, and the panchromatic resolution is 0.8m / 0.65m (forward view and back view). After the multispectral and panchromatic images are fused, a multispectral remote sensing image with a resolution of not less than 1m can be generated, which can meet the high-precision extraction of plastic film coverings.

[0063] S2. After the mulched crops were harvested and the land was tilled, a high-resolution second remote sensing image of the study area was acquired, and the spatial distribution map of the fields and the spatial distribution map of the residual film were extracted from the second remote sensing image.

[0064] The second remote sensing image can also be acquired via satellite or drone;

[0065] The aforementioned second remote sensing image may be an optical image, a thermal infrared image, or other modal images or multimodal fusion images, etc. The second remote sensing image is not limited to one or more types of images that can achieve the above-mentioned function of extracting the spatial distribution map of the field plots and the spatial distribution map of the residual film from the second remote sensing image.

[0066] Specifically, in this embodiment, the second remote sensing image was acquired by a DJI M300 RTK drone equipped with a Zenmuse H30T at a flight altitude of 50 meters. The spatial resolution of the visible light camera is approximately 0.45 cm, which can provide high-precision remote sensing data support for residual film monitoring.

[0067] In addition, field spatial distribution maps can also be obtained through remote sensing image acquisition or through other means from relevant management agencies.

[0068] S3, obtain the farmland quality grade map, permanent basic farmland delineation map, administrative unit boundary map, county farmland area, and county agricultural added value of the study area;

[0069] The administrative unit is the administrative village, administrative township, or administrative district / county;

[0070] Specifically, refer to Figures 3 to 5 In this embodiment, the cultivated land quality grade map, the permanent basic farmland delineation map, the administrative management unit boundary map, the county agricultural added value and county cultivated land area statistics all cover the entire area of ​​Handan City, Hebei Province, and the administrative management unit is the administrative district / county.

[0071] S4. The spatial distribution maps of plastic film in S1, field plots and residual plastic film in S2, and the cultivated land quality grade map, permanent basic farmland delineation map, and administrative unit boundary map in S3 are uniformly transformed into the same coordinate system to obtain the results maps of plastic film spatial distribution, field plot spatial distribution, residual plastic film spatial distribution, cultivated land quality grade, permanent basic farmland delineation, and administrative unit boundary.

[0072] S5, Delineate the assessment unit, which can be a natural field, an administrative unit, or a regular grid;

[0073] The study area can be divided into regular grids of n meters × m meters.

[0074] Specifically, regular fishnets are mainly created using the Create Fishnet tool in ArcGIS software, and the size of the fishnet can be set according to actual needs.

[0075] Specifically, in this embodiment, the evaluation unit is the county-level administrative unit of Handan City.

[0076] S6, based on the evaluation unit in step S5, calculate the residual film rate RLR and retention rate RRR;

[0077] ;

[0078] ;

[0079] S7. Construct a comprehensive risk assessment model for farmland residual film pollution that includes multiple evaluation factors, such as residual film pollution index (RPI), arable land quality grade, basic farmland attributes, and arable land productivity per unit area.

[0080] Specifically, the residual film fouling index (RPI) can be calculated using any of the following methods: a, b, or c.

[0081] a. Weighted summation method: ;

[0082] in and These are the weighting coefficients. ;

[0083] b. Multiplicative composition method: ;

[0084] c. Polar coordinate method: ;

[0085] In this embodiment, the residual film fouling index (RPI) is calculated using a weighted summation method, and the formula is as follows:

[0086] ,in and These are the weighting coefficients. ,and = =0.5.

[0087] Specifically, the spatial model of per capita productivity of arable land is as follows:

[0088] ;

[0089] In the formula, Indicates the first The per capita productivity of arable land in each evaluation unit Representing the evaluation unit subordinate counties Agricultural added value, and Representing the evaluation units respectively He County The cultivated land area includes grain crops, oil crops, vegetables, and fruits.

[0090] S8. For each evaluation factor in S7, range standardization is performed to obtain standardized data for residual film pollution index (RPI), farmland quality grade, basic farmland attribute, and farmland productivity per unit area. The formula for range standardization is as follows:

[0091] If the evaluation indicator is positive, then the formula is:

[0092] ;

[0093] If the evaluation metric is reversed, then the formula is:

[0094] ;

[0095] In the formula, , The first The maximum and minimum values ​​of each evaluation indicator across all evaluation units. For the first The evaluation index is in the first The original values ​​of each evaluation unit, For the first The evaluation index is in the first Standardized data for each evaluation unit;

[0096] S9 assigns weights and determines the weights of the standardized data of residual film pollution index RPI, standardized data of cultivated land quality grade, standardized data of basic farmland attributes, and standardized data of cultivated land productivity per unit area in S8.

[0097] Specifically, in this embodiment, the weighting method is the analytic hierarchy process (AHP).

[0098] In other feasible embodiments, the weighting method may also use other methods capable of achieving the above-described functions.

[0099] S10, based on the standardized data of residual film pollution index RPI, standardized data of cultivated land quality grade, standardized data of basic farmland attributes, and standardized data of cultivated land productivity per unit area in S8, as well as the weights corresponding to each standardized data determined in S9, the comprehensive index method is used to calculate the comprehensive risk index PFPI of plastic film pollution, as follows:

[0100] ;

[0101] In the formula, Indicates the first The comprehensive risk index of plastic film pollution for each evaluation unit; For the first The evaluation index is in the first Standardized data for each evaluation unit; For the first The weight of each evaluation indicator; The total number of evaluation indicators;

[0102] S11. The risk level of the PFPI obtained in S10 is divided by the natural segment method to generate a spatial distribution map of the comprehensive risk level of plastic film pollution.

[0103] Specifically, in this embodiment, the natural breakpoint method is used to divide PFPI into three levels: high risk, medium risk, and low risk.

[0104] S12, Spatial clustering analysis is performed on the PFPI obtained in S10. Based on the significance results, cold and hot spot spatial clustering types are divided, and a spatial distribution map of the spatial clustering types of comprehensive risk of plastic film pollution is generated.

[0105] Specifically, spatial clustering analysis uses the Getis-Ord Gi* method to identify cold and hot spots of plastic film pollution risk. Significantly positive indicates that the evaluation unit The surrounding comprehensive ecological risk difference is also relatively high, belonging to areas with significantly increased clustering (hot spots), and conversely, areas with significantly decreased clustering (cold spots). The formula is as follows:

[0106] ;

[0107] , ;

[0108] In the formula, Z is the score. The higher the Z score, the more clustered the high-value pixels are. The lower the Z score, the more clustered the low-value pixels are. When the Z score is close to 0, it means that there is no obvious spatial clustering of pixels. For the first GDI score for each evaluation unit; As an evaluation unit and The binary spatial weights between them are defined by distance rules, with 1 for adjacent spatial ranges and 0 for non-adjacent spatial ranges; The total number of evaluation units; This is the average value. Let Variance be the variance.

[0109] Specifically, in this embodiment, the Getis-Ord Gi* hotspot analysis method is mainly calculated using the HotSpot Analysis (Getis-Ord Gi*) tool in ArcGIS software. The Gi_Bin field in the calculation result can identify statistically significant hotspots and coldspots. Elements in the + / -3 interval reflect a 99% confidence level; elements in the + / -2 interval reflect a 95% confidence level; elements in the + / -1 interval reflect a 90% confidence level; and the clustering of elements in the 0 interval is not statistically significant. Therefore, in this embodiment, Gi_Bin greater than or equal to 1 indicates a significant hotspot; Gi_Bin less than or equal to 1 indicates a significant coldspot; and Gi_Bin equal to 0 indicates no significance.

[0110] S13. Based on the comprehensive analysis of the spatial distribution map of the comprehensive risk level of plastic film pollution in S11 and the spatial distribution map of the spatial clustering type of comprehensive risk of plastic film pollution in S12, the warning levels are classified, and a spatial distribution map of the warning level of residual plastic film pollution in farmland is generated.

[0111] Specifically, in this embodiment, the method for classifying warning levels is as follows:

[0112]

[0113] Although the invention has been described herein with reference to several illustrative embodiments, it should be understood that many other modifications and implementations can be devised by those skilled in the art, which will fall within the scope and spirit of the principles disclosed herein. More specifically, various variations and modifications can be made to the components and / or layout of the subject matter arrangement within the scope of the disclosure, drawings, and claims. Besides variations and modifications to the components and / or layout, other uses will be apparent to those skilled in the art.

Claims

1. A method for comprehensive risk assessment and early warning of farmland residual film pollution that couples space-air-ground data, characterized in that, Includes the following steps: S1. Acquire high-resolution first remote sensing images of the key growth period of crop mulching in farmland of the study area, and extract the spatial distribution map of mulch film from the first remote sensing images. S2. After the mulched crops were harvested and the land was tilled, a high-resolution second remote sensing image of the study area was acquired, and the spatial distribution map of the fields and the spatial distribution map of the residual film were extracted from the second remote sensing image. S3, obtain the farmland quality grade map, permanent basic farmland delineation map, administrative unit boundary map, county farmland area, and county agricultural added value of the study area; S4. Transform the spatial distribution map of plastic film in S1, the spatial distribution map of field plots and the spatial distribution map of residual plastic film in S2, and the cultivated land quality grade map, permanent basic farmland delineation map, and administrative management unit boundary map in S3 into the same coordinate system to obtain the results of the spatial distribution map of plastic film, the spatial distribution map of field plots, the spatial distribution map of residual plastic film, the results of the cultivated land quality grade, the results of the delineation map of permanent basic farmland, and the results of the boundary map of administrative management units. S5, Delineate the assessment unit, which is a natural field, an administrative unit, or a regular grid; S6, based on the evaluation unit in S5, calculate the residual film rate RLR and retention rate RRR; ； ； S7. Construct a comprehensive risk assessment model for farmland residual film pollution that includes multiple evaluation factors, including residual film pollution index (RPI), arable land quality grade, basic farmland attributes, and arable land productivity per unit area. S8. For each evaluation factor in S7, range standardization is performed to obtain standardized data for residual film pollution index (RPI), farmland quality grade, basic farmland attribute, and farmland productivity per unit area. The formula for range standardization is as follows: If the evaluation indicator is positive, then: ； If the evaluation metric is reversed, then: ； In the formula, , The first The maximum and minimum values ​​of each evaluation indicator across all evaluation units. For the first The evaluation index is in the first The original values ​​of each evaluation unit, For the first The evaluation index is in the first Standardized data for each evaluation unit; S9 assigns weights and determines the weights of the standardized data of residual film pollution index RPI, standardized data of cultivated land quality grade, standardized data of basic farmland attributes, and standardized data of cultivated land productivity per unit area in S8. S10, based on the standardized data of residual film pollution index RPI, standardized data of cultivated land quality grade, standardized data of basic farmland attributes, and standardized data of cultivated land productivity per unit area in S8, as well as the weights corresponding to each standardized data determined in S9, the comprehensive index method is used to calculate the comprehensive risk index PFPI of plastic film pollution, as follows: ； In the formula, Indicates the first The comprehensive risk index of plastic film pollution for each evaluation unit; For the first The evaluation index is in the first Standardized data for each evaluation unit; For the first The weight of each evaluation indicator; The total number of evaluation indicators; S11. The risk level of the PFPI obtained in S10 is divided by the natural segment method to generate a spatial distribution map of the comprehensive risk level of plastic film pollution. S12, Spatial clustering analysis is performed on the PFPI obtained in S10. Based on the significance results, cold and hot spot spatial clustering types are divided, and a spatial distribution map of the spatial clustering types of comprehensive risk of plastic film pollution is generated. S13. Based on the comprehensive analysis of the spatial distribution map of the comprehensive risk level of plastic film pollution in S11 and the spatial distribution map of the spatial clustering type of comprehensive risk of plastic film pollution in S12, the warning levels are classified, and a spatial distribution map of the warning level of residual plastic film pollution in farmland is generated.

2. The method for comprehensive risk assessment and early warning of farmland residual film pollution coupled with space-air-ground data as described in claim 1, characterized in that, In step S7, the residual film fouling index RPI is calculated by weighted summation, multiplication synthesis, or polar coordinate method. The weighted summation method is calculated as follows: ; in and These are the weighting coefficients. ; The calculation process of the multiplication composition method is as follows: ; The polar coordinate method calculation process is as follows: .

3. The method for comprehensive risk assessment and early warning of farmland residual film pollution coupled with space-air-ground data as described in claim 1, characterized in that, The spatial model of per capita productivity of cultivated land in step S7 is as follows: ； In the formula, Indicates the first The per capita productivity of arable land in each evaluation unit Representing the evaluation unit subordinate counties Agricultural added value, and Representing the evaluation units respectively He County The cultivated land area includes grain crops, oil crops, vegetables, and fruits.

4. The method for comprehensive risk assessment and early warning of farmland residual film pollution coupled with space-air-ground data as described in claim 1, characterized in that, The spatial clustering analysis method in step S12 is the Getis-Ord Gi* spatial hotspot analysis method, and the formula is as follows: ； ， ； In the formula, Z is the score. The higher the Z score, the more clustered the high-value pixels are. The lower the Z score, the more clustered the low-value pixels are. When the Z score is close to 0, it means that there is no obvious spatial clustering of pixels. For the first GDI score for each evaluation unit; As an evaluation unit and The binary spatial weights between them are defined by distance rules, with 1 for adjacent spatial ranges and 0 for non-adjacent spatial ranges; The total number of evaluation units; This is the average value. Let Variance be the variance.

Images