A comprehensive assessment method of atmospheric pollution level based on LUCC

Abstract

The application relates to a LUCC-based comprehensive atmospheric pollution level evaluation method, and relates to the technical field of data processing.The method comprises the following steps: acquiring regional feature sets, land use data sets and atmospheric pollution concentration sequences of a plurality of sample regions, determining target types corresponding to the sample regions, and dividing the plurality of sample regions into a plurality of sample region groups; for each sample region group, based on the target types corresponding to the sample regions included in the sample region group, the land use data sets and the atmospheric pollution concentration sequences, establishing an atmospheric pollution evaluation model, and based on the regional feature sets of the sample regions included in the sample region group, determining target regional feature factors corresponding to the sample region group; determining a target sample region group according to a regional feature set of a region to be evaluated; and determining the atmospheric pollution level of the region to be evaluated according to the land use data set of the region to be evaluated through the atmospheric pollution evaluation model corresponding to the target sample region group, which has the advantages of improving the efficiency and accuracy of comprehensive atmospheric pollution level evaluation.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a comprehensive assessment method for air pollution levels based on LUCC. Background Technology

[0002] With the rapid pace of global urbanization, land use and land cover change (LUCC) has an increasingly significant impact on the environment and ecosystems, particularly on air quality. During urbanization, vast amounts of natural land are converted into building land, significantly altering the land cover structure and profoundly affecting regional and global climate and air quality. This land use change directly influences the formation and dispersion of particulate matter and other pollutants. Under the dual pressures of rapid urbanization and industrialization, factors such as traffic congestion and increased energy consumption have led to a significant increase in atmospheric pollutant emissions, making air quality deterioration one of the most pressing global challenges. PM10, particulate matter with a diameter of 10 micrometers or less, has a significant impact on the environment and human health. PM10 sources are widespread, including industrial emissions, traffic pollution, and natural factors such as wind erosion (e.g., dust from sandstorms). Studies show that elevated PM10 concentrations not only pose a serious threat to health but also exacerbate regional air pollution, especially during sandstorms in spring and autumn. Understanding and effectively controlling PM10 pollution is crucial for maintaining good air quality and protecting public health.

[0003] Current comprehensive assessment methods for the atmospheric environment are generally highly subjective, lack scientific rigor, and are not easily convincing. Furthermore, traditional comprehensive assessment methods are not applicable to all regions, resulting in low assessment accuracy.

[0004] Therefore, there is a need to provide a comprehensive assessment method for air pollution levels based on LUCC to improve the efficiency and accuracy of comprehensive assessment of air pollution levels. Summary of the Invention

[0005] This invention provides a comprehensive assessment method for air pollution levels based on LUCC (Land Use Concentration Data Set), comprising: acquiring regional feature sets, land use datasets, and air pollution concentration sequences for multiple sample areas; wherein the regional feature sets include feature data corresponding to multiple regional feature factors, the land use datasets include land use data for various land types at multiple time points, and the air pollution concentration sequences include air pollution concentrations at multiple time points; for each sample area, determining the target type corresponding to the sample area based on the land use dataset and air pollution concentration sequence; and dividing the multiple sample areas into multiple sample area groups based on the target type corresponding to each sample area. For each sample area group, an air pollution assessment model is established based on the target type, land use dataset, and air pollution concentration sequence corresponding to the sample areas included in the sample area group; for each sample area group, target area feature factors corresponding to the sample area group are determined based on the regional feature set of the sample areas included in the sample area group; the regional feature set and land use dataset of the area to be assessed are obtained; a target sample area group is determined from the multiple sample area groups according to the regional feature set of the area to be assessed; and the air pollution level of the area to be assessed is determined based on the land use dataset of the area to be assessed using the air pollution assessment model corresponding to the target sample area group.

[0006] Furthermore, based on the land use dataset and air pollution concentration sequence of the sample area, the target type corresponding to the sample area is determined, including: for each type, determining a first correlation coefficient between the type and the air pollution concentration of the sample area based on the land use dataset and air pollution concentration sequence of the sample area; for every two types, determining a second correlation coefficient between the two types and the air pollution concentration of the sample area based on the land use dataset and air pollution concentration sequence of the sample area; and determining the target type corresponding to the sample area based on the first correlation coefficient between each type and the air pollution concentration of the sample area and the second correlation coefficient between every two types and the air pollution concentration of the sample area.

[0007] Furthermore, based on the land use dataset and air pollution concentration sequence of the sample area, a second correlation coefficient between the two types and the air pollution concentration of the sample area is determined, including: calculating land use transfer data at multiple time points corresponding to the two types based on the land use dataset of the sample area; and determining the second correlation coefficient between the two types and the air pollution concentration of the sample area based on the land use transfer data at multiple time points corresponding to the two types and the air pollution concentration sequence of the sample area.

[0008] Furthermore, based on the target type corresponding to each of the sample regions, the plurality of sample regions are divided into a plurality of sample region groups, including: for any two sample regions, calculating the type similarity between the two sample regions based on the target types corresponding to the two sample regions; clustering the plurality of sample regions according to the type similarity of any two sample regions using a first clustering algorithm to determine sample region clusters; for each sample region cluster, calculating the correlation coefficient similarity between any two sample regions included in the sample region cluster and the air pollution concentration of the sample region according to a first correlation coefficient between the type corresponding to the sample region included in the sample region cluster and the air pollution concentration of the sample region, and two second correlation coefficients between the two types and the air pollution concentration of the sample region; and clustering the plurality of sample regions included in the sample region cluster according to the correlation coefficient similarity of any two sample regions included in the sample region cluster using a second clustering algorithm to determine a plurality of sample region groups.

[0009] Furthermore, based on the target types, land use datasets, and air pollution concentration sequences corresponding to the sample areas included in the sample area group, an air pollution assessment model is established, including: determining the group target type corresponding to the sample area group based on the target types corresponding to the sample areas included in the sample area group; establishing multiple training samples according to the group target type corresponding to the sample area group and the land use datasets and air pollution concentration sequences of the sample areas included in the sample area group; establishing the air pollution assessment model, and training the air pollution assessment model using the multiple training samples.

[0010] Furthermore, based on the regional feature set of the sample regions included in the sample region group, the target regional feature factor corresponding to the sample region group is determined, including: for each regional feature factor, calculating the difference parameter corresponding to the regional feature factor according to the regional feature set of the sample regions included in the sample region group; and determining the target regional feature factor corresponding to the sample region group according to the difference parameter corresponding to each regional feature factor.

[0011] Furthermore, determining a target sample region group from the plurality of sample region groups based on the regional feature set of the region to be evaluated includes: for each sample region group, determining group feature data corresponding to the sample region group based on the target regional feature factor corresponding to the sample region group and the regional feature set of the sample regions included in the sample region group; for each sample region group, determining feature data corresponding to the region to be evaluated based on the target regional feature factor corresponding to the sample region group and the regional feature set of the region to be evaluated; calculating the feature similarity between the region to be evaluated and the sample region group based on the group feature data corresponding to the sample region group and the feature data corresponding to the region to be evaluated; and determining a target sample region group from the plurality of sample region groups based on the feature similarity between the region to be evaluated and each sample region group.

[0012] Furthermore, the atmospheric pollution level of the area to be assessed is determined by using the atmospheric pollution assessment model corresponding to the target sample area group based on the land use dataset of the area to be assessed, including: extracting the dynamic characteristics of land use in the area to be assessed based on the group target type corresponding to the target sample area group and the land use dataset of the area to be assessed; and determining the atmospheric pollution level of the area to be assessed based on the dynamic characteristics of land use in the area to be assessed by using the atmospheric pollution assessment model corresponding to the target sample area group.

[0013] Furthermore, based on the group target type corresponding to the target sample area group and the land use dataset of the area to be evaluated, the land use dynamic features of the area to be evaluated are extracted, including: for each group target type, based on the land use dataset of the area to be evaluated, determining the single land use change dynamic feature corresponding to the group target type.

[0014] Furthermore, based on the target sample area group corresponding to the group target type and the land use dataset of the area to be evaluated, the land use dynamic features of the area to be evaluated are extracted, including: for any two group target types, based on the land use dataset of the area to be evaluated, determining the land use transfer dynamic features corresponding to the two group target types.

[0015] Compared with existing technologies, the comprehensive assessment method for air pollution levels based on LUCC provided in this specification has at least the following advantages:

[0016] 1. By comprehensively considering land use datasets and air pollution concentration sequences, the relationship between air pollution and land use can be more accurately reflected, thus achieving a precise assessment of air pollution levels. Grouping multiple sample areas and establishing specific air pollution assessment models for each group simplifies the assessment process and improves efficiency. This method can handle various types of land use and air pollution concentration data, making it applicable to different geographical environments and time scales. By analyzing the regional feature set of the area to be assessed, the most suitable model can be selected from multiple sample area groups for assessment, thereby achieving customized air pollution assessment. Accurate air pollution assessment can provide scientific evidence for environmental protection policymakers, helping them formulate more effective emission reduction measures and environmental protection strategies. This method combines advanced technologies such as big data analysis, machine learning, and Geographic Information Systems (GIS), contributing to technological progress and innovation in the field of environmental protection.

[0017] 2. By considering land use datasets and air pollution concentration sequences for each sample area, this method enables a more refined assessment of air pollution levels. It relies not only on single air pollution data but also incorporates land use as a crucial factor, providing a more comprehensive assessment. By calculating the first correlation coefficient between land use type and air pollution concentration, and the second correlation coefficient between two land use types, this method captures the dynamic relationship between land use change and air pollution. This helps reveal which land use types or changes have a significant impact on air pollution, thus guiding policy-making and environmental protection measures. Grouping multiple sample areas based on target type and correlation coefficient similarity makes the assessment more targeted and comparable. Sample areas within the same group share similarities in land use and air pollution, which helps identify common problems and challenges and develop corresponding solutions. By comprehensively utilizing the first and second correlation coefficients and employing clustering algorithms to group sample areas, this method reduces subjectivity and uncertainty in the assessment process, improving the accuracy of the assessment.

[0018] 3. By establishing an air pollution assessment model and training it with multiple training samples, this method can provide more accurate and reliable assessment results. The model can capture the complex relationship between land use and air pollution, thereby achieving accurate prediction and assessment of air pollution levels. By determining the target regional characteristic factors corresponding to the sample area groups, this method can identify regional characteristics that have a significant impact on air pollution assessment. This helps to focus on these characteristics in subsequent assessments, improving the relevance and accuracy of the assessment. Based on the regional characteristic set of the area to be assessed, the target sample area group is determined from multiple sample area groups. This method can find sample area groups with similar characteristics to the area to be assessed. This makes the assessment results closer to the actual situation of the area to be assessed, improving the applicability and accuracy of the assessment. By extracting the dynamic characteristics of land use in the area to be assessed, this method can capture the dynamic impact of land use change on air pollution levels. This helps to reveal the temporal relationship between land use change and air pollution, providing a scientific basis for formulating effective environmental protection measures. This method not only considers the impact of a single land use type on air pollution, but also the impact of the transfer and change between land use types on air pollution. This makes the assessment results more comprehensive and integrated, and can more accurately reflect the actual situation of air pollution levels. By providing accurate assessments of air pollution levels and analyses of dynamic land use characteristics, this method can offer scientific evidence and decision support for governments in formulating environmental protection policies, urban planning schemes, and land use policies. This helps reduce air pollution, improve environmental quality, and promote sustainable development. Attached Figure Description

[0019] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0020] Figure 1 This is a flowchart illustrating a comprehensive assessment method for air pollution levels based on LUCC, as shown in one embodiment of this application. Detailed Implementation

[0021] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0022] Figure 1 This is a flowchart illustrating a comprehensive assessment method for air pollution levels based on LUCC, as shown in one embodiment of this application. Figure 1 As shown, a comprehensive assessment method for air pollution levels based on LUCC may include the following steps.

[0023] Step 110: Obtain the regional feature set, land use dataset, and air pollution concentration sequence of multiple sample areas.

[0024] The regional feature set includes feature data corresponding to multiple regional feature factors, the land use dataset includes land use data of various types at multiple time points, and the air pollution concentration sequence includes air pollution concentrations at multiple time points.

[0025] Specifically, multiple regional characteristic factors may include environmental temperature factors, environmental humidity factors, and environmental atmospheric pressure factors.

[0026] Land types can include farmland, forests, grasslands, water bodies, wasteland, and urban land. Land use data for these various land types can include the area used for each type.

[0027] Air pollution concentrations can include PM10 concentrations, etc.

[0028] Step 120: For each sample area, determine the target type corresponding to the sample area based on the land use dataset and air pollution concentration sequence of the sample area.

[0029] Specifically, it includes:

[0030] For each type, the first correlation coefficient between the type and the air pollution concentration in the sample area is determined based on the land use dataset and air pollution concentration sequence of the sample area.

[0031] For each pair of types, based on the land use dataset and air pollution concentration sequence of the sample area, the second correlation coefficient between the two types and the air pollution concentration of the sample area is determined;

[0032] The target type corresponding to the sample area is determined based on the first correlation coefficient between each type and the atmospheric pollution concentration of the sample area, and the second correlation coefficient between each pair of types and the atmospheric pollution concentration of the sample area.

[0033] Specifically, the first correlation coefficient between the type and the air pollution concentration in the sample area can be calculated using the following formula:

[0034]

[0035] Among them, R (i,1) S is the first correlation coefficient between the i-th type and the air pollution concentration in the sample area. (i,t) For, ρ t Let T be the atmospheric pollution concentration in the sample area at time t, and T be the total number of sampling time points.

[0036] Preferably, based on the land use dataset and air pollution concentration sequence of the sample area, two types of second correlation coefficients between the sample area and the air pollution concentration of the sample area are determined, including:

[0037] Based on the land use dataset of the sample area, calculate land use transfer data for multiple time points corresponding to the two types;

[0038] Based on land use transfer data at multiple time points corresponding to the two types and the atmospheric pollution concentration sequence of the sample area, the second correlation coefficient between the two types and the atmospheric pollution concentration of the sample area was determined.

[0039] Specifically, for each point in time, the difference between the land use areas of the two types at that point in time can be used as the land use transfer data for the two types at that point in time.

[0040] The second correlation coefficient between the two types of air pollution concentrations in the sample area can be determined using the following formula:

[0041]

[0042] Among them, R ((i,j),2) Let ΔS be the second correlation coefficient between the i-th and j-th types of air pollution and the concentration of air pollution in the sample area. ((i,j),t) Let be the difference between the utilization area of ​​the i-th type of land and the utilization area of ​​the j-th type of land in the sample area at time t.

[0043] Types with a first correlation coefficient greater than the first correlation coefficient threshold can be used as the first target type corresponding to the sample region.

[0044] For each first target type, a type whose second correlation coefficient with that first target type is greater than a second correlation coefficient threshold but is not a first target type can be considered as the second target type corresponding to that sample region. The target type corresponding to the sample region can include both the first target type and the second target type.

[0045] Step 130: Based on the target type corresponding to each sample region, divide the multiple sample regions into multiple sample region groups.

[0046] Specifically, it includes:

[0047] For any two sample regions, calculate the type similarity between the two sample regions based on the target types corresponding to the two sample regions;

[0048] The first clustering algorithm (e.g., K-means clustering, hierarchical clustering, etc.) is used to cluster multiple sample regions based on the type similarity between any two sample regions to determine the sample region clusters;

[0049] For each sample region cluster, based on the first correlation coefficient between the type of the sample region included in the cluster and the atmospheric pollution concentration of the sample region, and the second correlation coefficient between the two types and the atmospheric pollution concentration of the sample region, the similarity of the correlation coefficients between any two sample regions included in the cluster is calculated. Then, using a second clustering algorithm (e.g., K-means clustering, hierarchical clustering, etc.), multiple sample regions included in the cluster are clustered based on the similarity of the correlation coefficients between any two sample regions included in the cluster, thus determining multiple sample region groups.

[0050] Specifically, for any two sample regions, the type similarity between the two sample regions can be calculated using the following formula:

[0051]

[0052] Among them, S (e,f),type) Let N be the type similarity between the e-th sample region and the f-th sample region. (e,f) N represents the number of target types corresponding to both the e-th and f-th sample regions. e N represents the number of target types corresponding to the e-th sample region. f Let N be the number of target types corresponding to the f-th sample region, and max(N) e N f ) is to take N e and N f The larger value in the range.

[0053] Based on the target types corresponding to the sample regions included in the sample region cluster, the cluster target type corresponding to the sample region cluster is determined. For example, the target type corresponding to each sample region included in the sample region cluster is taken as the cluster target type.

[0054] The correlation coefficient similarity between any two sample regions included in the sample region cluster can be calculated based on the first correlation coefficient between the group target type and the atmospheric pollution concentration of the sample region corresponding to each sample region included in the sample region cluster, and the second correlation coefficient between the two group target types and the atmospheric pollution concentration of the sample region.

[0055] For example, the correlation coefficient similarity between any two sample regions included in a sample region cluster can be calculated using the following formula:

[0056]

[0057] Among them, S (e,f),correlation) R is the correlation coefficient similarity between any two sample regions included in the sample region cluster, where γ is a preset parameter, and γ is greater than 0. ((g,e),1)R is the first correlation coefficient between the target type of the g-th cluster corresponding to the e-th sample region and the atmospheric pollution concentration of the sample region. ((g,f),1) R is the first correlation coefficient between the target type of the g-th cluster corresponding to the f-th sample region and the atmospheric pollution concentration of the sample region. ((g,h,e),2) R is the second correlation coefficient between the target type of the g-th cluster and the target type of the h-th cluster corresponding to the e-th sample region and the atmospheric pollution concentration in the sample region. ((g,h,f),2) G represents the second correlation coefficient between the g-th and h-th cluster target types corresponding to the f-th sample region and the atmospheric pollution concentration in the sample region, where G is the total number of cluster target types corresponding to the sample region cluster.

[0058] Step 140: For each sample area group, establish an air pollution assessment model based on the target type, land use dataset, and air pollution concentration sequence corresponding to the sample areas included in the sample area group.

[0059] Specifically, it includes:

[0060] Based on the target types corresponding to the sample regions included in the sample region group, determine the group target type corresponding to the sample region group. For example, take the target type corresponding to each sample region included in the sample region group as the group target type.

[0061] Based on the target type of the sample area group and the land use dataset and air pollution concentration sequence of the sample area included in the sample area group, multiple training samples are established. The training samples may include land use data of the target type of the sample area group extracted from the land use dataset of the sample area according to the target type of the sample area group. The label of the training samples may be the air pollution concentration sequence of the sample area.

[0062] An air pollution assessment model is established and trained using multiple training samples. The air pollution assessment model can be a Long Short Term Memory Network (LSTM) model.

[0063] Step 150: For each sample region group, determine the target region feature factor corresponding to the sample region group based on the region feature set of the sample regions included in the sample region group.

[0064] Specifically, it includes:

[0065] For each regional feature factor, the difference parameter corresponding to the regional feature factor is calculated based on the regional feature set of the sample regions included in the sample region group.

[0066] Based on the difference parameters corresponding to each regional characteristic factor, the target regional characteristic factors corresponding to the sample regional group are determined.

[0067] Specifically, the difference parameters corresponding to the regional characteristic factors can be calculated using the following formula:

[0068]

[0069] Among them, D k Let V be the difference parameter corresponding to the k-th regional feature factor, U be the total number of sample regions included in the sample region group, and V be the difference parameter. (u,k) Let be the feature value of the k-th region feature factor corresponding to the u-th sample region included in the sample region group.

[0070] Regional feature factors whose difference parameters are less than the difference parameter threshold can be used as target regional feature factors for the sample region group.

[0071] Step 160: Obtain the regional feature set and land use dataset of the area to be evaluated.

[0072] Step 170: Determine the target sample region group from multiple sample region groups based on the region feature set of the region to be evaluated.

[0073] Specifically, it includes:

[0074] For each sample region group, the group feature data corresponding to the sample region group is determined based on the target region feature factor corresponding to the sample region group and the region feature set of the sample regions included in the sample region group. The group feature data corresponding to the sample region group may include the feature value of the target region feature factor corresponding to the sample region group. For example, the feature value of the target region feature factor corresponding to the sample region group can be obtained by averaging the feature values ​​of the target region feature factors corresponding to the sample regions included in the sample region group.

[0075] For each sample region group, the feature data corresponding to the region to be evaluated is determined based on the target region feature factors corresponding to the sample region group and the region feature set of the region to be evaluated. The feature data corresponding to the region to be evaluated may include the feature values ​​of the target region feature factors corresponding to the sample region group corresponding to the region to be evaluated. Based on the group feature data corresponding to the sample region group and the feature data corresponding to the region to be evaluated, the feature similarity between the region to be evaluated and the sample region group is calculated.

[0076] The target sample region group is determined from multiple sample region groups based on the feature similarity between the region to be evaluated and each sample region group.

[0077] Specifically, the cosine similarity between the feature data corresponding to the region to be evaluated and the group feature data corresponding to the sample region group can be calculated as the feature similarity between the region to be evaluated and the sample region group. The sample region group with the highest feature similarity is selected as the target sample region group.

[0078] Step 180: Determine the air pollution level of the area to be assessed based on the land use dataset of the area to be assessed using the air pollution assessment model corresponding to the target sample area group.

[0079] Specifically, it includes:

[0080] Based on the target sample area group corresponding to the group target type and the land use dataset of the area to be evaluated, extract the dynamic characteristics of land use in the area to be evaluated.

[0081] The air pollution level of the area to be assessed is determined by using the air pollution assessment model corresponding to the target sample area group and the dynamic characteristics of land use in the area to be assessed.

[0082] Preferably, based on the target sample area group corresponding to the target target type and the land use dataset of the area to be evaluated, the dynamic characteristics of land use in the area to be evaluated are extracted, including:

[0083] For each target group, the dynamic characteristics of single land use change corresponding to the target group are determined based on the land use dataset of the area to be evaluated.

[0084] Specifically, based on the land use dataset of the area to be evaluated, a single land use change dynamic index corresponding to multiple consecutive time periods for each target type can be calculated. Based on the single land use change dynamic index corresponding to multiple consecutive time periods for each target type, the sequence of single land use change dynamic indices corresponding to each target type can be determined.

[0085] The following formulas can be used to determine the dynamic index of a single land use change corresponding to a target type:

[0086]

[0087] Among them, K i U is the single land use change dynamic index corresponding to the i-th target type. (b,i) Let U be the area of ​​land of the i-th target type at the end of the time period. (a,i) Let T be the area of ​​the land of the i-th target type at the beginning of the time period, and T be the duration of the time period.

[0088] For each target group type, variational mode decomposition is performed on the single land use change dynamic index sequence corresponding to the target group type to obtain multiple single land use change dynamic index mode components. The time domain features and frequency domain features of the single land use change dynamic index mode components are extracted. The single land use change dynamic features corresponding to the target group type can include the time domain features and frequency domain features of each single land use change dynamic index mode component.

[0089] Preferably, based on the target sample area group corresponding to the target target type and the land use dataset of the area to be evaluated, the dynamic characteristics of land use in the area to be evaluated are extracted, including:

[0090] For any two target types, based on the land use dataset of the area to be evaluated, determine the dynamic characteristics of land use transfer corresponding to the two target types.

[0091] Specifically, for any two groups of target types, based on the land use dataset of the area to be evaluated, the land use transfer area corresponding to the two groups of target types is calculated, and a transfer area sequence corresponding to the two groups of target types is generated. Variational mode decomposition can be performed on the transfer area sequence corresponding to the two groups of target types to obtain multiple land use transfer mode components. The time domain features and frequency domain features of the land use transfer mode components are extracted. The dynamic features of land use transfer corresponding to the two groups of target types can include the time domain features and frequency domain features of each land use transfer mode component.

[0092] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A comprehensive assessment method for air pollution levels based on LUCC, characterized in that, include: The method involves acquiring regional feature sets, land use datasets, and air pollution concentration sequences for multiple sample areas. The regional feature sets include feature data corresponding to multiple regional feature factors, the land use datasets include land use data for various land types at multiple time points, and the air pollution concentration sequences include air pollution concentrations at multiple time points. For each sample region, the target type corresponding to the sample region is determined based on the land use dataset and air pollution concentration sequence of the sample region; including: For each type, a first correlation coefficient between the type and the air pollution concentration in the sample area is determined based on the land use dataset and air pollution concentration sequence of the sample area; for every two types, a second correlation coefficient between the two types and the air pollution concentration in the sample area is determined based on the land use dataset and air pollution concentration sequence of the sample area; and the target type corresponding to the sample area is determined based on the first correlation coefficient between each type and the air pollution concentration in the sample area and the second correlation coefficient between every two types and the air pollution concentration in the sample area. Based on the target type corresponding to each of the sample regions, the multiple sample regions are divided into multiple sample region groups; For each of the sample area groups, an air pollution assessment model is established based on the target type, land use dataset, and air pollution concentration sequence corresponding to the sample areas included in the sample area group; For each sample region group, the target region feature factor corresponding to the sample region group is determined based on the region feature set of the sample regions included in the sample region group. Obtain the regional feature set and land use dataset of the area to be evaluated; Based on the regional feature set of the region to be evaluated, a target sample region group is determined from the plurality of sample region groups; The air pollution level of the area to be assessed is determined using the air pollution assessment model corresponding to the target sample area group and based on the land use dataset of the area to be assessed. This includes: Based on the target sample area group corresponding to the group target type and the land use dataset of the area to be evaluated, the land use dynamic characteristics of the area to be evaluated are extracted. For each group target type, the single land use change dynamic characteristics corresponding to the group target type are determined based on the land use dataset of the area to be evaluated. For any two group target types, the land use transfer dynamic characteristics corresponding to the two group target types are determined based on the land use dataset of the area to be evaluated. The air pollution level of the area to be assessed is determined by using the air pollution assessment model corresponding to the target sample area group and the dynamic characteristics of land use in the area to be assessed.

2. The comprehensive assessment method for air pollution levels based on LUCC according to claim 1, characterized in that, Based on the land use dataset and air pollution concentration sequence of the sample area, a second correlation coefficient between the two types and the air pollution concentration of the sample area is determined, including: Based on the land use dataset of the sample area, calculate land use transfer data at multiple time points corresponding to the two types; Based on land use transfer data at multiple time points corresponding to the two types and the air pollution concentration sequence of the sample area, a second correlation coefficient between the two types and the air pollution concentration of the sample area is determined.

3. The comprehensive assessment method for air pollution levels based on LUCC according to claim 1, characterized in that, Based on the target type corresponding to each of the sample regions, the plurality of sample regions are divided into a plurality of sample region groups, including: For any two sample regions, calculate the type similarity between the two sample regions based on the target types corresponding to the two sample regions; The first clustering algorithm is used to cluster the multiple sample regions based on the type similarity between any two sample regions to determine the sample region clusters; For each sample region cluster, based on the first correlation coefficient between the type of the sample region and the atmospheric pollution concentration of the sample region, and the two second correlation coefficients between the type and the atmospheric pollution concentration of the sample region, the correlation coefficient similarity between any two sample regions included in the sample region cluster is calculated. Then, based on the correlation coefficient similarity between any two sample regions included in the sample region cluster, the multiple sample regions included in the sample region cluster are clustered using a second clustering algorithm to determine multiple sample region groups.

4. The comprehensive assessment method for air pollution levels based on LUCC according to claim 1, characterized in that, Based on the target type, land use dataset, and air pollution concentration sequence corresponding to the sample areas included in the sample area group, an air pollution assessment model is established, including: Based on the target types corresponding to the sample regions included in the sample region group, determine the group target type corresponding to the sample region group; Based on the target type of the sample area group and the land use dataset and air pollution concentration sequence of the sample areas included in the sample area group, multiple training samples are established; An air pollution assessment model is established, and the air pollution assessment model is trained using the multiple training samples.

5. A comprehensive assessment method for air pollution levels based on LUCC according to any one of claims 1-4, characterized in that, Based on the regional feature set of the sample regions included in the sample region group, the target region feature factors corresponding to the sample region group are determined, including: For each of the regional feature factors, the difference parameter corresponding to the regional feature factor is calculated based on the regional feature set of the sample regions included in the sample region group; Based on the difference parameter corresponding to each of the regional feature factors, the target regional feature factors corresponding to the sample regional group are determined.

6. A comprehensive assessment method for air pollution levels based on LUCC according to any one of claims 1-4, characterized in that, Based on the regional feature set of the region to be evaluated, a target sample region group is determined from the plurality of sample region groups, including: For each sample region group, the group feature data corresponding to the sample region group is determined based on the target region feature factor corresponding to the sample region group and the region feature set of the sample regions included in the sample region group. For each sample region group, the feature data corresponding to the region to be evaluated is determined based on the target region feature factor corresponding to the sample region group and the region feature set of the region to be evaluated. The feature similarity between the region to be evaluated and the sample region group is calculated based on the group feature data corresponding to the sample region group and the feature data corresponding to the region to be evaluated. The target sample region group is determined from the plurality of sample region groups based on the feature similarity between the region to be evaluated and each of the sample region groups.

Images