A method and system for identifying mismatch between urban heat environment and green environment

Abstract

The application provides a kind of city heat environment and green environment mismatch identification method and system, comprising: through the input of anonymous job and residence coordinates, remote sensing ground temperature and vegetation index extraction, job and residence heat and green environment mismatch index construction, rule model and clustering model joint identification, spatial unit aggregation and cooling and greening measures identification code output, realize the automatic identification of the relative difference between the heat environment and green environment of the residential area and the working place of urban residents, so as to improve the spatial accuracy, stability and planning application value of city heat and green environment risk identification.

Description

Technical Field

[0001] This invention relates to the field of information data processing technology, and in particular to a method and system for identifying mismatches between urban thermal environment and green environment. Background Technology

[0002] Against the backdrop of accelerated urbanization and the increasing impact of climate change, the urban heat island effect and insufficient green space have become significant issues in urban thermal environment management and thermal resilience planning. Residents' daily activities in cities are not confined to a single residence but typically shift between multiple stable activity scenarios, such as residence and workplace. For working individuals, residence and workplace are two primary daily activity spaces. If there are significant differences in surface temperature or vegetation cover between residence and workplace, some individuals may experience higher thermal exposure and lower green space availability during work or residential periods. Therefore, accurately identifying the thermal-green environmental differences and their spatial distribution between residence and workplace is crucial for optimizing the allocation of urban cooling and greening resources and improving the accuracy of thermal environmental risk identification.

[0003] Existing technologies include methods that retrieve land surface temperature from remote sensing images and use the normalized difference in vegetation index (NDVI) or other vegetation indices to evaluate the urban green environment. Some methods also incorporate population distribution, land use, or administrative unit data to identify areas with excessive heat exposure, insufficient green coverage, or high-risk population concentrations. While these methods can reflect the macroscopic distribution characteristics of the urban thermal environment and green spaces to some extent, they still have shortcomings in identifying environmental differences between work and residence scenarios and in their application in planning.

[0004] First, existing urban thermal risk assessment methods mostly use residential areas, communities, streets, or regular grids as the main spatial benchmarks. While these methods can reflect the environmental conditions around residential areas or within fixed spatial units, they struggle to identify changes in the thermal and green environments experienced by residents in their workplaces or commuting-related spaces. When the surface temperature at the workplace is higher than that at the residential area, or the vegetation index at the workplace is lower than that at the residential area, environmental assessments based solely on the residential area may underestimate the adverse thermal-green environmental conditions of the target population during daytime activities, thus affecting the accuracy of spatial allocation of cooling and greening resources.

[0005] Secondly, existing methods for identifying differences between work and residence environments mostly focus on aspects such as air pollution exposure, commuting distance, commuting carbon emissions, or employment accessibility. Methods for identifying work-residence mismatches in terms of surface temperature and green environment are still relatively insufficient. In reality, there are differences between work-related geothermal-low-green-environment states and residential geothermal-low-green-environment states in terms of spatial location, duration of action, and the configuration of remediation measures. If only the thermal environment or green environment is evaluated separately, or if the mismatch is not distinguished between the workplace and the residence, it is difficult to generate spatial identification results tailored to different scenarios.

[0006] Furthermore, existing spatial environment assessment methods typically output a single risk level map or a single indicator distribution map, rarely linking object-level work-residence dual-address information with spatial management units such as communities, streets, grids, and industrial parks. Even if high-temperature zones or low-green-area zones can be identified, it is difficult to further generate spatial unit-level governance priority results that match the direction and intensity of the mismatch, and even more difficult to form data-driven measure codes that can be directly written into GIS layer attribute tables or planning platforms.

[0007] Furthermore, data privacy protection is a technical issue that needs to be addressed when utilizing commuting location data, workplace / residence coordinate data, or mobile terminal location data. Directly using workplace / residence location data containing personally identifiable information poses a risk of privacy breaches; excessive obfuscation of location data may reduce the accuracy of spatial environmental feature extraction and mismatch identification. Therefore, a data processing method that balances privacy protection and identification accuracy needs to be developed, balancing the anonymization of target object identification numbers with the calculability of spatial coordinates.

[0008] In summary, there is an urgent need for a method and system that can automatically calculate the differences in thermal environment and green environment between workplace and residence based on anonymized workplace and residence coordinates, surface temperature raster data, and vegetation index raster data, identify the direction and intensity of mismatch, and aggregate the object-level identification results into a spatial unit-level governance priority layer and a cooling and greening measure identification code. Summary of the Invention

[0009] This invention provides a method and system for identifying mismatches between urban thermal and green environments. It addresses the shortcomings of existing technologies, such as the lack of ability to identify composite mismatches between thermal and green environments based on both residential and workplace locations, and the inability to transform object-level mismatch identification results into governance basis at the spatial management unit level. This invention enables automatic identification of the relative differences in thermal and green environments between urban residents' residential and workplace locations, thereby improving the spatial accuracy, stability, and planning application value of urban thermal and green environment risk identification.

[0010] In a first aspect, the present invention provides a method for identifying mismatches between urban thermal environment and green environment, comprising: Anonymized occupational and residential coordinate data and remote sensing image data are acquired and processed in a unified manner to generate anonymized occupational and residential spatial datasets, surface temperature raster data and vegetation index raster data; Based on the anonymized work-residence space dataset and raster data, a spatial neighborhood is constructed and environmental features are extracted. The work-residence environment difference is calculated, and thermal environment mismatch index and green environment mismatch index are generated. A standardized mismatch feature vector is constructed. Based on the standardized mismatch feature vector, a mismatch type identification process combining a threshold rule model and a clustering model is executed, and after consistency verification, the mismatch type and mismatch intensity results at the target object level are generated. Based on the mismatch type, determine the spatial aggregation location, aggregate the target object level results to the spatial management unit, and generate differentiated workplace governance priority scores and residential governance priority scores by combining the mismatch object ratio and average mismatch intensity of each unit. Based on the governance priority score, mismatch type, and dominant dimensions of thermal and green environment mismatch, the system automatically matches the identification code of cooling and greening measures, and generates spatial output results with governance priority and measure identification code. The above process is repeated at multiple spatial neighborhood scales to convert multi-scale identification consistency into confidence level and scale sensitivity label, which are then superimposed on the spatial output result.

[0011] Secondly, the present invention also provides a system for identifying mismatches between urban thermal environment and green environment, comprising: The acquisition module is used to acquire anonymized occupational and residential coordinate data and remote sensing image data and process them in a unified manner to generate anonymized occupational and residential spatial datasets, surface temperature raster data and vegetation index raster data. The construction module is used to construct a spatial neighborhood and extract environmental features based on the anonymized work-residence space dataset and raster data, calculate the work-residence environment difference, generate thermal environment mismatch index and green environment mismatch index, and construct a standardized mismatch feature vector. The identification module is used to perform a mismatch type identification process that combines a threshold rule model and a clustering model based on the standardized mismatch feature vector, and generate the target object level mismatch type and mismatch intensity results after consistency verification. The generation module is used to determine the spatial aggregation location based on the mismatch type, aggregate the target object-level results to the spatial management unit, and generate differentiated workplace governance priority scores and residential governance priority scores by combining the proportion of mismatched objects and the average mismatch intensity of each unit. The matching module is used to automatically match the identification code of cooling and greening measures based on the governance priority score, mismatch type and dominant dimension of thermal and green environment mismatch, and generate spatial output results with governance priority and measure identification code; The verification module is used to repeatedly execute the above process at multiple spatial neighborhood scales, converting multi-scale recognition consistency into confidence level and scale sensitivity label, which are then superimposed on the spatial output result.

[0012] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the urban thermal environment and green environment mismatch identification method as described above.

[0013] The method and system for identifying mismatches between urban thermal environment and green environment provided by this invention have the following beneficial effects: (1) Improve the ability to identify complex mismatches between work-residence heat and green environment. This invention uses residential coordinates and work coordinates as dual spatial benchmarks to construct thermal environment mismatch indicators and green environment mismatch indicators. These indicators can simultaneously characterize the direction and magnitude of differences in both thermal and green environment dimensions between work and residence, and identify complex mismatch states such as thermal-low green type at work and thermal-low green type at residence. Compared with evaluation methods that rely solely on a single residence or a single environmental indicator, this invention can more accurately reflect the differences in thermal and green environment between the target object and the actual daily activity space.

[0014] (2) Balancing interpretability, adaptability, and result stability in mismatch identification. This invention employs a dual-mode identification approach combining rule-based and clustering models. The threshold rule-based model provides stable and reproducible baseline classification results, while the prior-guided clustering model adapts to the mismatch identification needs under different city and sample distribution conditions. Simultaneously, this invention repeatedly performs environmental feature extraction, mismatch type identification, and governance priority judgment at multiple spatial neighborhood scales, converting the consistency of identification at different scales into confidence levels and scale sensitivity labels. Through this processing method, this invention can reduce the subjective influence of selecting a single buffer scale or a single neighborhood range; for mismatch results that remain consistent across multiple scales, a higher confidence level can be assigned; for spatial units that undergo type conversion or priority changes with scale, they can be labeled as scale-sensitive objects, thus providing a basis for planners to distinguish between stable high-risk units and scale-sensitive units, improving the stability, interpretability, and reliability of mismatch identification results for planning use.

[0015] (3) Improve the precision and spatial orientation of governance priority judgment. This invention determines the spatial aggregation location based on the mismatch type, aggregating unfavorable conditions of the workplace to the spatial management unit where the workplace is located, and aggregating unfavorable conditions of the residence to the spatial management unit where the residence is located, thereby avoiding misattributing workplace risks to residence units or vice versa. Furthermore, this invention generates a governance priority score by comprehensively considering the proportion of mismatch objects, average mismatch intensity, severe mismatch proportion, and type confidence, which can simultaneously reflect the coverage, severity, and identification reliability of mismatch risks. Compared with the evaluation method that only uses the proportion of objects, this invention can avoid simply judging spatial units with a large number of low-intensity mismatch objects as high priority; compared with the evaluation method that only uses the average mismatch intensity, this invention can also avoid ignoring spatial units with a large coverage of mismatch objects. Therefore, this invention can more precisely identify priority governance units that have both high mismatch intensity and a large impact range, and whose identification results are relatively stable.

[0016] (4) Enhance the application and transformation capabilities of assessment results and reduce the risk of privacy leakage. Based on the governance orientation, mismatch intensity level, and the dominant dimension of thermal and green environment mismatch, this invention automatically matches the identification code for cooling and greening measures, and writes the output results into GIS layer attribute fields, tables, reports, interface data, or visualization interfaces, enabling the identification results to be directly imported into the urban planning GIS platform. Simultaneously, this invention performs spatial calculations based on anonymized target object numbers, residential coordinates, and workplace coordinates, without reading personal identification information such as names, ID numbers, mobile phone numbers, or detailed house numbers, thereby reducing the risk of personal privacy leakage while ensuring the accuracy of spatial calculations. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the method for identifying mismatches between urban thermal environment and green environment provided by the present invention. Figure 2 This is a schematic diagram of the module structure provided by the present invention; Figure 3 This is a schematic diagram of the spatial neighborhood construction and environmental feature extraction of residential and workplace locations provided by the present invention; Figure 4 This is a schematic diagram illustrating the identification of mismatches between work-residence heat and green environment provided by the present invention; Figure 5 This is a schematic diagram illustrating the aggregation of object-level mismatch results to the space management unit provided by the present invention; Figure 6 This is a schematic diagram of the multi-scale consistency verification and final spatial output provided by the present invention; Figure 7 This is a schematic diagram of the urban thermal environment and green environment mismatch identification system provided by the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0020] To address the shortcomings of existing technologies in identifying complex mismatches between thermal and green environments using both residential and workplace locations as dual spatial benchmarks, and the inability to translate object-level mismatch identification results into spatial management unit-level governance criteria, this invention aims to provide a method and system for identifying urban thermal and green environment mismatches based on dual residential and workplace addresses and remote sensing data. This addresses the problems in existing technologies, such as the inability of a single residential spatial benchmark to reflect differences between residential and workplace environments, the inability to distinguish the direction and intensity of mismatches, and the difficulty in translating object-level identification results into spatial governance criteria. This method achieves automatic identification of the relative differences in thermal and green environments between urban residents' residential and workplace locations through anonymized residential and workplace coordinate input, extraction of remote sensing surface temperature and vegetation indices, construction of residential and workplace thermal and green environment mismatch indicators, joint identification using rule-based and clustering models, spatial unit aggregation, and output of cooling and greening measure identification codes. This improves the spatial accuracy, stability, and planning application value of urban thermal and green environment risk identification.

[0021] Figure 1 This is a flowchart illustrating the method for identifying urban thermal environment and green environment mismatch provided in an embodiment of the present invention, as shown below. Figure 1 As shown, it includes: Step 100: Obtain anonymized occupational and residential coordinate data and remote sensing image data and process them uniformly to generate anonymized occupational and residential spatial dataset, surface temperature raster data and vegetation index raster data; Step 200: Based on the anonymized work-residence space dataset and raster data, construct a spatial neighborhood and extract environmental features, calculate the work-residence environment difference, generate thermal environment mismatch index and green environment mismatch index, and construct a standardized mismatch feature vector. Step 300: Based on the standardized mismatch feature vector, perform a mismatch type identification process that combines a threshold rule model and a clustering model, and generate the target object-level mismatch type and mismatch intensity results after consistency verification; Step 400: Determine the spatial aggregation location based on the mismatch type, aggregate the target object-level results to the spatial management unit, and generate differentiated workplace governance priority scores and residential governance priority scores by combining the mismatch object ratio and average mismatch intensity of each unit. Step 500: Based on the governance priority score, mismatch type, and dominant dimension of thermal and green environment mismatch, automatically match the cooling and greening measure identification code to generate a spatial output result with governance priority and measure identification code; Step 600: Repeat the above process at multiple spatial neighborhood scales to convert the multi-scale identification consistency into confidence level and scale sensitivity label, and superimpose them onto the spatial output result.

[0022] Specifically, embodiments of the present invention include, for example... Figure 1As shown, the process includes: acquiring and uniformly processing anonymized work-residence coordinate data and remote sensing image data; constructing spatial neighborhoods of residential and workplace locations and extracting environmental features; generating a feature vector of mismatch between work-residence heat and green environment; performing mismatch type identification by combining rule models and clustering models; aggregating target object-level results into spatial management units; matching identification codes for cooling and greening measures; and performing multi-scale consistency verification.

[0023] Step 1: Acquire anonymized work-residence coordinate data and remote sensing image data for identifying mismatches between work-residence heat and green environment, and perform unified processing on the data based on preset spatial reference parameters, target time range, and data quality control rules. Specifically, this includes: Step 11: Obtain the work and residence coordinates of multiple anonymized target objects within the target area.

[0024] In this embodiment, the work-residence coordinate data includes anonymized target object ID, residential coordinates, and workplace coordinates. The anonymized target object ID is used to associate the residential and workplace information of the same target object in subsequent data processing, but does not include name, ID number, mobile phone number, detailed house number, or other personally identifiable information.

[0025] The work-residence coordinate data is derived from anonymized spatial location data, including but not limited to questionnaire survey data, commuting OD data, spatial data of employees of enterprises and institutions, mobile terminal positioning data, or other data that can characterize the spatial location of residence and workplace. The data processing steps of this invention do not depend on a specific data source; any spatial dataset containing anonymized target object numbers, anonymized residential coordinates, and anonymized workplace coordinates can be used as input data for this invention.

[0026] When receiving the aforementioned data, the name, ID number, mobile phone number, detailed address, or other personally identifiable information in the data has been removed or replaced with an anonymized target object ID. During execution, this method does not read personally identifiable information; it only uses the anonymized target object ID, residential coordinates, and workplace coordinates for spatial calculations.

[0027] Step 12: Perform coordinate format standardization and field integrity verification on the work-residence coordinate data.

[0028] In this embodiment, the system first checks whether the work-residence coordinate data includes anonymized target object ID, residential coordinate field, and workplace coordinate field. Records with missing residential coordinates, missing workplace coordinates, incorrect coordinate format, empty target object ID, or duplicate target object ID are removed or marked.

[0029] Furthermore, coordinate data from different sources are unified into a preset coordinate format. This coordinate format can be latitude and longitude coordinates, projected coordinates, or other spatial coordinate formats readable by GIS platforms. After completing field integrity verification and coordinate format unification, an initial job and residence coordinate dataset is generated.

[0030] Step 13: Based on preset spatial reference parameters, perform spatial reference transformation and spatial range verification on the initial work-residence coordinate dataset.

[0031] In this embodiment, the initial work-residence coordinate dataset is converted to a spatial reference system uniformly used in the target area. The spatial reference system can be determined according to the map cartographic standards, remote sensing data projection methods, or GIS platform processing requirements of the country or region where the target area is located.

[0032] After completing the spatial reference transformation, the study area scope is verified for the residential and workplace coordinates. Coordinate records located outside the target area are discarded or marked. Coordinate records falling into inaccessible areas are discarded, marked, or manually verified. Inaccessible areas include water bodies, mountains, restricted areas, or other spatial areas unsuitable as residences or workplaces. Through the above processing, an anonymized work-residence spatial dataset is generated. This anonymized work-residence spatial dataset includes at least the anonymized target object ID, residential coordinates, workplace coordinates, and spatial reference information.

[0033] Step 14: Acquire remote sensing image data covering the target area within the target time range.

[0034] In this embodiment, the remote sensing image data includes thermal infrared remote sensing images for retrieving land surface temperature and multispectral remote sensing images for calculating vegetation indices. The target time range can be summer, heat wave period, monthly scale period, quarterly scale period, or user-specified period. In a preferred embodiment, the target time range is selected as the high-temperature period of summer to improve the ability to identify differences in urban thermal environment. The thermal infrared remote sensing images can be Landsat 8 or Landsat 9 thermal infrared band images. The multispectral remote sensing images can be Landsat series images, Sentinel-2 images, Harmonized Landsat-Sentinel images, or other remote sensing images that can be used to calculate vegetation indices.

[0035] Step 15: Based on data quality control rules, perform image-level screening and pixel-level preprocessing on the remote sensing image data.

[0036] In this embodiment, the image-level filtering includes image-level cloud cover filtering and coverage filtering. Image-level cloud cover filtering is used to remove remote sensing images with cloud cover rates higher than a preset threshold. In one example, the preset threshold can be 20% or 30%. Coverage filtering is used to ensure that the selected remote sensing images can cover the target area or the main analysis range of the target area.

[0037] The pixel-level preprocessing includes pixel-level cloud masking, cloud shadow removal, outlier pixel removal, radiometric correction, image cropping, image stitching, and temporal composite. Specifically, pixel-level cloud masking and cloud shadow removal reduce the impact of clouds and cloud shadows on the calculation of land surface temperature and vegetation index; outlier pixel removal removes pixels that significantly exceed reasonable value ranges; radiometric correction improves the comparability between different images; image cropping and image stitching generate continuous images covering the target area; and temporal composite forms stable input data for land surface temperature and vegetation index within the target time range.

[0038] Step 16: Generate land surface temperature raster data and vegetation index raster data based on the preprocessed remote sensing images.

[0039] In this embodiment, surface temperature raster data is generated based on preprocessed thermal infrared remote sensing images. The surface temperature raster data is used to characterize the surface thermal environment at different spatial locations within the target area.

[0040] Simultaneously, vegetation index raster data is generated based on the preprocessed multispectral remote sensing image. The vegetation index can be the normalized vegetation index, or it can be the enhanced vegetation index, the soil-regulated vegetation index, or other remote sensing indices used to characterize vegetation cover.

[0041] Taking the Normalized Difference Vegetation Index (NDVI) as an example, its calculation formula is:

[0042] Wherein, NIR represents the reflectance in the near-infrared band, and Red represents the reflectance in the red band.

[0043] After generating land surface temperature raster data and vegetation index raster data, spatial reference, spatial extent, pixel scale, and spatial alignment processing are performed on both to ensure that the land surface temperature raster data, vegetation index raster data, and anonymized occupational and residential spatial dataset have a unified spatial reference, spatial extent, and computational scale.

[0044] Step 17: Output the basic data after unified processing.

[0045] In this embodiment, step 1 ultimately outputs an anonymized work-residence space dataset, surface temperature raster data, and vegetation index raster data. The anonymized work-residence space dataset is used for subsequent construction of residential and workplace spatial neighborhoods, while the surface temperature raster data and vegetation index raster data are used for subsequent environmental feature extraction and identification of mismatches between work-residence thermal and green environments.

[0046] Step 2: Based on the anonymized work-residence space dataset, surface temperature raster data, and vegetation index raster data, construct spatial neighborhoods around the residential and work coordinates of each anonymized target object, extract surface temperature and vegetation index values ​​within the residential and work neighborhoods, and generate a mismatch feature vector between work-residence heat and green environment. Figure 3 This diagram illustrates the spatial neighborhood construction and environmental feature extraction of residential and workplace locations. (See also...) Figure 1 and Figure 3 Step 2 specifically includes: Step 21: Read the anonymized occupational and residential spatial dataset, surface temperature raster data, and vegetation index raster data generated in Step 1, and perform spatial overlay verification.

[0047] In this embodiment, the anonymized work-residence space dataset includes at least the anonymized target object ID, residential coordinates, and workplace coordinates. The surface temperature raster data is used to characterize the thermal environment state within the target area, and the vegetation index raster data is used to characterize the green environment state within the target area.

[0048] After reading the above data, spatial overlay verification is performed on the anonymized residential and work space dataset, land surface temperature raster data, and vegetation index raster data to determine whether the coordinates of the residence and workplace fall within the effective coverage area of ​​the land surface temperature raster data and vegetation index raster data. Records that do not fall within the effective coverage area are removed or marked. Through this processing, the coordinates of the residence and workplace, the land surface temperature raster data, and the vegetation index raster data are placed under the same spatial reference, spatial extent, and calculation scale.

[0049] Step 22: Construct a spatial neighborhood centered on the residential coordinates and workplace coordinates of each anonymized target object.

[0050] In this embodiment, for the i-th anonymized target object, its residential coordinates H are respectively... i and work location coordinates W i A residential spatial neighborhood and a workplace spatial neighborhood are constructed around the central location. The spatial neighborhood can be a circular buffer zone, a regular grid, a street block, or a service area generated based on the road network accessibility.

[0051] In one exemplary embodiment, a 500 m circular buffer zone is used as the main analysis scale to construct 500 m spatial neighborhoods for residential areas and 500 m spatial neighborhoods for workplaces. The 500 m scale can be used to characterize the surrounding environment of a residential or workplace. In other embodiments, the spatial neighborhood scale can be set to 100 m, 300 m, 1000 m, 1500 m, or other scales, depending on urban planning and management needs, walkable range, remote sensing data resolution, or user-defined parameters.

[0052] Step 23: Extract the surface temperature values ​​in the residential spatial neighborhood and the workplace spatial neighborhood respectively.

[0053] In this embodiment, the spatial neighborhoods of the residential location and the spatial neighborhood of the workplace are spatially overlaid with surface temperature raster data, and the surface temperature raster pixel values ​​within the neighborhood range are extracted. For the i-th anonymized target object, the surface temperature feature of the residential location is obtained. and surface temperature characteristics of the work site .

[0054] The surface temperature feature can be the mean, median, maximum, quantile, or area-weighted mean of surface temperature raster pixels in the neighborhood. In one exemplary embodiment, the mean of the effective surface temperature raster pixels in the neighborhood is used as the surface temperature feature of the spatial neighborhood.

[0055] Step 24: Extract vegetation index values ​​in the residential spatial neighborhood and the workplace spatial neighborhood respectively.

[0056] In this embodiment, the spatial neighborhoods of the residential area and the spatial neighborhood of the workplace are spatially overlaid with vegetation index raster data, and the vegetation index raster pixel values ​​within the neighborhood are extracted. For the i-th anonymized target object, the residential vegetation index features are obtained. and vegetation index characteristics of the work site .

[0057] The vegetation index feature can be the mean, median, maximum, quantile, or area-weighted mean of vegetation index raster pixels within the neighborhood. In one exemplary embodiment, the mean of the effective vegetation index raster pixels within the neighborhood is used as the vegetation index feature of the spatial neighborhood.

[0058] Step 25: Verify the validity of the extracted surface temperature characteristics of residential areas, surface temperature characteristics of workplaces, vegetation index characteristics of residential areas, and vegetation index characteristics of workplaces.

[0059] In this embodiment, for each anonymized target object's residential spatial neighborhood and workplace spatial neighborhood, the number of effective raster cells and the percentage of effective raster cell area are counted respectively. When the number of effective raster cells in the spatial neighborhood is lower than a preset threshold, or the percentage of effective raster cell area is lower than a preset percentage threshold, the corresponding record of the target object is removed or marked.

[0060] In one exemplary implementation, the effective raster cell area ratio threshold is set to 80%. When the effective cell area ratio in a certain spatial neighborhood is less than 80%, the environmental feature extraction result in that neighborhood is considered to be unstable, and a low effective cell mark is added to the record, or the record is removed from the subsequent mismatch index calculation process.

[0061] Step 26: Generate a thermal environment mismatch index based on the difference between the surface temperature characteristics of the workplace and the surface temperature characteristics of the residence.

[0062] In this embodiment, for the i-th anonymization target object, the thermal environment mismatch index Calculated using the following formula:

[0063] in, This represents the surface temperature characteristics within the spatial neighborhood of the work location of the i-th anonymized target object. This represents the surface temperature characteristics within the spatial neighborhood of the residence of the i-th anonymized target object.

[0064] when A value greater than 0 indicates that the surface temperature of the target object's work location is higher than the surface temperature of its residential location; when... When the value is less than 0, it indicates that the surface temperature of the target object's residential area is higher than the surface temperature of the working area; when When the value is close to 0, it indicates that the difference in thermal environment between the target's residence and workplace is small.

[0065] Step 27: Based on the differences between the vegetation index characteristics of the workplace and the vegetation index characteristics of the residence, a green environment mismatch index is generated.

[0066] In this embodiment, for the i-th anonymized target object, the green environment mismatch index Calculated using the following formula:

[0067] in, This represents the vegetation index characteristics within the spatial neighborhood of the work location of the i-th anonymized target object. This represents the vegetation index characteristics within the spatial neighborhood of the residence of the i-th anonymized target object.

[0068] when A value greater than 0 indicates that the vegetation index of the target object's workplace is higher than that of its residential area; when... A value less than 0 indicates that the vegetation index of the target object's workplace is lower than that of its residential area; when When the value is close to 0, it indicates that the difference in green environment between the target's residence and workplace is small.

[0069] Step 28: Based on the thermal environment mismatch index and the green environment mismatch index, construct the job-housing-thermal-green-environment mismatch feature vector for each anonymized target object.

[0070] In this embodiment, the feature vector of job-housing-heat and green environment mismatch includes at least a thermal environment mismatch index. Mismatched indicators with green environment To improve the interpretability of subsequent identification results, the job-housing-heat and green environment mismatch feature vector can further include the surface temperature characteristics of the residential area. Surface temperature characteristics of the work site Characteristics of vegetation index in residential areas Vegetation index characteristics of the work site Numerical features.

[0071] Meanwhile, the mismatch feature vector between the job-housing heat and green environment can be associated with identification fields such as spatial neighborhood scale identifier, anonymized target object number, and spatial management unit identifier of the target object. The numerical features are used to participate in subsequent standardization processing, mismatch type identification, and mismatch intensity calculation; the identification fields do not participate in standardization processing but are retained as grouping, association, or verification keys.

[0072] Step 29: Standardize the numerical features in the mismatch feature vector between work-housing-heat and green environment to generate a standardized mismatch feature vector between work-housing-heat and green environment.

[0073] In this embodiment, the thermal environment mismatch index is... Mismatched indicators with green environment Standardization is performed. Taking Z-score standardization as an example, the standardized thermal environment mismatch characteristic ZH is... i The standardized green environment mismatch feature ZG_i is calculated using the following formulas:

[0074]

[0075] in, and These represent the mean and standard deviation of the thermal environment mismatch index for all valid target objects within the target area, respectively. and These represent the mean and standard deviation of the green environment mismatch index for all valid target objects within the target area, respectively.

[0076] In other implementations, range standardization, quantile standardization, or robust standardization methods can be used to process numerical features. Standardization is not simply a data normalization operation, but rather an effort to eliminate dimensional differences between thermal environment mismatch indicators and green environment mismatch indicators. This allows surface temperature differences and vegetation index differences to participate in mismatch direction judgment, mismatch intensity calculation, and mismatch type identification on the same scale. This approach avoids the unreasonable amplification or weakening of a particular environmental dimension in composite mismatch judgment due to differences in the ranges of surface temperature and vegetation index values.

[0077] Step 210: Associate the standardized job-housing-heat and green environment mismatch feature vector with the corresponding identifier field to output a standardized job-housing-heat and green environment mismatch feature dataset with identifier information.

[0078] In this embodiment, the standardized job-housing-heat and green environment mismatch feature dataset with identification information includes at least anonymized target object number, residential coordinates, workplace coordinates, spatial neighborhood scale identifier, thermal environment mismatch index, green environment mismatch index, standardized thermal environment mismatch feature, standardized green environment mismatch feature, residential surface temperature feature, workplace surface temperature feature, residential vegetation index feature, and workplace vegetation index feature.

[0079] The standardized dataset of job-housing heat and green environment mismatch features with identification information serves as input data for subsequent mismatch type identification, mismatch intensity calculation, spatial management unit aggregation, and governance priority score calculation.

[0080] Step 3: Based on the standardized job-housing-heat and green environment mismatch feature vector, perform a mismatch type identification process that combines rule model and clustering model to generate target object-level job-housing-heat and green environment mismatch type and mismatch intensity results. Figure 4 A schematic diagram illustrating the identification of mismatches between work-life balance and green environment is shown. (See also...) Figure 1 and Figure 4 Step 3 specifically includes: Step 31: Read the standardized job-housing-thermal-green environment mismatch feature dataset with identification information generated in Step 2, and extract the standardized thermal environment mismatch features and standardized green environment mismatch features for mismatch type identification.

[0081] In this embodiment, for the i-th anonymization target object, the standardized thermal environment mismatch feature is denoted as . The standardized green environment mismatch characteristics are denoted as The aforementioned Used to characterize the direction and standardized magnitude of thermal environment differences between the workplace and residence. Used to characterize the direction and standardized magnitude of the green environment difference between the workplace and the residence.

[0082] Among them, when A value greater than 0 indicates that the workplace has a higher thermal environment exposure compared to the residence; when A value less than 0 indicates that the residence has a higher thermal exposure compared to the workplace. A value greater than 0 indicates that the workplace has a higher supply of green space compared to the residence; when A value less than 0 indicates that the workplace has a lower supply of green space compared to the residence.

[0083] Step 32: Set a threshold parameter for mismatch direction determination, and determine the mismatch direction for each anonymized target object based on the threshold parameter.

[0084] In this embodiment, a first hot mismatch threshold T is set. H and the first green mismatch threshold T G Among them, the first hot mismatch threshold T H Also known as the hot mismatch direction threshold, the first green mismatch threshold T G Also known as the green mismatch direction threshold. The T... H and T G Used to determine the direction of mismatch.

[0085] In one exemplary implementation, T H and T G All are taken as 0. At this time, when... Greater than 0 and When the value is less than 0, it indicates that the workplace of the i-th anonymized target object is in a hotter and lower greener state than its residence, and it is determined to be in the hot-low greener direction of the workplace; when Less than 0 and When the value is greater than 0, it indicates that the residence of the i-th anonymized target object is in a state of being hotter and greener than its workplace, and it is determined to be in the hot-low green direction of the residence.

[0086] In other implementations, T H and T G Alternatively, a value between 0.1 and 0.5 can be selected based on sample distribution, urban planning needs, or expert settings to reduce the impact of small differences close to zero on the judgment of mismatch direction.

[0087] Step 33: Based on the standardized thermal environment mismatch characteristics and the standardized green environment mismatch characteristics, calculate the mismatch intensity of each anonymized target object.

[0088] In this embodiment, the mismatch strength of the i-th anonymization target object Calculated using the following formula:

[0089] in, This represents the standardized thermal environment mismatch feature of the i-th anonymized target object. Let represent the standardized green environment mismatch feature of the i-th anonymized target object. It is used to characterize the overall mismatch intensity of the target object in both thermal and green environments.

[0090] Furthermore, a severity threshold T is set. R The severity threshold T R Used to determine the level of mismatch intensity. In one exemplary implementation, T R Take 1.5. When Less than T R When the mismatch intensity of the target object is classified into a moderate mismatch level; when Greater than or equal to T R When the mismatch intensity of the target object is classified into a severe mismatch level.

[0091] In other implementations, T R You can also base your decision on the target area. The quantiles, sample distribution characteristics, urban thermal risk management needs, or expert settings can be adjusted. For example, the quantiles, sample distribution characteristics, urban thermal risk management needs, or expert settings can be adjusted. Target objects located in the higher quantile range are classified as severely mismatched.

[0092] Step 34: Based on the mismatch direction and mismatch intensity level, generate the baseline mismatch type for each anonymized target object using a threshold rule model.

[0093] In this embodiment, the threshold rule model is executed according to the following rules: when >T H and < T G At that time, it was determined to be a working geothermal-low green direction; if <T R Then the target object is identified as a moderately working geothermal-low green type; if ≥ T R If so, the target object is identified as a severely working geothermal-low green type.

[0094] when < T H and >TG At that time, it was determined to be in the direction of residential geothermal energy - low green space; if <T R Then the target object is identified as a moderately residential geothermal-low green type; if ≥ T R If so, the target object is identified as a severely residential geothermal-low green type.

[0095] When a target object does not meet the above-mentioned criteria for judging the thermal-low green direction of a work site or a residential site, it is identified as a thermal-green compensation type, a low mismatch type, or a non-dominant thermal-low green mismatch type. Specifically, the thermal-green compensation type refers to a type where thermal environment differences and green environment differences do not simultaneously overlap in the unfavorable direction; the low mismatch type refers to a type where both thermal environment mismatch and green environment mismatch are at a low level; and the non-dominant thermal-low green mismatch type refers to a type where a clear dominant thermal-low green direction has not been formed for either the work site or the residential site.

[0096] Step 35: Construct a clustering model for adaptive identification, and input the standardized job-housing-heat and green environment mismatch feature vector into the clustering model.

[0097] In this embodiment, the clustering model is used to adaptively identify the mismatch between job-housing heat and green environment under different city conditions, sample distributions, or spatial neighborhood scales. The clustering model can be any one of the following: K-means clustering model, hierarchical clustering model, spectral clustering model, or unsupervised clustering model, with K-means clustering model being preferred.

[0098] In one exemplary implementation, a K-means clustering model is employed to standardize thermal environment mismatch features. Mismatch characteristics with standardized green environment As a clustering input variable, each anonymized target object is represented as a two-dimensional mismatch feature vector. .

[0099] In this embodiment, the number of clusters K is determined through a data-driven approach. Specifically, K=2, 3, 4, 5, and 6 are set respectively to perform K-means clustering on the standardized job-housing-heat and green environment mismatch feature vectors, and the clustering effect evaluation index under different K values ​​is calculated. The clustering effect evaluation index can be the silhouette coefficient, the Davis-Bolding index, the sum of squared errors corresponding to the elbow rule, or the information criterion. The K value that results in a high silhouette coefficient, a low Davis-Bolding index, or an inflection point in the decrease of the sum of squared errors is selected as the candidate number of clusters, and the optimal number of clusters is determined by combining the position of the cluster centers in the heat and green environment mismatch feature space.

[0100] In one exemplary embodiment, considering the clustering effect evaluation index, the spatial meaning of cluster centers, and the interpretation requirements of mismatch types, the number of clusters K is determined to be 4. Observing the clustering results when K=4, it is found that the four cluster centers naturally differentiate into two directional regions and two intensity levels in the standardized thermal and green environment mismatch feature space, corresponding to the moderate and severe levels of working geothermal-low green direction, and the moderate and severe levels of residential geothermal-low green direction, respectively, which is consistent with the four core mismatch type system of the present invention.

[0101] In other implementations, if the data distribution in the target region is different, the optimal number of clusters determined by the above data-driven method can also be 3, 5, or other values. In this case, the label assignment of the cluster categories can be interpreted according to the sign and modulus of the cluster centers, referring to the method in step 38.

[0102] Step 36, in a preferred embodiment, uses a priori-guided K-means clustering model to cluster the two-dimensional mismatch feature vectors.

[0103] In this embodiment, when the optimal number of clusters is determined to be 4 through the above data-driven approach, and the distribution pattern of the cluster centers is consistent with the known mismatch direction-intensity structure, the prior-guided K-means clustering model can be directly used to improve computational efficiency and result stability when identifying mismatches in new cities, new years, or new datasets.

[0104] In one exemplary implementation, the cluster number K is set to 4, and the prior cluster centers are determined based on the directional and intensity characteristics of the mismatch between heat and green environments. Specifically, the prior cluster centers can be set as follows: C1 = (0.5, 0.5) C2 = ( 0.5, 0.5) C3 = (1.5, 1.5) C4 = ( 1.5, 1.5) Among them, C1 and C3 are located in the fourth quadrant (ZH is positive and ZG is negative), corresponding to the working geothermal - low green direction. C1 is closer to the origin and corresponds to the moderate level, while C3 is farther from the origin and corresponds to the severe level. C2 and C4 are located in the second quadrant (ZH is negative and ZG is positive), corresponding to the residential geothermal - low green direction. C2 is closer to the origin and corresponds to the moderate level, while C4 is farther from the origin and corresponds to the severe level.

[0105] The sign orientation of the aforementioned prior cluster centers is set according to the physical meaning of the thermal and green environment mismatch index: the working site thermal-low green direction corresponds to a state where the working site is hotter and has lower greenness, i.e., ZH>0 and ZG<0; the residential site thermal-low green direction corresponds to a state where the residential site is hotter and has lower greenness, i.e., ZH<0 and ZG>0. By setting prior cluster centers consistent with the physical meaning of the mismatch, the clustering results can directly obtain a clear spatial environmental meaning and reduce the uncertainty of manual interpretation of labels after clustering.

[0106] In other implementations, the cluster number K can be adjusted to 3, 5, or other values ​​based on the sample distribution characteristics within the target region, silhouette coefficient, elbow rule, information criterion, or user-defined parameters. Cluster centers can also be generated using K-means++ initialization to further improve the stability of the clustering results.

[0107] Step 37: Perform clustering iterations and generate cluster categories.

[0108] In this embodiment, the feature vector X of each two-dimensional mismatch is first calculated. i The Euclidean distance between X and each cluster center, and X i The target object is assigned to the category corresponding to the nearest cluster center. Then, the cluster centers are recalculated based on the two-dimensional mismatch feature vectors of the target objects within each category. This assignment and update process is repeated until the change in cluster centers is less than a preset convergence threshold, or the number of iterations reaches a preset maximum value.

[0109] In one exemplary implementation, the preset convergence threshold is 10. -4 The maximum number of iterations is 300. After clustering is completed, the cluster category corresponding to each anonymized target object and the cluster center of that cluster category are obtained.

[0110] Step 38: Assign labels to each cluster category output by the clustering model.

[0111] In this embodiment, the cluster center of the k-th cluster category is denoted as . ,in This represents the center value of the cluster in the thermal environment mismatch dimension. This represents the center value of the cluster center on the green environment mismatch dimension.

[0112] when Greater than 0 and When the value is less than 0, the cluster category is marked as working geothermal - low green direction; when Less than 0 and When the value is greater than 0, the cluster category is labeled as the residential thermal-low green direction. Further, based on the cluster center strength... Determine the mismatch strength level of this cluster category, where Calculated using the following formula:

[0113] For clustering categories with the same incorrect formula, they can be classified according to... The levels of mismatch, from smallest to largest, are labeled as moderate mismatch and severe mismatch, respectively. This yields the cluster mismatch type output by the clustering model.

[0114] Step 39: Perform consistency verification between the baseline mismatch type output by the threshold rule model and the clustering mismatch type output by the clustering model, and determine the final mismatch type.

[0115] In this embodiment, when the baseline mismatch type output by the threshold rule model is consistent with the clustering mismatch type output by the clustering model, the mismatch type is taken as the final mismatch type of the target object, and its type confidence is marked as high.

[0116] When the identification results of the two are inconsistent, the final mismatch type is determined according to a preset priority rule, a distance determination rule, or a confidence determination rule. The preset priority rule can specify that either the threshold rule model result or the clustering model result should be used as the standard. The distance determination rule uses the distance between the two-dimensional mismatch feature vector of the target object and each cluster center as the basis, determining the type corresponding to the closest cluster center as the final mismatch type. The confidence determination rule uses cluster membership distance, silhouette coefficient, or model output stability as the basis, determining the model output result with higher confidence as the final mismatch type.

[0117] In one exemplary implementation, when the threshold rule model and the clustering model output inconsistent results, but both point to the same mismatch direction and differ only in intensity, the mismatch direction is retained, and the clustering is adjusted according to the mismatch intensity. With severity threshold T R The relationship is used to determine the severity level. When the mismatch directions are inconsistent, the mismatch type corresponding to the nearest cluster center is preferred, and the type confidence is marked as medium or low.

[0118] Step 310: Generate and output the results of the job-housing-heat and green environment mismatch types and mismatch intensity at the target object level.

[0119] In this embodiment, the results of the mismatch type and intensity of the workplace-residential-thermal and green environment at the target object level include at least the anonymized target object number, residential coordinates, workplace coordinates, spatial neighborhood scale identifier, thermal environment mismatch index, green environment mismatch index, standardized thermal environment mismatch characteristics, standardized green environment mismatch characteristics, mismatch direction, mismatch intensity, mismatch intensity level, baseline mismatch type, clustered mismatch type, final mismatch type, and type confidence.

[0120] The results of the mismatch type and intensity of the job-housing-heat and green environment at the target object level are used as input data for subsequent spatial management unit aggregation, governance priority scoring calculation and cooling and greening measure identification code matching.

[0121] Through the processing in step 3, this embodiment can identify working geothermal-low green type, residential geothermal-low green type and other mismatch types at the target object level, and further distinguish between moderate mismatch and severe mismatch. The rule model ensures the interpretability and reproducibility of the mismatch direction judgment, and the prior-guided clustering model improves the adaptability to different sample distribution conditions. The combination of the two can enhance the stability of the mismatch type identification results.

[0122] Step 4: Based on the results of the mismatch type and intensity of the job-housing heat and green environment at the target object level, aggregate the target objects into spatial management units such as communities, streets, rule grids, industrial parks or planning management units, and comprehensively calculate the number of objects of different mismatch types, object proportion, average mismatch intensity and severe mismatch proportion in each spatial management unit to generate workplace governance priority score and residential governance priority score.

[0123] Figure 5 This diagram illustrates the aggregation of object-level mismatch results to the space management unit. See also... Figure 1 and Figure 5 Step 4 specifically includes the following steps.

[0124] Step 41: Read the target object-level job-housing-heat and green environment mismatch identification results generated in Step 3.

[0125] In this embodiment, the target object-level job-housing-heat and green environment mismatch identification results include at least the anonymized target object number, residential coordinates, workplace coordinates, spatial neighborhood scale identifier, thermal environment mismatch index, green environment mismatch index, standardized thermal environment mismatch features, standardized green environment mismatch features, mismatch direction, mismatch intensity, mismatch intensity level, baseline mismatch type, clustered mismatch type, final mismatch type, and type confidence.

[0126] Among them, the final mismatch type is used to determine whether the target object belongs to the working geothermal-low green type, residential geothermal-low green type, thermal green compensation type, low mismatch type, or non-dominant thermal-low green mismatch type; the mismatch intensity is used to calculate the average mismatch intensity within the space management unit; the mismatch intensity level is used to identify severely mismatched objects; and the type confidence is used to characterize the reliability of the target object mismatch type identification results.

[0127] Step 42: Obtain spatial management unit data within the target area and preprocess the spatial management unit data.

[0128] In this embodiment, the spatial management unit can be a community, street, administrative district, regular grid, land use patch, industrial park, employment cluster, or other planning management unit. The spatial management unit data includes at least a spatial management unit identifier and a spatial management unit boundary.

[0129] The spatial management unit data undergoes spatial reference unification, boundary topology verification, and attribute field integrity verification to ensure it is within the same spatial reference system as the target object-level mismatch identification result generated in step 3. Records with overlapping boundaries, missing boundaries, or missing spatial management unit identifiers are corrected, removed, or marked.

[0130] Step 43: Determine the spatial aggregation location based on the final mismatch type of the target object.

[0131] In this embodiment, for target objects whose final mismatch type is "Working Geothermal - Low Greenery Type," their working location coordinates are used as the spatial aggregation location. The "Working Geothermal - Low Greenery Type" includes both moderately "Working Geothermal - Low Greenery Type" and severely "Working Geothermal - Low Greenery Type." Through this processing method, unfavorable environmental conditions at the working location are aggregated to the spatial management unit where the working location is located.

[0132] For target objects whose final mismatch type is residential geothermal-low green space, their residential coordinates are used as the spatial aggregation location. The residential geothermal-low green space type includes moderate residential geothermal-low green space and severe residential geothermal-low green space. Through this processing method, unfavorable environmental conditions of the residential location are aggregated to the spatial management unit where the residential location is located.

[0133] For target objects whose final mismatch type is hot-green compensation type, low mismatch type, or non-dominant hot-low green mismatch type, the residential coordinates, workplace coordinates, or both can be selected as spatial aggregation locations according to preset rules. These target objects can be used as reference samples or background distributions in the governance priority scoring calculation, or they can be excluded from the core scoring of workplace governance priority and residential governance priority.

[0134] Step 44: Based on the spatial aggregation location, associate the target object-level mismatch identification result with the corresponding spatial management unit.

[0135] In this embodiment, spatial overlay analysis, point-to-surface association, or spatial connection methods are used to match the spatial aggregation position of each target object with the boundary of the spatial management unit to determine its corresponding spatial management unit. This forms a set of mismatched objects at the spatial management unit level.

[0136] For target objects located at the intersection of multiple spatial management units, their affiliation can be determined according to the spatial management unit boundary inclusion relationship, nearest neighbor rule, area dominance rule, or preset priority rule. Target objects that cannot be matched to any spatial management unit are either removed or marked as unmatched.

[0137] Step 45: For each space management unit, count the number of objects of different mismatch types, the proportion of objects, the average mismatch intensity, the number of severely mismatched objects, the proportion of severely mismatched objects, and the average type confidence.

[0138] In this embodiment, for the j-th space management unit, the total number N of target objects within it is counted. j Number of working geothermal - low green type objects N W, j Number N of residential geothermal-low green type objects H, j Number N of severely working geothermal - low green type objects SW, j And the number N of severely residential geothermal-low green type objects SH, j .

[0139] Furthermore, the proportions of working geothermal-low green space objects, residential geothermal-low green space objects, severely working geothermal-low green space objects, and severely residential geothermal-low green space objects are calculated. These object proportions can be calculated by dividing the number of objects of the corresponding mismatch type by the total number N of target objects within the space management unit. j get.

[0140] Simultaneously, the average mismatch intensity of working geothermal-low green type objects, the average mismatch intensity of residential geothermal-low green type objects, the proportion of severe mismatches, and the average type confidence are calculated within each space management unit. The average type confidence can be obtained by averaging the type confidence at the target object level.

[0141] Step 46: Based on the spatial management unit-level mismatch statistics, calculate the workplace governance priority score and the residential governance priority score respectively.

[0142] In this embodiment, the governance priority score is not a general comprehensive evaluation score, but rather a score constructed based on the governance needs of urban heat and green environment mismatch in spatial management units. This score simultaneously incorporates the proportion of mismatched objects, average mismatch intensity, proportion of severe mismatches, and type confidence level to characterize the coverage, intensity level, severity, and identification reliability of mismatch risks, thereby converting the target object-level mismatch identification results into spatial governance priority results that can be used for screening and ranking on urban planning GIS platforms.

[0143] The priority score for site-specific governance is determined by comprehensively considering the proportion of geothermal-type working sites with low green coverage, the average mismatch intensity of geothermal-type working sites with low green coverage, the proportion of severely geothermal-type working sites with low green coverage, and the type confidence level. The priority score for residential site-specific governance is determined by comprehensively considering the proportion of geothermal-type residential sites with low green coverage, the average mismatch intensity of geothermal-type residential sites with low green coverage, the proportion of severely severely geothermal-type residential sites with low green coverage, and the type confidence level.

[0144] In one exemplary implementation, a weighted scoring method is used to calculate the workplace governance priority score. :

[0145] in, Indicates the first The proportion of working geothermal energy and low-green objects within each space management unit. Indicates the first Average mismatch intensity of working geothermal-low green objects within a spatial management unit. Indicates the first The proportion of severely active geothermal-related and low-green objects within each space management unit. Indicates the first Average type confidence of working geothermal-low green objects within a spatial management unit. Represents the normalized function. , , and These are weight parameters, and + + + = 1.

[0146] Accordingly, a weighted scoring method is used to calculate the residential governance priority score P_home,j:

[0147] in, Indicates the first The proportion of residential geothermal-low green space within each spatial management unit. Indicates the first Average mismatch intensity of residential geothermal-low green type objects within a spatial management unit Indicates the first The proportion of severely residential geothermal-affected, low-greenery-type properties within each spatial management unit. Indicates the first Average type confidence of residential geothermal-low green type objects within a spatial management unit. Represents the normalized function. , , and These are weight parameters, and + + + = 1.

[0148] In one exemplary implementation , , and These can be set to 0.30, 0.30, 0.25, and 0.15 respectively. , , and The values ​​can be set to 0.30, 0.30, 0.25, and 0.15 respectively. In other implementations, the weight parameters can also be adjusted based on urban thermal risk management objectives, expert knowledge, analytic hierarchy process (AHP), entropy weighting method, or user-defined parameters.

[0149] Step 47: Based on the workplace governance priority score and the residential site governance priority score, divide each space management unit into different governance priority levels.

[0150] In this embodiment, the space management unit can be divided into high-priority work site governance units, medium-priority work site governance units, low-priority work site governance units, high-priority residential site governance units, medium-priority residential site governance units, or low-priority residential site governance units based on a preset scoring threshold or quantile threshold.

[0151] In one exemplary implementation, the levels are determined according to the quantile of the governance priority score. For workplace governance priority scores... The top 25% of space management units are marked as high-priority units for work site governance; for Spatial management units falling between 25% and 50% are marked as medium-priority units for workplace governance; the remaining spatial management units are marked as low-priority units for workplace governance. Residential Governance Priority Scoring The same method can be used for classifying levels.

[0152] In other implementations, the priority level of governance can also be divided based on a fixed threshold, a threshold set by experts, or a threshold specified by the city management department.

[0153] Step 48: Perform integrity verification on the mismatch statistics and governance priority results at the space management unit level.

[0154] In this embodiment, for each space management unit, it is determined whether the total number of target objects within it reaches a preset minimum sample size threshold. When the number of target objects is lower than the preset minimum sample size threshold, a low sample size mark is added to the space management unit, or it is removed from the governance priority scoring results, in order to reduce the impact of insufficient sample size on the governance priority judgment.

[0155] In one exemplary implementation, a preset minimum sample size threshold is set to 30. When the total number of target objects within a spatial management unit is less than 30, a low sample size marker is added to that spatial management unit. Spatial management units with low sample size markers can still be displayed in the GIS layer, but their governance priority results are not used as the primary basis for decision-making.

[0156] Step 49: Output the space management unit-level governance priority results.

[0157] In this embodiment, the spatial management unit-level governance priority results include spatial management unit identifier, spatial management unit boundary, proportion of working geothermal-low green type objects, proportion of residential geothermal-low green type objects, average mismatch intensity, severe mismatch proportion, working site governance priority score, residential site governance priority score, governance priority level, and sample size label.

[0158] The spatial management unit-level governance priority results serve as input data for subsequent cooling and greening measures identification code matching, spatial mapping output, and multi-scale consistency verification.

[0159] Through the processing in step 4, this embodiment can convert the target object-level identification results of the mismatch between workplace / residential heat and green environment into spatial management unit-level governance priority results. This processing method can distinguish the spatial location of unfavorable workplace and residential environmental conditions, and comprehensively reflect the proportion of mismatched objects, mismatch intensity, degree of severe mismatch, and identification confidence, thereby providing a spatial basis for the subsequent configuration of cooling and greening measures.

[0160] Step 5: Based on the work site governance priority score, residential site governance priority score, mismatch type, average mismatch intensity, and the dominant dimension of thermal and green environment mismatch, automatically match the cooling and greening measure identification code, and write the cooling and greening measure identification code into the GIS layer attribute fields, tables, reports, interface data, or visualization interface corresponding to the spatial management unit, generating a spatial output result with mismatch type, mismatch intensity, governance priority, and measure identification code.

[0161] See Figure 1 Step 5 specifically includes the following steps.

[0162] Step 51: Read the space management unit-level governance priority results generated in step 4.

[0163] In this embodiment, the spatial management unit-level governance priority results include spatial management unit identifier, spatial management unit boundary, proportion of working geothermal-low green type objects, proportion of residential geothermal-low green type objects, average mismatch intensity, severe mismatch proportion, working site governance priority score, residential site governance priority score, governance priority level, and sample size label.

[0164] Among them, the workplace governance priority score is used to determine whether the space management unit needs to prioritize the configuration of workplace cooling and greening measures; the residential area governance priority score is used to determine whether the space management unit needs to prioritize the configuration of residential area cooling and greening measures; the average mismatch intensity and the severe mismatch ratio are used to determine the intensity of measure configuration; and the sample size label is used to indicate whether the identification results of the space management unit have sufficient sample support.

[0165] Step 52: Build or read the preset identification code library for cooling and greening measures.

[0166] In this embodiment, the cooling and greening measure identification code library includes a measure identification code, measure name, applicable scenario, corresponding mismatch type, corresponding mismatch intensity level, and corresponding dominant dimensions of thermal and green environment. The measure identification code is used to convert spatial identification results into measure codes that can be read by urban planning GIS platforms, planning auxiliary decision-making systems, or spatial governance databases.

[0167] In one exemplary embodiment, the cooling and greening measure identification code library includes workplace cooling and greening measure identification codes and residential cooling and greening measure identification codes.

[0168] The identification codes for workplace cooling and greening measures include: C01, Shading facilities; Applicable scenario: workplace; Corresponding mismatch type: workplace thermal-low green type; Corresponding mismatch intensity level: basic configuration level, medium configuration level, or enhanced configuration level; Corresponding thermal and green environment dominant dimension: thermal environment dominant type or thermal-green co-dominant type.

[0169] C02, factory area greening; applicable scenario: workplace; corresponding mismatch type: workplace thermal-low green type; corresponding mismatch intensity level: basic configuration level, medium configuration level, or enhanced configuration level; corresponding thermal and green environment dominant dimension: green environment dominant type or thermal-green co-dominant type.

[0170] C03, Cool Roof; Applicable Scenario: Work Site; Corresponding Mismatch Type: Work Site Thermal-Low Green Type; Corresponding Mismatch Intensity Level: Medium Configuration Level or Enhanced Configuration Level; Corresponding Thermal and Green Environment Dominant Dimension: Thermal Environment Dominant Type.

[0171] C04, high albedo paving; applicable scenario: work site; corresponding mismatch type: work site thermal-low green type; corresponding mismatch intensity level: medium configuration level or enhanced configuration level; corresponding thermal and green environment dominant dimension: thermal environment dominant type.

[0172] C05, permeable pavement; applicable scenario: work site; corresponding mismatch type: work site thermal-low green type; corresponding mismatch intensity level: medium configuration level or enhanced configuration level; corresponding thermal and green environment dominant dimension: green environment dominant type or thermal-green co-dominant type.

[0173] C06, ventilation corridor; applicable scenario: workplace; corresponding mismatch type: workplace thermal-low green type; corresponding mismatch intensity level: enhanced configuration level; corresponding thermal and green environment dominant dimension: thermal environment dominant type or thermal-green co-dominant type.

[0174] C07, a midday retreat space; applicable scenario: workplace; corresponding mismatch type: workplace thermal-low green type; corresponding mismatch intensity level: medium configuration level or enhanced configuration level; corresponding thermal and green environment dominant dimension: thermal environment dominant type.

[0175] The identification codes for residential area cooling and greening measures include: R01, Community Nighttime Cooling Facilities; Applicable Scenario: Residential Areas; Corresponding Mismatch Type: Residential Area Heat-Low Green Type; Corresponding Mismatch Intensity Level: Basic Configuration Level, Medium Configuration Level, or Enhanced Configuration Level; Corresponding Thermal and Green Environment Dominance Dimension: Thermal Environment Dominance Type or Thermal-Green Co-dominance Type.

[0176] R02, Neighborhood Green Space Supplement; Applicable Scenario: Residential Area; Corresponding Mismatch Type: Residential Area Hot-Low Green Type; Corresponding Mismatch Intensity Level: Basic Configuration Level, Medium Configuration Level, or Enhanced Configuration Level; Corresponding Thermal and Green Environment Dominance Dimension: Green Environment Dominance Type or Thermal-Green Co-dominance Type.

[0177] R03, Improved Park Accessibility; Applicable Scenario: Residential Areas; Corresponding Mismatch Type: Residential Area Hot-Low Green Type; Corresponding Mismatch Intensity Level: Medium Configuration Level or Enhanced Configuration Level; Corresponding Dominant Dimension of Hot and Green Environment: Green Environment Dominant Type.

[0178] R04, Pocket Park; Applicable scenario: Residential area; Corresponding mismatch type: Residential area hot-low green type; Corresponding mismatch intensity level: Medium configuration level or enhanced configuration level; Corresponding thermal and green environment dominant dimension: Green environment dominant type or thermal-green co-dominant type.

[0179] R05, Restorative Public Space; Applicable Scenario: Residential Area; Corresponding Mismatch Type: Residential Area Hot-Low Green Type; Corresponding Mismatch Intensity Level: Medium Configuration Level or Enhanced Configuration Level; Corresponding Thermal and Green Environment Dominant Dimension: Green Environment Dominant Type or Thermal-Green Co-dominant Type.

[0180] R06, pedestrian shade facilities; applicable scenario: residential area; corresponding mismatch type: residential area heat-low green type; corresponding mismatch intensity level: basic configuration level, medium configuration level or enhanced configuration level; corresponding thermal and green environment dominant dimension: thermal environment dominant type or thermal-green co-dominant type.

[0181] In other embodiments, the measure names, number of measures, and coding methods in the measure identification code library can be adjusted according to the planning standards, climate zone characteristics, land use types, management department needs, or expert knowledge of different cities.

[0182] Step 53: Based on the workplace governance priority score and the residential governance priority score, determine the governance direction of each spatial management unit.

[0183] In this embodiment, when the workplace governance priority score reaches a preset priority threshold, the space management unit is marked as a workplace governance target; when the residential governance priority score reaches a preset priority threshold, the space management unit is marked as a residential governance target; when both the workplace governance priority score and the residential governance priority score reach the preset priority threshold, the space management unit is marked as a dual-scenario governance target.

[0184] In one exemplary implementation, the preset priority threshold can be determined based on the quantile of the governance priority score. For example, spatial management units with a workplace governance priority score in the top 25% are marked as workplace governance targets, and spatial management units with a residential governance priority score in the top 25% are marked as residential governance targets. If the same spatial management unit meets both of the above conditions, it is marked as a dual-scenario governance target.

[0185] Step 54: Based on the average mismatch intensity and the severe mismatch ratio, determine the measure configuration intensity level for each space management unit.

[0186] In this embodiment, the measure configuration intensity levels include basic configuration level, medium configuration level, and enhanced configuration level. The measure configuration intensity level is used to determine the number of measure identification codes and the intensity of measure intervention that the space management unit needs to match.

[0187] In one exemplary implementation, when both the average mismatch intensity and the severe mismatch ratio of a space management unit are in a low range, it is marked as a basic configuration level; when either the average mismatch intensity or the severe mismatch ratio is in a medium range, it is marked as a medium configuration level; and when both the average mismatch intensity and the severe mismatch ratio are in a high range, it is marked as an enhanced configuration level. The low, medium, and high ranges can be determined based on quantile thresholds, fixed thresholds, or expert-set thresholds.

[0188] In a specific example, space management units with average mismatch intensity or severe mismatch ratio in the top 25% can be marked as enhanced configuration level, space management units in the 25% to 50% range can be marked as medium configuration level, and the remaining space management units that still belong to governance priority units can be marked as basic configuration level.

[0189] Step 55: Based on the relative contributions of thermal environment mismatch index and green environment mismatch index within the space management unit, determine the dominant dimension of thermal and green environment mismatch.

[0190] In this embodiment, for space management unit j, the average standardized thermal environment mismatch intensity and the average standardized green environment mismatch intensity of the target mismatch type objects within the space management unit are calculated respectively. For a working geothermal-low green type space management unit, the average standardized thermal environment mismatch intensity is the average standardized thermal environment mismatch value of the working geothermal-low green type target objects within the space management unit, and the average standardized green environment mismatch intensity is the average absolute value of the standardized green environment mismatch value of the working geothermal-low green type target objects within the space management unit.

[0191] For a residential geothermal-low green space management unit, the average standardized thermal environment mismatch intensity is the average absolute value of the standardized thermal environment mismatch value of the residential geothermal-low green target object within the space management unit, and the average standardized green environment mismatch intensity is the average standardized green environment mismatch value of the residential geothermal-low green target object within the space management unit.

[0192] Further, the ratio between the average intensity of thermal environment mismatch and the average intensity of green environment mismatch is calculated. When the ratio is greater than a first dominant threshold, the space management unit is marked as thermal environment dominant; when the ratio is less than a second dominant threshold, the space management unit is marked as green environment dominant; when the ratio is between the first and second dominant thresholds, the space management unit is marked as thermal-green co-dominant.

[0193] The first and second dominant thresholds are determined based on sample distribution characteristics, urban planning needs, or expert experience. In a specific embodiment, the first dominant threshold is set to 1.2, and the second dominant threshold is set to 0.8. When the ratio of the mean intensity of thermal environment mismatch to the mean intensity of green environment mismatch is greater than 1.2, it indicates that the unfavorable thermal environment of the space management unit is more prominent; when the ratio is less than 0.8, it indicates that the insufficient green environment of the space management unit is more prominent; when the ratio is between 0.8 and 1.2, it indicates that unfavorable thermal environment and insufficient green environment together constitute the main mismatch characteristics.

[0194] The above calculation method is used to uniformly convert the adverse thermal environment and the adverse green environment under different mismatch formulas into positive values ​​for comparison, so that the mean intensity of thermal environment mismatch and the mean intensity of green environment mismatch are comparable.

[0195] Step 56: Based on the governance direction, the intensity level of the measures configuration, and the dominant dimension of the mismatch between heat and green environment, automatically match one or more cooling and greening measure identification codes from the cooling and greening measure identification code library.

[0196] In this embodiment, when a space management unit is marked as a work site governance target, a measure identification code matching its measure configuration intensity level and dominant dimension is selected from the work site cooling and greening measure identification codes. When a space management unit is marked as a residential site governance target, a measure identification code matching its measure configuration intensity level and dominant dimension is selected from the residential site cooling and greening measure identification codes. When a space management unit is marked as a dual-scenario governance target, both the work site cooling and greening measure identification codes and the residential site cooling and greening measure identification codes are matched simultaneously.

[0197] "Priority matching" refers to including the measure identification codes corresponding to the dominant dimension in the preferred recommendation list, provided that the measure configuration intensity level constraints are met. In the output results, priority-matched measure identification codes can be marked as dominant measures, while other measure identification codes that meet the governance orientation and configuration intensity constraints but do not belong to the dominant dimension can be marked as auxiliary measures.

[0198] In one exemplary implementation, for high-priority work site governance units dominated by thermal environment, C01, C03, C04, C06, and C07 are preferentially matched; for high-priority work site governance units dominated by green environment, C02 and C05 are preferentially matched; for high-priority work site governance units dominated by both thermal and green environment, multiple identification codes among C01, C02, C03, C04, C05, C06, and C07 can be matched simultaneously.

[0199] For high-priority residential governance units dominated by thermal environment, R01 and R06 are matched first; for high-priority residential governance units dominated by green environment, R02, R03, R04 and R05 are matched first; for high-priority residential governance units dominated by both thermal and green environment, multiple identification codes from R01, R02, R03, R04, R05 and R06 can be matched simultaneously.

[0200] Step 57: Perform integrity and consistency checks on the automatically matched identification codes for cooling and greening measures.

[0201] In this embodiment, the integrity check is used to determine whether each governance priority spatial management unit has been matched with at least one measure identification code. The consistency check is used to determine whether the matched measure identification code is consistent with the governance direction, final mismatch type, measure configuration intensity level, and dominant dimension of thermal and green environment mismatch of the spatial management unit.

[0202] When a measure identification code is inconsistent with the governance orientation of a space management unit, the measure identification code is removed. When a space management unit fails to match any measure identification code, a mark for manual review is added, or a default measure identification code is matched based on its governance orientation and dominant dimension.

[0203] Step 58: Write the identification code of the cooling and greening measures into the attribute field corresponding to the space management unit to generate the space output result.

[0204] In this embodiment, the governance orientation, mismatch type, average mismatch intensity, governance priority level, measure configuration intensity level, dominant dimension of thermal and green environment mismatch, and cooling and greening measure identification code corresponding to each spatial management unit are written into the GIS layer attribute fields.

[0205] The attribute fields may include a spatial management unit identifier field, a workplace governance priority field, a residential area governance priority field, a governance direction field, a measure configuration intensity level field, a dominant dimension field, a measure identification code field, a sample size marker field, and a manual review marker field. If the same spatial management unit corresponds to multiple measure identification codes, they can be recorded using a delimiter concatenation, multi-field storage, or associated sub-tables.

[0206] Step 59: Output the spatial output results in the form of GIS layers, attribute tables, statistical tables, analysis reports, interface data, or visualization interfaces.

[0207] In this embodiment, the spatial output results can be saved as vector layers, raster layers, database tables, GeoJSON files, Shapefile files, geodatabase feature classes, WebGIS interface data, or other data formats readable by GIS platforms. The spatial output results are used to support spatial queries, layer overlays, risk screening, and the layout of cooling and greening measures in urban planning GIS platforms.

[0208] Through the processing in step 5, this embodiment can further convert the identification results of the mismatch between urban living and working environment and green environment from risk identification results into measure coding results that can be read and called by the planning system. This processing method enables the spatial management unit to not only have mismatch type and governance priority attributes, but also corresponding cooling and greening measure identification codes, thereby improving the application and transformation capabilities of this invention in urban thermal environment governance, green space configuration, and planning auxiliary decision-making.

[0209] Step 6: Perform consistency verification on mismatch type and governance priority results at multiple spatial neighborhood scales, and convert the consistency verification results into confidence fields or scale sensitivity labels to generate the final spatial output results with stability information.

[0210] Figure 6 A schematic diagram illustrating multi-scale consistency verification and final spatial output is shown. See also... Figure 1 and Figure 6 Step 6 specifically includes the following steps.

[0211] Step 61: Set multiple spatial neighborhood scales.

[0212] In this embodiment, the spatial neighborhood scale includes at least two circular buffer zones of different scales, regular grids, street blocks, or service areas generated based on road network accessibility. Multiple spatial neighborhood scales are used to characterize the thermal and green environment characteristics around residential and workplace locations within different spatial ranges.

[0213] In one exemplary embodiment, five spatial neighborhood scales are defined: 100 m, 300 m, 500 m, 1000 m, and 1500 m. The 100 m and 300 m scales are used to characterize micro-scale environmental features within the vicinity of a residence or workplace; the 500 m scale is used to characterize neighborhood environmental features within a walking distance; and the 1000 m and 1500 m scales are used to characterize surrounding environmental features over a larger area. In other embodiments, other scale combinations can be selected based on urban spatial morphology, walking accessibility, remote sensing image spatial resolution, or user-defined parameters.

[0214] Step 62: Repeat the process of identifying and spatially aggregating the mismatch between job and residence heat and green environment at the multiple spatial neighborhood scales.

[0215] In this embodiment, at each spatial neighborhood scale, the environmental feature extraction, thermal environment mismatch index generation, green environment mismatch index generation, standardized feature vector construction, mismatch type identification, mismatch intensity calculation, and spatial management unit aggregation processes in steps 2 to 4 are repeated to obtain target object-level mismatch type results and spatial management unit-level governance priority results at different spatial neighborhood scales.

[0216] Furthermore, at each spatial neighborhood scale, the matching process for cooling and greening measure identification codes in step 5 can be repeated to obtain measure identification code results at different spatial neighborhood scales. These measure identification code results at different scales can be used to verify the stability of the measure matching results at the main analysis scale.

[0217] Step 63: Determine the main analysis scale and generate the main layer results.

[0218] In this embodiment, one spatial neighborhood scale is selected as the primary analysis scale from multiple spatial neighborhood scales. The primary analysis scale can be determined based on urban planning and management needs, neighborhood walking range, data resolution, existing planning standards, or user-defined parameters.

[0219] In one exemplary implementation, a spatial neighborhood scale of 500 m is selected as the primary analysis scale. The mismatch type results, governance priority results, and cooling and greening measure identification code results generated at the 500 m scale are used as the main layer results. Results at other spatial neighborhood scales are used as auxiliary verification results to evaluate the scale stability of the main layer results.

[0220] Step 64: Compare the mismatch type results at other spatial neighborhood scales with the main layer results for consistency.

[0221] In this embodiment, for the same anonymized target object, the final mismatch type at other spatial neighborhood scales is compared with the final mismatch type at the main analysis scale. If the target object is identified as having the same mismatch type at multiple scales as at the main analysis scale, it indicates that its mismatch type identification result is not sensitive to changes in spatial neighborhood scale. If the target object is identified as having a different mismatch type at other spatial neighborhood scales than at the main analysis scale, it indicates that its mismatch type identification result is significantly affected by changes in spatial neighborhood scale.

[0222] For spatial management units, it is also possible to compare whether the dominant mismatch type at different spatial neighborhood scales is consistent with the dominant mismatch type at the main analysis scale. The dominant mismatch type can be the mismatch type with the highest proportion of objects within the spatial management unit, or it can be the main mismatch type corresponding to the governance priority score.

[0223] Step 65: Based on the mismatch type consistency comparison results, generate the mismatch type confidence level.

[0224] In this embodiment, when the same anonymized target object or the same spatial management unit is identified as the same mismatch type at two or more spatial neighborhood scales, the confidence level corresponding to the mismatch type is increased; when the same anonymized target object or the same spatial management unit is identified as the same mismatch type at multiple spatial neighborhood scales, the mismatch type is marked as a high-stability mismatch type.

[0225] In one exemplary implementation, if the total number of spatial neighborhood scales is set to M, and the number of scales with the same mismatch type as the main analysis scale for a certain target object or spatial management unit is set to m, then the mismatch type consistency ratio CT can be expressed as:

[0226] CT is used to characterize the consistency of mismatch type results across different spatial neighborhood scales. When CT is greater than or equal to 0.80, the mismatch type confidence level is marked as high; when CT is greater than or equal to 0.50 and less than 0.80, the mismatch type confidence level is marked as medium; and when CT is less than 0.50, the mismatch type confidence level is marked as low. In other embodiments, the confidence level threshold can also be adjusted according to the sample size, number of scales, or user-defined parameters.

[0227] Step 66: When the mismatch type results are inconsistent at different spatial neighborhood scales, generate a mismatch type scale sensitivity label.

[0228] In this embodiment, when the mismatch type of the same anonymized target object or the same spatial management unit changes at different spatial neighborhood scales, a mismatch type scale sensitivity marker is generated. This mismatch type scale sensitivity marker indicates that the mismatch type identification result of the target object or the spatial management unit is significantly affected by changes in the spatial neighborhood scale.

[0229] In one exemplary implementation, when the mismatch type confidence level is low, or when the mismatch type in the main analysis scale only occurs at a spatial neighborhood scale, the target object or the spatial management unit is marked as mismatch type scale sensitive. When the mismatch type confidence level is medium, it can be marked as medium scale sensitive. When the mismatch type confidence level is high, it can be marked as low scale sensitive or scale stable.

[0230] Step 67: Compare the governance priority results at different spatial neighborhood scales for consistency, and generate governance priority confidence level and governance priority scale sensitivity label.

[0231] In this embodiment, for the same spatial management unit, the governance priority level of its work location and its governance priority level of its residence are compared at different spatial neighborhood scales to see if they are consistent. If the same spatial management unit is identified as having the same governance priority level at two or more spatial neighborhood scales, the confidence level of the governance priority result of that spatial management unit is increased. If the governance priority level of the spatial management unit changes at different spatial neighborhood scales, a governance priority scale sensitivity marker is generated.

[0232] In one exemplary implementation, the governance priority level of the work site and the governance priority level of the residence of each spatial management unit are calculated at scales of 100 m, 300 m, 500 m, 1000 m, and 1500 m, respectively. When the same spatial management unit is identified as a high-priority unit at three or more scales, its governance priority confidence level is marked as high; when the same spatial management unit is identified as a high-priority unit at only one scale, it is marked as governance priority scale sensitive; when the same spatial management unit changes between high, medium, and low priorities at different scales, its governance priority scale sensitivity is marked as yes, and the scale range of the change is recorded in the attribute field.

[0233] Step 68: Write the mismatch type confidence level, governance priority confidence level, mismatch type scale sensitivity flag, and governance priority scale sensitivity flag into the attribute fields of the spatial output results.

[0234] In this embodiment, the stability information generated in steps 65 to 67 is correlated with the spatial output result generated in step 5. The spatial output result, in addition to the existing fields such as mismatch type, mismatch intensity, governance priority level, and cooling and greening measure identification code, further includes mismatch type confidence level, governance priority confidence level, mismatch type scale sensitivity marker, and governance priority scale sensitivity marker.

[0235] In one exemplary implementation, if a spatial management unit is identified as a high-priority unit for workplace governance at the main analysis scale and is also identified as a high-priority unit for workplace governance at multiple other spatial neighborhood scales, then the workplace governance priority confidence level of the spatial management unit is marked as high. If the spatial management unit is identified as a high-priority unit for workplace governance only at the main analysis scale, but is identified as a medium-priority or low-priority unit at other scales, then the spatial management unit is marked with a governance priority scale sensitivity tag.

[0236] Step 69: Output the final spatial output result with stability information.

[0237] In this embodiment, the final spatial output results include spatial management unit identifier, spatial management unit boundary, mismatch type, mismatch intensity, governance priority level, cooling and greening measure identification code, mismatch type confidence level, governance priority confidence level, mismatch type scale sensitivity label, governance priority scale sensitivity label, and sample size label.

[0238] The final spatial output results can be output in the form of GIS layers, attribute tables, statistical tables, analysis reports, interface data, or visualization interfaces, to support spatial queries, layer overlays, risk screening, and the layout of cooling and greening measures in urban planning GIS platforms.

[0239] Through the processing in step 6, this embodiment can transform the identification consistency under different spatial neighborhood scales into confidence level and scale sensitivity markers that can be read by the GIS platform, thereby reducing the impact of a single spatial scale selection on the identification results and improving the stability, interpretability, and planning reliability of the identification results of the mismatch between job and housing heat and green environment and the judgment results of governance priority.

[0240] Figure 7 This is a schematic diagram of the urban thermal environment and green environment mismatch identification system provided in an embodiment of the present invention, as shown below. Figure 7 As shown, it includes: an acquisition module 71, a construction module 72, an identification module 73, a generation module 74, a matching module 75, and a verification module 76, wherein: The acquisition module 71 is used to acquire anonymized occupational and residential coordinate data and remote sensing image data and perform unified processing to generate anonymized occupational and residential spatial datasets, surface temperature raster data, and vegetation index raster data; the construction module 72 is used to construct spatial neighborhoods and extract environmental features based on the anonymized occupational and residential spatial datasets and raster data, calculate occupational and residential environment differences, generate thermal environment mismatch indicators and green environment mismatch indicators, and construct standardized mismatch feature vectors; the identification module 73 is used to perform a mismatch type identification process combining a threshold rule model and a clustering model based on the standardized mismatch feature vectors, and generate target object-level mismatch type and mismatch intensity results after consistency verification; the generation module... Block 74 is used to determine the spatial aggregation location based on the mismatch type, aggregate the target object-level results to the spatial management unit, and generate differentiated workplace governance priority scores and residential governance priority scores by combining the mismatch object ratio and average mismatch intensity of each unit; Matching module 75 is used to automatically match the cooling and greening measure identification code based on the governance priority score, mismatch type and the dominant dimension of thermal and green environment mismatch, and generate spatial output results with governance priority and measure identification code; Verification module 76 is used to repeat the above process at multiple spatial neighborhood scales, convert the multi-scale identification consistency into confidence level and scale sensitivity label, and superimpose it on the spatial output results.

[0241] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0242] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0243] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying mismatches between urban thermal environment and green environment, characterized in that, include: Anonymized occupational and residential coordinate data and remote sensing image data are acquired and processed in a unified manner to generate anonymized occupational and residential spatial datasets, surface temperature raster data and vegetation index raster data; Based on the anonymized work-residence space dataset and raster data, a spatial neighborhood is constructed and environmental features are extracted. The work-residence environment difference is calculated, and thermal environment mismatch index and green environment mismatch index are generated. A standardized mismatch feature vector is constructed. Based on the standardized mismatch feature vector, a mismatch type identification process combining a threshold rule model and a clustering model is executed, and after consistency verification, the mismatch type and mismatch intensity results at the target object level are generated. Based on the mismatch type, determine the spatial aggregation location, aggregate the target object level results to the spatial management unit, and generate differentiated workplace governance priority scores and residential governance priority scores by combining the mismatch object ratio and average mismatch intensity of each unit. Based on the governance priority score, mismatch type, and dominant dimensions of thermal and green environment mismatch, the system automatically matches the identification code of cooling and greening measures, and generates spatial output results with governance priority and measure identification code. The above process is repeated at multiple spatial neighborhood scales to convert multi-scale identification consistency into confidence level and scale sensitivity label, which are then superimposed on the spatial output result.

2. The method for identifying mismatches between urban thermal environment and green environment according to claim 1, characterized in that, The process of acquiring anonymized occupational and residential coordinate data and remote sensing image data and performing unified processing includes: Perform field integrity checks and spatial reference transformations on the work-residence coordinate data, remove abnormal coordinate records, and generate an anonymized work-residence spatial dataset; Cloud masking, radiometric correction, and temporal synthesis are performed on remote sensing image data to generate surface temperature raster data and vegetation index raster data, and the anonymized occupational and residential spatial dataset and the raster data are made to have a unified spatial reference.

3. The method for identifying mismatches between urban thermal environment and green environment according to claim 1, characterized in that, The construction of the standardized mismatch feature vector includes: Surface temperature and vegetation index values ​​are extracted from the spatial neighborhoods of residential and workplace locations, respectively. The surface temperature difference and vegetation index difference between workplace and residential locations are calculated to generate thermal environment mismatch index and green environment mismatch index. The mismatch index is then standardized to eliminate the dimensional differences between surface temperature and vegetation index.

4. The method for identifying mismatches between urban thermal environment and green environment according to claim 1, characterized in that, The threshold rule model is used to determine the mismatch direction of each target object based on the preset hot mismatch direction threshold and green mismatch direction threshold, and to divide the mismatch intensity level based on the preset intensity threshold to generate the baseline mismatch type. The clustering model adopts a priori-guided K-means clustering model, in which the prior cluster centers are set according to the directional and intensity features of mismatches, so as to cluster and identify the standardized mismatch feature vectors.

5. The method for identifying mismatches between urban thermal environment and green environment according to claim 1, characterized in that, The consistency check includes: When the mismatch type output by the threshold rule model and the clustering model is consistent, the mismatch type is taken as the final mismatch type and marked as high confidence. When the two are inconsistent, the final mismatch type is determined according to the distance determination rule or the confidence determination rule.

6. The method for identifying mismatches between urban thermal environment and green environment according to claim 1, characterized in that, The step of determining the spatial aggregation location based on the mismatch type includes: For working geothermal-low green type target objects, their working location coordinates are used as the spatial aggregation location; For residential hot-low green type target objects, their residential coordinates are used as spatial aggregation locations, so that unfavorable environmental conditions of the workplace are aggregated to the unit where the workplace is located, and unfavorable environmental conditions of the residential area are aggregated to the unit where the residential area is located.

7. The method for identifying mismatches between urban thermal environment and green environment according to claim 1, characterized in that, The automatically matched cooling and greening measure identification code includes: A pre-defined database of identification codes for cooling and greening measures is provided, where cooling and greening measures for workplaces include at least shading facilities and factory greening, and cooling and greening measures for residential areas include at least community nighttime cooling facilities and supplementary neighborhood green spaces. The governance priority score is determined by a comprehensive calculation of the proportion of mismatched objects, average mismatch intensity, proportion of severe mismatches, and type confidence within the corresponding spatial management unit.

8. The method for identifying mismatches between urban thermal environment and green environment according to claim 1, characterized in that, The multi-scale consistency check includes: The environmental feature extraction, mismatch index calculation, mismatch type identification and spatial aggregation processes are repeated at at least two different spatial neighborhood scales, and the identification results at each scale are compared with the results at the main analysis scale. When the same object or unit is identified as the same mismatch type at two or more scales, the confidence level of that mismatch type is increased; when the identification results are inconsistent, a scale-sensitive label is generated.

9. A system for identifying mismatches between urban thermal environment and green environment, characterized in that, include: The acquisition module is used to acquire anonymized occupational and residential coordinate data and remote sensing image data and process them in a unified manner to generate anonymized occupational and residential spatial datasets, surface temperature raster data and vegetation index raster data. The construction module is used to construct a spatial neighborhood and extract environmental features based on the anonymized work-residence space dataset and raster data, calculate the work-residence environment difference, generate thermal environment mismatch index and green environment mismatch index, and construct a standardized mismatch feature vector. The identification module is used to perform a mismatch type identification process that combines a threshold rule model and a clustering model based on the standardized mismatch feature vector, and generate the target object level mismatch type and mismatch intensity results after consistency verification. The generation module is used to determine the spatial aggregation location based on the mismatch type, aggregate the target object-level results to the spatial management unit, and generate differentiated workplace governance priority scores and residential governance priority scores by combining the proportion of mismatched objects and the average mismatch intensity of each unit. The matching module is used to automatically match the identification code of cooling and greening measures based on the governance priority score, mismatch type and dominant dimension of thermal and green environment mismatch, and generate spatial output results with governance priority and measure identification code; The verification module is used to repeatedly execute the above process at multiple spatial neighborhood scales, converting multi-scale recognition consistency into confidence level and scale sensitivity label, which are then superimposed on the spatial output result.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the urban thermal environment and green environment mismatch identification method as described in any one of claims 1 to 8.

Images