Intelligent monitoring and data evaluation method for ecological benefits of urban green space

Abstract

The application discloses a method for intelligent monitoring and data evaluation of urban green land ecological benefits, and belongs to the technical field of ecological monitoring and data analysis. The method comprises the following steps: acquiring multi-source remote sensing image data of a target urban area, analyzing a vegetation coverage distribution map therefrom, and dividing the target urban area into multiple ecological evaluation units; extracting a vegetation index time sequence and a ground temperature time sequence for each ecological evaluation unit, and calculating an ecological service function equivalent cumulative value; collecting meteorological monitoring data to perform spatio-temporal dynamic correction on the ecological service function equivalent cumulative value, and generating a corrected cumulative equivalent value; constructing an ecological benefit spatial distribution matrix based on the corrected cumulative equivalent value and a land use type map; and determining an ecological benefit comprehensive index of the target urban area according to the distribution matrix. The method realizes fine and dynamic evaluation of the ecological benefits of urban green land by coupling the vegetation and temperature time sequences, dynamically correcting the meteorological data, and constructing the distribution matrix by using the land use and spatial autocorrelation information.

Description

Technical Field

[0001] This invention relates to the field of ecological protection, specifically to a method for intelligent monitoring and data evaluation of the ecological benefits of urban green spaces. Background Technology

[0002] In the field of ecological and environmental protection, the monitoring and assessment of the ecological benefits of urban green spaces typically rely on extracting single indicators such as vegetation indices or vegetation cover from remote sensing images, and obtaining regional ecological function evaluation values ​​through simple statistical analysis. This approach treats ecological benefits as a static reflection of vegetation status, failing to model the coupling relationship between vegetation and surface temperature and the cumulative effect of this relationship over time. Consequently, the assessment results cannot reflect the dynamic changes in ecosystem service functions at different times. Furthermore, the real-time disturbances of meteorological factors to the ecological benefits of green spaces are ignored, and rainfall and temperature fluctuations are directly removed from the assessment process. This results in the assessment output lacking a response to climate variability in both time scale and spatial location, making it difficult to reflect the fluctuations in ecosystem service functions under real environmental conditions.

[0003] Furthermore, the spatial allocation of ecological benefits often adopts a uniform or simple area ratio approach, failing to consider the heterogeneity of land use types within the city and the spatial dependence of ecological processes. This results in assessment results that deviate from the actual diffusion patterns of ecological functions in terms of spatial distribution, and cannot accurately reveal the differentiated contributions of different land parcels to the overall ecological benefits. Urban green space planning and management require an assessment method that can integrate the temporal response characteristics of vegetation growing seasons, the impact of meteorological dynamic fluctuations, and the spatial pattern of land use. This is to address the shortcomings of existing technologies in characterizing the cumulative effects over time, lacking correction for meteorological fluctuations, and lacking ecological rationality in spatial allocation. Summary of the Invention

[0004] This paper proposes an intelligent monitoring and data evaluation method for the ecological benefits of urban green spaces. By establishing a time-series cumulative equivalent conversion mechanism that couples vegetation index with surface temperature, and combining it with dynamic correction driven by rainfall and temperature anomalies, and introducing land use type information and spatial autocorrelation iterative allocation strategy, the paper achieves dynamic evaluation of the ecological benefits of urban green spaces that can reflect the impact of climate fluctuations and spatial heterogeneity patterns.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for intelligent monitoring and data evaluation of the ecological benefits of urban green spaces is disclosed. This method achieves spatiotemporal dynamic evaluation and comprehensive index quantification of the ecological service functions of urban green spaces through the fusion processing of multi-source remote sensing image data and meteorological monitoring data. The implementation steps include: acquiring multi-source remote sensing image data of the target urban area; parsing a vegetation cover distribution map from the multi-source remote sensing image data; dividing the vegetation cover distribution map into multiple ecological assessment units; for each ecological assessment unit, extracting its corresponding vegetation index time series and land surface temperature time series; calculating the cumulative equivalent value of the ecological service function for each ecological assessment unit based on the vegetation index time series and the land surface temperature time series; collecting meteorological monitoring data corresponding to each ecological assessment unit; performing spatiotemporal dynamic correction on the cumulative equivalent value of the ecological service function based on the meteorological monitoring data; generating a corrected cumulative equivalent value for each ecological assessment unit; constructing an ecological benefit spatial allocation matrix based on the corrected cumulative equivalent value of each ecological assessment unit and a land use type map of the target urban area; and determining the comprehensive ecological benefit index of the target urban area based on the ecological benefit spatial allocation matrix.

[0006] As a technical solution of this invention, when acquiring multi-source remote sensing image data and dividing it into ecological assessment units, the multi-source remote sensing image data is radiometrically calibrated and atmospherically corrected to obtain corrected image data. Vegetation cover areas are extracted from the corrected image data, and vegetation cover is calculated to obtain the vegetation cover distribution map. Based on the spatial variation characteristics of vegetation cover in the vegetation cover distribution map, a spatial grid segmentation method is used to divide the target urban area into multiple ecological assessment units of equal area. This processing method can eliminate the influence of atmospheric conditions on image quality, making the subsequently extracted vegetation and temperature information more consistent, and providing a standardized basis for spatial analysis using a unified grid.

[0007] Preferably, when calculating the cumulative equivalent value of ecosystem service functions, the normalized vegetation index (NVI) of each ecological assessment unit in the corrected image data is extracted based on the corresponding pixel position of that unit in multiple historical time phases to form the vegetation index time series; simultaneously, the thermal infrared radiance value of that unit in multiple historical time phases is extracted to obtain the land surface temperature time series; for each ecological assessment unit, the NVI and land surface temperature in the same time phase are substituted into the equivalent transformation function to obtain the unit area equivalent value for that time phase; finally, the unit area equivalent value of each ecological assessment unit in all time phases is integrated over time to obtain the cumulative equivalent value of ecosystem service functions. Through the joint time series analysis of vegetation index and land surface temperature, the linkage between vegetation growth and thermal environmental effects can be captured, thereby more accurately characterizing the comprehensive ecosystem service capacity of green spaces in areas such as cooling and carbon sequestration.

[0008] Furthermore, in the specific design of the equivalent conversion function, the difference between the normalized vegetation index and the land surface temperature at that time phase is obtained. This difference is divided by the sum of the normalized vegetation index and the land surface temperature to obtain the vegetation temperature structure coefficient. The product of the normalized vegetation index and the land surface temperature is obtained, and the product is multiplied by the vegetation temperature structure coefficient to obtain the equivalent value per unit area at that time phase. This conversion method can effectively highlight the nonlinear coupling relationship between vegetation cover and land surface temperature, making the equivalent value per unit area more sensitive to the ecological benefits of green space.

[0009] To improve upon the aforementioned technical solution, in the spatiotemporal dynamic correction stage, for each ecological assessment unit, time-series rainfall and temperature data from the meteorological monitoring station at the unit's location are acquired as the corresponding meteorological monitoring data for that unit. Based on the rainfall and temperature time-series data, the rainfall anomaly and temperature anomaly for that unit are calculated respectively. A correction coefficient for that unit is determined based on the rainfall and temperature anomalies. The cumulative equivalent value of the ecosystem service function is multiplied by the correction coefficient to obtain the corrected cumulative equivalent value. By introducing anomaly information from rainfall and temperature, the estimation bias of ecosystem service functions under different climatic conditions can be corrected, making the assessment results more reflective of the actual impact of climate fluctuations.

[0010] When determining the correction coefficient, the rainfall anomaly is multiplied by a first weighting factor to obtain the rainfall component, and the temperature anomaly is multiplied by a second weighting factor to obtain the temperature component. The rainfall component and the temperature component are added together to obtain the climate fluctuation factor. The preset benchmark coefficient is added to the climate fluctuation factor to obtain the correction coefficient. This method quantifies the synergistic impact of hydrothermal conditions on the ecological benefits of green spaces using a weighted synthesis approach, and the weighting factors can be flexibly set to adapt to the assessment needs of different climate zones.

[0011] When constructing the spatial allocation matrix of ecological benefits, the land use type map of the target urban area is obtained, and the land use category code corresponding to each ecological assessment unit is extracted from the land use type map. Based on the corrected cumulative equivalent value and the land use category code of each ecological assessment unit, a feature vector for that unit is generated. The feature vectors of all ecological assessment units are arranged in spatial order to form an initial ecological benefit matrix. Spatial autocorrelation iteration is performed on the initial ecological benefit matrix, and the spatial allocation matrix of ecological benefits is obtained after convergence. This step integrates the point-based assessment results with land use type information and uses spatial autocorrelation to smooth and optimize the allocation pattern, making the spatial distribution of ecological benefits more continuous and consistent with geographical laws.

[0012] In the specific implementation of the above spatial autocorrelation iterative operation, a neighborhood-weighted average is performed on the initial ecological benefit matrix using a spatial weight matrix to obtain a first iterative matrix; the difference between the first iterative matrix and the initial ecological benefit matrix is ​​calculated to obtain an iterative residual matrix; the iterative residual matrix is ​​multiplied by a relaxation factor and added to the initial ecological benefit matrix to obtain a second iterative matrix; the second iterative matrix is ​​used as a new initial ecological benefit matrix, and the above iterative steps are repeated until the norm of the iterative residual matrix is ​​less than a preset threshold, thus converging to obtain the ecological benefit spatial allocation matrix. This iterative process can effectively eliminate random noise, strengthen local spatial clustering characteristics, and improve the robustness and interpretability of the allocation matrix.

[0013] When determining the comprehensive ecological benefit index, each element in the ecological benefit spatial allocation matrix is ​​normalized to obtain a normalized allocation value. Based on the normalized allocation value of each ecological assessment unit, the ecological benefit contribution weight of that unit is determined. Specifically, the normalized allocation value of each ecological assessment unit is divided by the sum of the normalized allocation values ​​of all ecological assessment units to obtain the ecological benefit contribution weight of that unit. The ecological benefit contribution weight of each ecological assessment unit is then weighted and summed with its corresponding corrected cumulative equivalent value to obtain the comprehensive ecological benefit index. This comprehensive index integrates spatial allocation information and corrected cumulative equivalent, enabling a holistic quantification of the ecological benefit level of the regional green space system and providing an intuitive decision-making basis for urban planning and management.

[0014] This invention utilizes the joint processing of multi-source remote sensing and meteorological data to construct an assessment framework from multiple dimensions, including vegetation cover, thermal environment effects, and climate fluctuations. This transforms the evaluation of urban green space ecological benefits from a single indicator to a comprehensive spatiotemporal dynamic assessment. By dividing the assessment into gridded units and correcting the spatiotemporal equivalent cumulative values, the comparability of assessment results across different locations and periods is ensured. The introduction of vegetation temperature structure coefficients and climate fluctuation factors enhances the accuracy of quantifying ecosystem service functions. Furthermore, land use type coding and spatial autocorrelation iterative calculations optimize the spatial allocation of ecological benefits, ensuring that the final comprehensive ecological benefit index accurately reflects the overall ecosystem service level of the urban green space system, supporting precise ecological space planning and performance evaluation.

[0015] The technical effects and advantages provided by the present invention in the above technical solution are as follows: Based on the time series of vegetation indices and land surface temperature, normalized vegetation indices (NVI) and land surface temperatures retrieved from thermal infrared bands for multiple historical time phases are extracted for each ecological assessment unit. A vegetation-temperature structure coefficient is generated by dividing the difference between the NVI and land surface temperature at the same time phase by their sum. The equivalent value per unit area is then obtained by multiplying the product of the NVI and land surface temperature by this structure coefficient. Finally, the equivalent value per unit area for all time phases is integrated over time to obtain the cumulative equivalent value of ecosystem service function. This process expands the vegetation status of a single time phase to a cumulative measure of the coupling between vegetation and the thermal environment over time. This allows for comparison of the ecological benefits of green spaces under different growing seasons or meteorological backgrounds on the same cumulative dimension, avoiding the instantaneous and one-sided nature of using only a single vegetation indicator. Using time-series rainfall and temperature data collected from meteorological monitoring stations, rainfall and temperature anomalies are calculated respectively. A first and second weighting factor are introduced to convert these anomalies into rainfall and temperature components, which are then added together to obtain a climate fluctuation factor. This factor is then summed with a preset benchmark coefficient to form a correction coefficient, which corrects the equivalent cumulative value. The correction process directly embeds the degree of deviation of climate elements from normal conditions, enabling the assessment results of the same ecological assessment unit under different precipitation and temperature conditions to adaptively reflect the increase or decrease in ecological benefits caused by meteorological fluctuations. This overcomes the information loss caused by neglecting interannual and seasonal climate variability in traditional assessments.

[0016] Based on the corrected cumulative equivalent value, and combined with the land use type map of the target urban area, land use category codes are extracted for each ecological assessment unit. These codes, together with the corrected cumulative equivalent value, constitute a feature vector. The feature vectors of all units are arranged in spatial order to form an initial ecological benefit matrix. A neighborhood weighted average is then performed using the spatial weight matrix to obtain the first iteration matrix. The iteration residual matrix is ​​calculated and multiplied by a relaxation factor, then added to the initial matrix to obtain the second iteration matrix. This process is repeated until the norm of the iteration residual matrix is ​​less than a preset threshold, resulting in convergence and the ecological benefit spatial allocation matrix. In the construction of this allocation matrix, the land use category code endows the matrix elements with differentiated attributes constrained by human activities and ecological function types. Spatial autocorrelation iteration makes the ecological benefit allocation among neighboring units mutually constrained and tends to be smooth and reasonable. The final allocation matrix can present the distribution characteristics of urban ecological benefits jointly shaped by land use and spatial proximity. The ecological benefit contribution weight of each unit is determined based on the normalized allocation value, and the ecological benefit comprehensive index is obtained by weighted summation with the corrected cumulative equivalent value. This index integrates information from three dimensions: time accumulation, climate correction and spatial allocation, and fully expresses the spatiotemporal heterogeneity of urban green space ecological benefits. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0018] Figure 1 This is a flowchart of a method for intelligent monitoring and data evaluation of the ecological benefits of urban green spaces; Figure 2 This is a flowchart for calculating the equivalent value per unit area; Figure 3 This is a flowchart of the calculation process for the comprehensive ecological benefit index; Figure 4 This is a schematic diagram of the cumulative frequency curve and threshold of the normalized vegetation index; Figure 5 This is a bar chart showing the distribution of ecological benefit contribution weights. Detailed Implementation

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

[0020] See Figure 1 This invention provides a method for intelligent monitoring and data evaluation of the ecological benefits of urban green spaces, comprising: acquiring multi-source remote sensing image data of a target urban area; parsing a vegetation cover distribution map from the multi-source remote sensing image data; and dividing the area into multiple ecological assessment units based on the vegetation cover distribution map; for each ecological assessment unit, extracting its corresponding vegetation index time series and land surface temperature time series; calculating the cumulative equivalent value of ecological service function for each ecological assessment unit based on the vegetation index time series and land surface temperature time series; collecting meteorological monitoring data corresponding to each ecological assessment unit; performing spatiotemporal dynamic correction on the cumulative equivalent value of ecological service function based on the meteorological monitoring data; generating a corrected cumulative equivalent value for each ecological assessment unit; constructing an ecological benefit spatial allocation matrix based on the corrected cumulative equivalent value of each ecological assessment unit and a land use type map of the target urban area; and determining the comprehensive ecological benefit index of the target urban area based on the ecological benefit spatial allocation matrix.

[0021] Example 1: In specific implementation, the process of acquiring multi-source remote sensing image data of the target city area includes receiving visible light, near-infrared, and thermal infrared images covering the target city area from satellite and airborne platforms, and collecting auxiliary information such as solar altitude angle and atmospheric parameters at the time of image acquisition. When radiometrically calibrating the multi-source remote sensing image data, the digital count values ​​of each band in the image are converted into physically meaningful radiance values. The conversion relationship is based on sensor calibration coefficients, which are extracted from image metadata. Atmospheric correction employs a correction method based on a radiative transfer model. Using atmospheric parameters, aerosol optical thickness, and surface elevation data, the influence of atmospheric scattering and absorption on radiance values ​​is eliminated, resulting in corrected image data. The reflectance and radiance values ​​of each pixel in the corrected image data reflect the true surface condition.

[0022] When extracting vegetation cover areas from corrected image data, the Normalized Difference Vegetation Index (NDI) threshold method is used, defining areas with an NDI greater than 0.2 as vegetation cover areas. Vegetation cover is calculated using a pixel-based binary model, decomposing the NDI of mixed pixels into vegetation-contributing and non-vegetation-contributing components. The vegetation cover calculation formula is as follows:

[0023] in, Indicates vegetation coverage. The normalized vegetation index (NDI) represents the value of a given pixel. This represents the normalized vegetation index value of a pixel representing pure bare soil. The normalized vegetation index value represents the pure vegetation pixel. and The value of was determined by analyzing the cumulative frequency histogram of the normalized vegetation index in the corrected image data, and the normalized vegetation index value corresponding to a cumulative frequency of 5% was selected as . The normalized vegetation index value corresponding to a cumulative frequency of 95% was selected as... The vegetation cover is calculated for each pixel in all vegetation-covered areas to generate a vegetation cover distribution map. Each pixel in the vegetation cover distribution map records the corresponding vegetation cover value.

[0024] When dividing the vegetation cover distribution map into multiple ecological assessment units, a spatial grid segmentation method is adopted. First, the coefficient of variation of the vegetation cover coverage of the entire region in the vegetation cover distribution map is calculated, and a semi-variogram is constructed to analyze the spatial autocorrelation distance of the vegetation cover. This spatial autocorrelation distance is used as a reference value for the grid edge length. In specific implementation, the spatial autocorrelation distance is multiplied by an adjustment factor to obtain the grid edge length. The adjustment factor is set to 1, making the grid edge length equal to the spatial autocorrelation distance. According to the determined grid edge length, starting from the upper left corner of the target city area, the entire area is divided into regular grids in the row and column directions. Each grid is an ecological assessment unit, and all ecological assessment units have equal areas. If the grid at the boundary of the target city area is incomplete, the incomplete grid is retained as an independent ecological assessment unit, with its area still consistent with the complete grid. The insufficient part is filled in by the outer boundary of the area, and the vegetation cover of the part outside the area is assigned a value of zero.

[0025] For each ecological assessment unit, when extracting the vegetation index time series, based on the set of pixel locations covered by the ecological assessment unit in the corrected image data, the normalized vegetation index (NVI) values ​​for these pixel locations are obtained one by one in each historical time phase. The arithmetic mean of the NVI values ​​for all pixel locations is used as the representative NVI value for the ecological assessment unit in that time phase. All historical time phases are arranged in chronological order to form the vegetation index time series. The number and interval of historical time phases are determined based on the acquisition cycle of the multi-source remote sensing image data.

[0026] For each ecological assessment unit, when extracting the land surface temperature time series, the radiance value of the thermal infrared band under each historical time phase is also extracted based on the set of pixel locations covered by the ecological assessment unit in the corrected image data. Land surface temperature inversion uses the atmospheric correction method, utilizing atmospheric profile data synchronized with the image, and calculating the land surface radiance value by subtracting the effects of upward and downward atmospheric radiation through the radiative transfer equation. Then, the land surface radiance value is converted into land surface temperature based on the Planck function and land surface emissivity. The land surface emissivity is estimated based on the empirical relationship between vegetation cover and land surface; the emissivity of pure vegetation is taken as 0.986, and the emissivity of pure bare soil is taken as 0.972. Pixel emissivity is obtained by weighting and mixing according to vegetation cover. The arithmetic mean of the land surface temperature of all pixels within each ecological assessment unit is used as the representative value of the land surface temperature of the ecological assessment unit at that time phase. The representative values ​​of the land surface temperature of all historical time phases are arranged in chronological order to form a land surface temperature time series.

[0027] Calculating the cumulative equivalent value of ecosystem service function for each ecological assessment unit requires traversing every historical time phase. For a given time phase, the normalized difference in vegetation index (NDVI) and land surface temperature (RTV) of the ecological assessment unit at that time phase are substituted into the equivalent conversion function. The equivalent conversion function is a pre-defined nonlinear function that jointly maps vegetation growth status and thermal environmental effects to the equivalent value of ecosystem service function. Its output is the equivalent value per unit area of ​​the ecological assessment unit at that time phase. The equivalent values ​​per unit area for all historical time phases of the ecological assessment unit are then integrated over time. The time integration method is phase-by-phase summation; that is, from the first time phase to the last, the equivalent value per unit area for each time phase is multiplied by the time interval between adjacent time phases and then summed. When the time intervals of all time phases are equal, the sum of the equivalent values ​​per unit area can be directly multiplied by the time interval. The result is the cumulative equivalent value of ecosystem service function. The time interval is in years or months, consistent with the temporal resolution of the historical time phases.

[0028] In some embodiments, the normalized vegetation index (NZDI) and land surface temperature (LTV) values, which are input parameters to the equivalent conversion function, are both normalized and scaled. This maps the NZDI values ​​to the interval between 0 and 1. Similarly, the LTV value is calculated by subtracting the minimum LTV value from the LTV value in the image and then dividing by the LTV range, also mapping the LTV value to the interval between 0 and 1. These values ​​are then substituted into the equivalent conversion function. This approach reduces the impact of parameter dimension differences on equivalent calculations.

[0029] See Figure 4 In the figure, the vertical axis represents the cumulative frequency of the Normalized Difference Vegetation Index (NDVI), expressed as a percentage (%); the horizontal axis represents the NDVI value, ranging from approximately -0.2 to 0.8. The blue solid line is the cumulative frequency curve of NDVI, depicting the cumulative distribution of NDVI values ​​for each pixel within the target city area. This curve gradually rises from the lower left to the upper right, exhibiting an S-shaped trend, indicating that areas with low NDVI values ​​account for a small proportion, while the proportion of pixels with high NDVI values ​​gradually increases until the cumulative frequency approaches 100%.

[0030] The red dashed line in the figure marks the 5th percentile of NDVI, which is approximately -0.031, reflecting the NDVI value of the lowest 5% of pixels in the target area. According to the vegetation cover calculation method in Example 1, the 5th percentile is selected as the NDVI of a pure bare soil pixel. The NDVI (Non-Vegetation Variable Index) is used to distinguish areas without vegetation cover. The green dashed line represents the 95th percentile of NDVI, which is approximately 0.769, corresponding to the NDVI of a pure vegetation pixel (…). ), which serves as the upper limit for vegetation cover decomposition.

[0031] Example 2, in specific implementation, refer to Figure 2When calculating the unit area equivalent value for a specific time phase for each ecological assessment unit, the representative values ​​of the Normalized Difference Vegetation Index (NDVI) and Land Surface Temperature (LST) are derived from the corresponding records in the vegetation index time series and land surface temperature time series, respectively. To obtain the difference between the NDVI and LST for that time phase, the NDVI is subtracted from the LST. To obtain the sum of the NDVI and LST, they are added together. The difference is then divided by the sum to obtain the vegetation temperature structure coefficient.

[0032] Simultaneously, when obtaining the product of the normalized vegetation index representative value and the land surface temperature representative value, the normalized vegetation index representative value is multiplied by the land surface temperature representative value to obtain the representative value product. Multiplying the representative value product by the vegetation temperature structure coefficient yields the unit area equivalent value for that time phase.

[0033] The above calculation process can be described by a single expression: the equivalent value per unit area. The calculation formula is:

[0034] in, This represents the unit area equivalent value of a specific ecological assessment unit at a specific time phase. This represents the normalized difference vegetation index (NDVI) value for the ecological assessment unit at that time phase. This represents the representative value of the land surface temperature for that ecological assessment unit at that time phase. The normalized difference vegetation index (NDVI) is also included. The value range is -1 to 1. When actually used in calculations, if... If the value is less than 0, set it to 0; Surface temperature representative value The unit is Kelvin, which needs to be converted to Celsius temperature value before being normalized to the 0 to 1 range.

[0035] Before substituting the normalized vegetation index representative value and the land surface temperature representative value into the calculation formula, the land surface temperature representative value is linearly normalized. The normalization method is to obtain the minimum land surface temperature representative value among all ecological assessment units covering the target urban area at that time phase. and maximum value , representing the surface temperature according to Mapped to normalized surface temperature values Then, using normalized surface temperature values Replace the original expression Participating in calculations. Similarly, the normalized vegetation index represents the value. It also needs to be mapped from the original value range to the 0 to 1 range, and the mapping method is as follows: Normalized value of vegetation index Alternative If it is involved in the calculation, the expression used in the actual calculation is:

[0036] at this time, The value range is from 0 to 1. The value range is also from 0 to 1. When At that time, set directly .

[0037] Once the normalized vegetation index (NVI) and land surface temperature (RTD) values ​​for a given time phase have been normalized, the following calculation steps are followed: First, calculate the difference between the normalized NVI value and the normalized RTD value. Then, calculate the sum of the two values ​​and obtain the ratio to get the vegetation temperature structure coefficient. Finally, multiply the product of the normalized NVI value and the normalized RTD value by the vegetation temperature structure coefficient to obtain the unit area equivalent value for that time phase.

[0038] In Example 3, during the specific implementation, when collecting meteorological monitoring data for each ecological assessment unit, the location information and observation records of all stations in the national or regional meteorological monitoring network covering the target urban area are obtained. For a given ecological assessment unit, the geometric center coordinates of the ecological assessment unit are determined, and the spatial distance between the geometric center coordinates and each meteorological monitoring station is calculated. The meteorological monitoring station with the smallest spatial distance is selected as the meteorological monitoring station corresponding to the ecological assessment unit. Daily rainfall data and daily temperature data within the time range completely consistent with the calculation period of the cumulative equivalent value of ecological service function are extracted from the historical observation database of the corresponding meteorological monitoring station. The daily rainfall data is the daily cumulative rainfall in millimeters, and the daily temperature data is the daily average temperature in degrees Celsius. The daily rainfall data is arranged in chronological order to form the rainfall time series data corresponding to the ecological assessment unit. The daily temperature data is arranged in chronological order to form the temperature time series data corresponding to the ecological assessment unit.

[0039] When calculating rainfall and temperature anomalies based on time-series rainfall and temperature data, a baseline period must first be determined. The baseline period is selected from the historical rainfall and temperature records of the same period over the past thirty years from meteorological monitoring stations. The past thirty years refer to a historical period within the same month or season as the rainfall time-series data. The monthly or seasonal average rainfall within the baseline period is calculated as the average rainfall for that period; the monthly or seasonal average temperature within the baseline period is also calculated as the average temperature for that period. The rainfall anomaly is calculated as follows: sum the cumulative daily rainfall in the rainfall time-series data to obtain the cumulative rainfall for that period; subtract the average rainfall for that period from the cumulative rainfall for that period, and then divide by the average rainfall for that period; the resulting ratio is the rainfall anomaly. If the average rainfall for that period is zero, the rainfall anomaly is set to zero. The temperature anomaly is calculated as follows: calculate the arithmetic mean of the daily average temperatures in the temperature time-series data to obtain the average temperature for that period; subtract the average temperature for that period from the average temperature for that period; the difference is the temperature anomaly.

[0040] The correction factor is determined based on the rainfall anomaly and the temperature anomaly, using the following formula:

[0041] in: The correction coefficient for the ecological assessment unit is dimensionless. This represents the preset baseline coefficient, which is set to 1.0. The basis for this value is that under ideal normal climate conditions where both rainfall anomaly and temperature anomaly are zero, the correction coefficient should not change the original cumulative value of the equivalent of ecosystem service function. The first weighting factor is 0.5. This value is based on the assumption that the marginal impact of rainfall fluctuations on ecosystem services is as important as that of temperature fluctuations on ecosystem services, each contributing half the weight, and that the range of variation of rainfall anomalies and temperature anomalies is similar after normalization. This represents the second weighting factor, with a value of 0.5, determined based on the same criteria as the first weighting factor. This represents the rainfall anomaly, which is dimensionless and is obtained as described above as the relative deviation between the cumulative rainfall over a period and the average rainfall during the same period. This represents the temperature anomaly, expressed in degrees Celsius. It is obtained as described above as the difference between the average temperature of the time period and the average temperature of the same period.

[0042] In some embodiments, due to temperature anomalies Anomaly with rainfall The numerical magnitudes and ranges of variation may differ. Before substituting into the above formula, the temperature anomaly values ​​are standardized by dividing the temperature anomaly value by the standard deviation of the temperature anomaly within the base period to obtain the standardized temperature anomaly value. This standardized temperature anomaly value is then multiplied by an adjustment factor of 0.5 to match its amplitude with the precipitation anomaly value. The standardized temperature anomaly value is used as the standard temperature anomaly value. Substitute into the formula. At this point, the first weighting factor... Second weighting factor All remain unchanged at 0.5.

[0043] Correction coefficients for ecological assessment units were obtained. Then, the cumulative equivalent value of ecosystem service functions is multiplied by a correction factor. The product is the corrected cumulative equivalent value of the ecological assessment unit. When the correction coefficient... A value greater than 1 indicates that climatic conditions are more favorable than usual for the functioning of green space ecosystem services, and the cumulative equivalent value is adjusted upwards after correction; when the correction coefficient is greater than 1, it indicates that the climatic conditions are more favorable than usual for the functioning of green space ecosystem services. A value less than 1 indicates unfavorable climate conditions, and the corrected cumulative equivalent value is adjusted downwards. The climate fluctuation factor is determined by... Calculations show that when the climate fluctuation factor is positive, the correction coefficient is increased, and when it is negative, the correction coefficient is decreased.

[0044] Example 4: In specific implementation, when obtaining the land use type map of the target urban area, the latest version of land use status vector data or raster data covering the target urban area is obtained from the geographic information database of the natural resources or planning department. The classification system of the land use type map adopts the national standard land use status classification, including categories such as cultivated land, orchards, forest land, grassland, commercial service land, industrial and mining land, residential land, public management and public service land, transportation land, and water area and water conservancy facilities land. Each map patch or pixel records a corresponding land use category code. The land use type map is spatially overlaid with the spatial grid of the ecological assessment unit for analysis. If the land use type map is in vector format, the vector data is intersected with the grid polygon of the ecological assessment unit, and the land use category code with the largest area proportion is taken as the land use category code corresponding to the ecological assessment unit. If the land use type map is in raster format, the land use category code that appears most frequently within the grid of the ecological assessment unit is taken as the land use category code corresponding to the ecological assessment unit.

[0045] For a given ecological assessment unit, after obtaining the corrected cumulative equivalent value and land use category code, a feature vector for the ecological assessment unit is generated. The feature vector is a two-dimensional column vector, where the first dimension represents the corrected cumulative equivalent value, and the second dimension represents the numerical form of the land use category code. The numerical form of the land use category code directly adopts the numerical code from the national standard land use status classification. Before adding the land use category code to the feature vector, a normalization process is performed on the land use category code. The normalization method is to divide the land use category code by the maximum category code value in the land use classification system, mapping the land use category code value to the interval between 0 and 1.

[0046] After generating corresponding feature vectors for all ecological assessment units, all feature vectors are arranged in spatial order. The spatial order is determined as follows: starting with the northernmost and westernmost ecological assessment units in the target urban area, they are arranged column by column from west to east, and within each column, they are arranged row by row from north to south. All ecological assessment units are numbered sequentially as 1, 2, ..., M, where M is the total number of ecological assessment units. The feature vectors of all ecological assessment units are arranged from left to right according to their numbering order, forming a 2xM matrix. This matrix is ​​the initial ecological benefit matrix. In the initial ecological benefit matrix, the element in the first row and m-th column is the first dimension element of the feature vector of the m-th ecological assessment unit, and the element in the second row and m-th column is the second dimension element of the feature vector of the m-th ecological assessment unit, where m is an integer from 1 to M.

[0047] After constructing the initial ecological benefit matrix, spatial autocorrelation iterative calculation is performed on the initial ecological benefit matrix. The spatial autocorrelation iterative calculation adopts an iterative format based on neighborhood smoothing and residual feedback. Each iterative step includes three sub-processes: neighborhood weighted averaging, iterative residual calculation, and residual relaxation superposition.

[0048] In the neighborhood weighted averaging subprocess, a spatial weight matrix is ​​constructed. The spatial weight matrix is ​​an M x M square matrix, where the element in the u-th row and v-th column represents the spatial adjacency weight between the u-th and v-th ecological assessment units, where u and v are integers from 1 to M. Spatial adjacency relationships are defined using the Rook adjacency rule: if the u-th and v-th ecological assessment units share an edge, the spatial adjacency weight is 1; otherwise, the spatial adjacency weight is 0. Row sum normalization is performed on each row of the spatial weight matrix by dividing each element in the u-th row by the sum of all elements in the u-th row, resulting in a normalized spatial weight matrix. The normalized spatial weight matrix is ​​then multiplied by the initial ecological benefit matrix by left-multiplying the normalized spatial weight matrix, resulting in a 2 x M first iteration matrix. The element in each column of the first iteration matrix is ​​the weighted average of the eigenvectors of the corresponding ecological assessment unit and its neighboring ecological assessment units.

[0049] In the iterative residual calculation process, the first iterative matrix is ​​subtracted from the initial ecological benefit matrix element by element. That is, the element in the first row and m column of the first iterative matrix is ​​subtracted from the element in the first row and m column of the initial ecological benefit matrix, and the element in the second row and m column of the first iterative matrix is ​​subtracted from the element in the second row and m column of the initial ecological benefit matrix. The result of the element-by-element subtraction forms an iterative residual matrix with 2 rows and M columns.

[0050] In the residual relaxation superposition process, a relaxation factor is introduced. The value of the relaxation factor determines the proportion of neighborhood information introduced in each iteration. The relaxation factor is set to 0.3. This value is based on the fact that in the spatial autocorrelation iteration process, to ensure the stability of iteration convergence, the relaxation factor generally ranges from 0 to 1. A value that is too small will result in slow convergence, while a value that is too large may lead to iteration oscillations. 0.3 strikes a balance between convergence speed and stability. Each element in the iterative residual matrix is ​​multiplied by the relaxation factor and then added to the corresponding element of the initial ecological benefit matrix to obtain the second iteration matrix. The expression for calculating the second iteration matrix is:

[0051] in, This represents the element in the r-th row and m-th column of the second iteration matrix, where r takes the value 1 or 2, corresponding to the first and second dimensions of the eigenvector, respectively, and m is the ecological assessment unit number. This represents the element in the r-th row and m-th column of the initial ecological benefit matrix at the start of the current iteration step. This represents the element in the r-th row and m-th column of the first iteration matrix. This represents the relaxation factor, with a value of 0.3.

[0052] After one iteration, the second iteration matrix is ​​used as the new initial ecological benefit matrix, and the neighborhood weighted averaging subprocess, iterative residual calculation subprocess, and residual relaxation superposition subprocess are repeated. The iteration terminates when the norm of the iterative residual matrix is ​​less than a preset threshold. The norm of the iterative residual matrix is ​​the Frobenius norm, which is calculated by taking the square root of the sum of the squares of all elements in the iterative residual matrix. The preset threshold is set to 0.001. This value is based on the fact that when the overall magnitude of the change in matrix elements between two iterations is less than one-thousandth, the spatial allocation structure can be considered to have stabilized, and the impact of continued iteration on the final ecological benefit spatial allocation matrix is ​​negligible. When the Frobenius norm is less than 0.001, the iteration terminates, and the second iteration matrix obtained from the last iteration is output as the ecological benefit spatial allocation matrix.

[0053] In some embodiments, to avoid non-convergence of the iterative process, a maximum number of iterations is set to 500. If the Frobenius norm is still not lower than a preset threshold of 0.001 when the number of iterations reaches 500, the iteration is forcibly terminated, and the current second iteration matrix is ​​used as the ecological benefit spatial allocation matrix. The ecological benefit spatial allocation matrix has 2 rows and M columns. The element in the first row and m-th column records the cumulative equivalent value allocation of the m-th ecological assessment unit after spatial autocorrelation iteration. The element in the second row and m-th column records the land use category code allocation of the m-th ecological assessment unit after spatial autocorrelation iteration.

[0054] Example 5, in specific implementation, refer to Figure 3 When determining the comprehensive ecological benefit index of a target urban area based on the spatial allocation matrix of ecological benefits, the spatial allocation matrix of ecological benefits is a 2-row, M-column matrix, where M is the total number of ecological assessment units. The element in the first row and m-th column of the spatial allocation matrix of ecological benefits is the cumulative equivalent value allocation after spatial autocorrelation iteration for the m-th ecological assessment unit, and the element in the second row and m-th column is the land use category code allocation for the m-th ecological assessment unit after spatial autocorrelation iteration.

[0055] When normalizing each element in the spatial allocation matrix of ecological benefits, the maximum and minimum values ​​of all elements in the first row and the second row are obtained respectively. For the element in the first row and m-th column, linear normalization is applied by subtracting the minimum value of the first row from the element value and then dividing by the difference between the maximum and minimum values ​​of the first row. The result is the normalized allocation value corresponding to the element in the first row and m-th column. The same method is used for the element in the second row and m-th column: subtracting the minimum value of the second row from the element value and then dividing by the difference between the maximum and minimum values ​​of the second row. The result is the normalized allocation value corresponding to the element in the second row and m-th column. If the maximum and minimum values ​​in a row are equal, the normalized allocation value for all elements in that row is set to 0.5.

[0056] For a given ecological assessment unit, the ecological assessment unit corresponds to two normalized allocation values ​​in the ecological benefit spatial allocation matrix, namely the normalized allocation value in the first row and the normalized allocation value in the second row. The arithmetic mean of these two normalized allocation values ​​is taken as the normalized allocation value of the ecological assessment unit.

[0057] When determining the ecological benefit contribution weight of each ecological assessment unit based on its normalized allocation value, the sum of the normalized allocation values ​​of all M ecological assessment units is first calculated and recorded as the sum of all normalized allocation values. For the m-th ecological assessment unit, its normalized allocation value is divided by the sum of all normalized allocation values; the resulting ratio is the ecological benefit contribution weight of that ecological assessment unit. The ecological benefit contribution weight ranges from 0 to 1, and the sum of the ecological benefit contribution weights of all M ecological assessment units equals 1.

[0058] The comprehensive ecological benefit index is obtained by weighting and summing the ecological benefit contribution weight of each ecological assessment unit with the corresponding corrected cumulative equivalent value. Here, the corrected cumulative equivalent value for each ecological assessment unit is the corrected cumulative equivalent value of the m-th ecological assessment unit generated in the spatiotemporal dynamic correction step, not the corrected cumulative equivalent value allocation in the ecological benefit spatial allocation matrix. The formula for the weighted summation is:

[0059] in: This represents a dimensionless comprehensive index of ecological benefits for the target urban area. Indicates the sequence number of the ecological assessment unit. The value of is an integer from 1 to M; This indicates the total number of ecological assessment units; Indicates the first The ecological benefit contribution weight of each ecological assessment unit is dimensionless and is obtained by dividing the normalized allocation value by the sum of all normalized allocation values. Indicates the first The corrected cumulative equivalent value of each ecological assessment unit is expressed in equivalent units. The corrected cumulative equivalent value is obtained by multiplying the cumulative equivalent value of ecological service functions by the correction factor.

[0060] After weighted summation, the resulting comprehensive ecological benefit index As a numerical value, it is used to characterize the overall strength of green space ecosystem service functions in the target urban area during the period of interest. The higher the value, the higher the overall ecological benefits of the green space ecosystem within the target city area.

[0061] See Figure 5 This figure is a bar chart showing the distribution of ecological benefit contribution weights for each ecological assessment unit within the target city area after processing the ecological benefit spatial allocation matrix in Example 5. The horizontal axis represents the range of ecological benefit contribution weights, which are approximately between 0.0003 and 0.009. The vertical axis represents the frequency of the ecological assessment unit within the corresponding weight range. The legend indicates that the data shown in the bar chart is "Distribution of Ecological Benefit Contribution Weights".

[0062] As shown in the figure, the contribution weights of ecological benefits exhibit a clear right-skewed distribution. The contribution weights of the vast majority of ecological assessment units are concentrated in the lower range, especially between 0.0003 and 0.002, with a high frequency peak of approximately 80, indicating that the number of ecological assessment units is highest in this weight range. As the contribution weight increases, the number of ecological assessment units gradually decreases, with units having a weight exceeding 0.004 becoming significantly rare. The highest weight value is approximately 0.009, with a frequency below 5, indicating that there are relatively few ecological assessment units with large weights.

[0063] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for intelligent monitoring and data evaluation of the ecological benefits of urban green spaces, characterized in that, The method includes: Acquire multi-source remote sensing image data within the target city area, parse the vegetation cover distribution map from the multi-source remote sensing image data, and divide multiple ecological assessment units based on the vegetation cover distribution map; For each ecological assessment unit, its corresponding vegetation index time series and land surface temperature time series are extracted. Based on the vegetation index time series and the land surface temperature time series, the cumulative value of the equivalent value of the ecological service function of each ecological assessment unit is calculated. The cumulative value of the equivalent value of the ecological service function is obtained by dividing the difference between the normalized vegetation index and the land surface temperature in the same time phase by the sum of the two to obtain the vegetation temperature structure coefficient. The coefficient is multiplied by the product of the two to obtain the unit area equivalent value. The unit area equivalent value of all time phases is obtained by time integration. Meteorological monitoring data corresponding to each of the ecological assessment units is collected, and the cumulative equivalent value of the ecosystem service function is dynamically corrected in time and space based on the meteorological monitoring data to generate the corrected cumulative equivalent value of each of the ecological assessment units. Based on the corrected cumulative equivalent value of each ecological assessment unit, and combined with the land use type map of the target urban area, an ecological benefit spatial allocation matrix is ​​constructed. Based on the aforementioned ecological benefit spatial allocation matrix, the comprehensive ecological benefit index of the target urban area is determined.

2. The intelligent monitoring and data evaluation method for the ecological benefits of urban green spaces according to claim 1, characterized in that, The process of acquiring multi-source remote sensing image data within the target city area, parsing a vegetation cover distribution map from the multi-source remote sensing image data, and dividing the vegetation cover distribution map into multiple ecological assessment units specifically includes: Radiometric calibration and atmospheric correction are performed on the multi-source remote sensing image data to obtain corrected image data; The vegetation cover area is extracted from the corrected image data, and the vegetation cover degree is calculated to obtain the vegetation cover distribution map; Based on the spatial variation characteristics of vegetation cover in the vegetation cover distribution map, the target urban area is divided into multiple ecological assessment units of equal area using a spatial grid segmentation method.

3. The intelligent monitoring and data evaluation method for the ecological benefits of urban green spaces according to claim 2, characterized in that, For each ecological assessment unit, its corresponding vegetation index time series and land surface temperature time series are extracted. Based on the vegetation index time series and the land surface temperature time series, the cumulative equivalent value of the ecosystem service function of each ecological assessment unit is calculated, specifically including: Based on the pixel position corresponding to each ecological assessment unit in the corrected image data, the normalized vegetation index of the unit under multiple historical time phases is extracted to form the vegetation index time series. Based on the pixel position of each ecological assessment unit in the corrected image data, the thermal infrared radiance value of the unit under multiple historical time phases is extracted, and the land surface temperature time series is obtained by inversion. For each ecological assessment unit, the normalized vegetation index and land surface temperature at the same time phase are substituted into the equivalent transformation function to obtain the unit area equivalent value for that time phase. The cumulative equivalent value of the ecosystem service function is obtained by integrating the unit area equivalent value of each ecological assessment unit across all time phases.

4. The intelligent monitoring and data evaluation method for the ecological benefits of urban green spaces according to claim 3, characterized in that, Meteorological monitoring data corresponding to each of the ecological assessment units is collected. Based on the meteorological monitoring data, the cumulative equivalent value of the ecosystem service function is dynamically corrected in time and space to generate the corrected cumulative equivalent value for each of the ecological assessment units. Specifically, this includes: For each ecological assessment unit, the time-series data of rainfall and temperature of the meteorological monitoring station at the location of the unit are obtained as the meteorological monitoring data corresponding to the unit; Based on the time-series rainfall data and the time-series temperature data, calculate the rainfall anomaly and temperature anomaly for this unit, respectively. The correction coefficient for this unit is determined based on the rainfall anomaly and the temperature anomaly. The corrected cumulative equivalent value is obtained by multiplying the cumulative value of the ecosystem service function equivalent by the correction coefficient.

5. The intelligent monitoring and data evaluation method for the ecological benefits of urban green spaces according to claim 4, characterized in that, Based on the corrected cumulative equivalent value of each ecological assessment unit, and in conjunction with the land use type map of the target urban area, a spatial allocation matrix of ecological benefits is constructed, specifically including: Obtain the land use type map of the target urban area, and extract the land use category code corresponding to each ecological assessment unit from the land use type map; Based on the corrected cumulative equivalent value and the land use category code of each ecological assessment unit, a feature vector of the unit is generated; Arrange the feature vectors of all the ecological assessment units in spatial order to form an initial ecological benefit matrix; Spatial autocorrelation iteration is performed on the initial ecological benefit matrix, and the spatial allocation matrix of ecological benefits is obtained after convergence.

6. The intelligent monitoring and data evaluation method for the ecological benefits of urban green spaces according to claim 5, characterized in that, Based on the aforementioned spatial allocation matrix of ecological benefits, the comprehensive ecological benefit index of the target urban area is determined, specifically including: Each element in the ecological benefit spatial allocation matrix is ​​normalized to obtain a normalized allocation value; The ecological benefit contribution weight of each ecological assessment unit is determined based on the normalized allocation value of each unit. The ecological benefit contribution weight of each ecological assessment unit is weighted and summed with the corresponding corrected cumulative equivalent value to obtain the comprehensive ecological benefit index.

7. The intelligent monitoring and data evaluation method for the ecological benefits of urban green spaces according to claim 3, characterized in that, For each ecological assessment unit, the normalized vegetation index and land surface temperature at the same time phase are substituted into the equivalent transformation function to obtain the unit area equivalent value for that time phase, specifically including: The difference between the normalized vegetation index and the land surface temperature at that time phase is obtained, and the difference is divided by the sum of the normalized vegetation index and the land surface temperature to obtain the vegetation temperature structure coefficient. The product of the normalized vegetation index and the land surface temperature is obtained, and the product is multiplied by the vegetation temperature structure coefficient to obtain the unit area equivalent value for that time phase.

8. The intelligent monitoring and data evaluation method for the ecological benefits of urban green spaces according to claim 4, characterized in that, Based on the rainfall anomaly and the temperature anomaly, the correction coefficient for this unit is determined, specifically including: Multiply the rainfall anomaly by the first weighting factor to obtain the rainfall component, and multiply the temperature anomaly by the second weighting factor to obtain the temperature component; The climate fluctuation factor is obtained by adding the rainfall component to the temperature component. The correction coefficient is obtained by adding the preset benchmark coefficient to the climate fluctuation factor.

9. The intelligent monitoring and data evaluation method for the ecological benefits of urban green spaces according to claim 5, characterized in that, The step of performing spatial autocorrelation iterative calculation on the initial ecological benefit matrix, and obtaining the ecological benefit spatial allocation matrix after convergence, specifically includes: The initial ecological benefit matrix is ​​weighted by a neighborhood-weighted average using a spatial weight matrix to obtain the first iteration matrix; Calculate the difference between the first iteration matrix and the initial ecological benefit matrix to obtain the iteration residual matrix; The iterative residual matrix is ​​multiplied by a relaxation factor and then added to the initial ecological benefit matrix to obtain the second iterative matrix; Using the second iterative matrix as the new initial ecological benefit matrix, the above iterative steps are repeated until the norm of the iterative residual matrix is ​​less than a preset threshold, thus converging to obtain the ecological benefit spatial allocation matrix.

10. The intelligent monitoring and data evaluation method for the ecological benefits of urban green spaces according to claim 6, characterized in that, Based on the normalized allocation value of each ecological assessment unit, the ecological benefit contribution weight of that unit is determined, specifically including: The ecological benefit contribution weight of a unit is obtained by dividing the normalized allocation value of each ecological assessment unit by the sum of the normalized allocation values ​​of all ecological assessment units.

Images