A spatiotemporal assessment method, system, equipment, and medium for the ecological quality of watershed human settlement systems.

Abstract

This invention discloses a method, system, equipment, and medium for spatiotemporal assessment of the ecological quality of a watershed human settlement system, relating to the field of ecological quality assessment technology. The method includes: acquiring target area data, with the Yellow River source region as a typical watershed human settlement system; calculating the Ecological Quality Index (EQI) based on the target area data; using 2m resolution patches based on real ecological boundaries to replace the idealized large-scale raster as the target area assessment unit, and integrating the EQI and collected potential influencing factor data into the target area assessment unit to generate ecological patch units; determining the spatiotemporal assessment results of the ecological quality based on the ecological patch units using Moran's I index and an ecological quality change matrix; the spatiotemporal assessment results of the ecological quality are used to describe the spatial autocorrelation and spatiotemporal changes of ecological quality. This invention can improve the accuracy of spatiotemporal assessment of the ecological quality of watershed human settlement systems.

Description

Technical Field

[0001] This invention relates to the field of ecological quality assessment technology, and in particular to a spatiotemporal assessment method, system, equipment and medium for ecological quality assessment of watershed human settlement systems. Background Technology

[0002] Ecological quality assessment is an important means of quantitatively evaluating the ecological status of watershed human settlement systems. Although existing studies have widely used principal component analysis (PCA) to assess ecological quality, it has limitations in time-series analysis, such as inaccurate trend analysis and insufficient representation of spatial heterogeneity.

[0003] Currently, most ecological quality assessments and influencing factor analyses are based on large-scale raster data units, commonly with resolutions of 30m and 1000m, or on administrative units to assess the ecological quality of the entire region. These artificially defined, idealized grids or administrative units exhibit significant heterogeneity in their internal ecological spaces and complex relationships between geographical elements. This can lead to interference from incompatible information when describing ecological quality characteristics, and the possibility that locally high and low ecological quality points may be largely ignored due to being randomly assigned to the same unit and averaged.

[0004] Therefore, existing research still has the following problems: the application of PCA to assess ecological quality is inaccurate in terms of the changing trends over time series, and assessments based on artificially set grid units may lead to insufficient characterization of spatial heterogeneity. Summary of the Invention

[0005] The purpose of this invention is to provide a method, system, device, and medium for spatiotemporal assessment of the ecological quality of watershed human settlement systems, which can improve the accuracy of spatiotemporal assessment of the ecological quality of watershed human settlement systems.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] A spatiotemporal assessment method for the ecological quality of watershed human settlement systems includes:

[0008] Data on the target area of ​​the Yellow River source region as a typical watershed human settlement system was obtained; the target area data was provided by the MODIS dataset and the LANDSAT dataset.

[0009] Calculate the Ecological Quality Index (EQI) based on the data from the target area.

[0010] A 2m resolution patch based on real ecological boundaries is used instead of an idealized large-scale raster as the target area assessment unit. The ecological quality index (EQI) and the collected potential influencing factor data are integrated into the target area assessment unit to generate an ecological patch unit.

[0011] Based on the ecological patch units, the spatiotemporal assessment results of ecological quality are determined using Moran's I index and the ecological quality change matrix; the spatiotemporal assessment results of ecological quality are used to describe the spatial autocorrelation and spatiotemporal changes of ecological quality.

[0012] Optionally, the calculation of the Ecological Quality Index (EQI) based on the target area data specifically includes:

[0013] Based on the MODIS dataset, the leaf area index (LAI) was obtained using MCD15A3H, and the total primary productivity (GPP) was obtained using MOD17A2H.

[0014] The NDVI value was obtained using TI_L2 data from the LANDSAT series satellites, and the vegetation cover FVC was calculated.

[0015] Initial index data were constructed based on the leaf area index (LAI), total primary productivity (GPP), and vegetation cover (FVC).

[0016] The initial index data from different years were normalized in the 0-1 range, and PCA analysis was used to unify the dimensions and weights of the same index across different years, thus forming the initial ecological quality index EQI. c , represented as:

[0017]

[0018] in, a , b , c PC represents the weighting coefficient. n For new explanatory variables formed during PCA analysis, n∈{1,2,3};

[0019] For the initial ecological quality index EQI c Normalization is performed to obtain the final ecological quality index EQI.

[0020] Optionally, the step of using 2m resolution patches based on real ecological boundaries to replace the idealized large-scale raster as the target area assessment unit, and integrating the Ecological Quality Index (EQI) and the collected potential influencing factor data into the target area assessment unit to generate ecological patch units, specifically includes:

[0021] Data on potential influencing factors of ecological quality were selected based on the data of the target area.

[0022] First, all raster images are resampled to 2m. Then, each ecological patch is assigned a unique ID. According to spatial location, the Ecological Quality Index (EQI) and the potential impact factor data are integrated within each ecological patch. The mean value of the impact factor is calculated based on the potential impact factor data. Ecological patch units are generated based on the Ecological Quality Index (EQI) and the mean value of the impact factor. The Ecological Quality Index (EQI) within each patch is divided into five levels: 0.8-1 is the best, 0.6-0.8 is good, 0.4-0.6 is medium, 0.2-0.4 is poor, and 0-0.2 is the worst.

[0023] Optionally, the step of determining the spatiotemporal assessment results of ecological quality based on the ecological patch units, using Moran's I index and the ecological quality change matrix, specifically includes:

[0024] Based on the ecological patch units, firstly, the global Moran's I is used to perform spatial autocorrelation analysis of the overall ecological quality of the target area. The overall spatial autocorrelation pattern is determined by Moran's I index, z-score, and p-value. Secondly, the local Moran's I is used to analyze the local spatial ecological quality correlation and heterogeneity of each spatial unit to obtain the LISA clustering diagram.

[0025] An ecological quality change matrix is ​​constructed based on the ecological quality level in the ecological patch unit; the ecological quality change matrix is ​​the corresponding area of ​​the ecological quality level transformation.

[0026] Optionally, it also includes: determining the main influencing factors of ecological quality based on multiple regression methods and the spatiotemporal assessment results of ecological quality; the main influencing factors of ecological quality are used to provide a reference for resource allocation for ecological protection; the multiple regression methods include OLS, GWR and MGWR.

[0027] This invention also provides a spatiotemporal assessment system for the ecological quality of watershed human settlement systems, comprising:

[0028] The data acquisition unit is used to acquire target area data of the Yellow River source region; the target area data is provided by the MODIS dataset and the LANDSAT dataset.

[0029] An index construction unit is used to calculate the Ecological Quality Index (EQI) based on the target area data.

[0030] An ecological patch construction unit is used to replace the idealized large-scale raster with a 2m resolution patch based on the real ecological boundary as the evaluation unit, and integrates the ecological quality index (EQI) and potential impact factor data into the ecological patch unit.

[0031] An assessment unit is used to determine the spatiotemporal assessment results of ecological quality based on the ecological patch units, using Moran's I index and the ecological quality change matrix; the spatiotemporal assessment results of ecological quality are used to describe the spatial autocorrelation and spatiotemporal changes of ecological quality.

[0032] The present invention also provides an electronic device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to perform the spatiotemporal assessment method for the ecological quality of watershed human settlement systems as described above.

[0033] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for spatiotemporal assessment of the ecological quality of watershed human settlement systems.

[0034] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0035] This invention discloses a method, system, device, and medium for spatiotemporal assessment of the ecological quality of a watershed human settlement system. The method includes acquiring target area data, with the Yellow River source region as a typical watershed human settlement system; the target area data is provided by the MODIS and LANDSAT datasets; calculating the Ecological Quality Index (EQI) based on the target area data; replacing the idealized large-scale raster with 2m resolution patches based on real ecological boundaries as the target area assessment unit, and integrating the EQI and collected potential influencing factor data into the target area assessment unit to generate ecological patch units; determining the spatiotemporal assessment results of the ecological quality based on the ecological patch units using Moran's I index and an ecological quality change matrix; the spatiotemporal assessment results of the ecological quality are used to describe the spatial autocorrelation and spatiotemporal variation of ecological quality. This invention can improve the accuracy of spatiotemporal assessment of the ecological quality of watershed human settlement systems. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is a flowchart of the spatiotemporal assessment method for the ecological quality of watershed human settlement systems according to the present invention;

[0038] Figure 2 This is a schematic diagram of the logic flow in this embodiment;

[0039] Figure 3This is the PCA optimization flowchart in this embodiment; where (a) is the traditional flowchart and (b) is the optimized flowchart.

[0040] Figure 4 This is a diagram illustrating the ecological quality area change matrix method in this embodiment;

[0041] Figure 5 This is a schematic diagram of the PCA results in this embodiment;

[0042] Figure 6 This is a Lisa clustering diagram of the local Moran's I results for 2015 and 2022 in this embodiment;

[0043] Figure 7 This is a matrix diagram showing the changes in the area of ​​ecological quality at various levels from 2015 to 2022 in this embodiment. Detailed Implementation

[0044] 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, and 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.

[0045] The purpose of this invention is to provide a method, system, device, and medium for spatiotemporal assessment of the ecological quality of watershed human settlement systems, which can improve the accuracy of spatiotemporal assessment of the ecological quality of watershed human settlement systems.

[0046] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] like Figure 1 As shown, this invention provides a spatiotemporal assessment method for the ecological quality of watershed human settlement systems, comprising:

[0048] Step 100: Obtain target area data of the Yellow River source region; the target area data is provided by the MODIS dataset and the LANDSAT dataset.

[0049] Step 200: Calculate the Ecological Quality Index (EQI) based on the target area data.

[0050] Step 300: Replace the idealized large-scale raster with a 2m resolution patch based on the real ecological boundary as the target area assessment unit, and integrate the ecological quality index EQI and the collected potential influencing factor data into the target area assessment unit to generate ecological patch units.

[0051] Step 400: Based on the ecological patch units, the spatiotemporal assessment results of ecological quality are determined using Moran's I index and the ecological quality change matrix; the spatiotemporal assessment results of ecological quality are used to describe the spatial autocorrelation and spatiotemporal changes of ecological quality.

[0052] As a specific implementation method, the above steps will be explained in detail using Maduo County as an example.

[0053] In this implementation, the data used to characterize EQI were obtained from the MODIS and LANDSAT datasets. Leaf area index (LAI) was obtained from MCD15A3H (500m resolution and 4-day interval) in the MODIS dataset, total primary productivity (GPP) was obtained from MOD17A2H (500m resolution and 8-day interval), and NDVI was obtained from TI_L2 (30m resolution) in the LANDSAT series satellite data. Vegetation cover (FVC) was calculated from these results. All datasets were from public archives on the GEE platform, and a modified normalized water index (MNDWI) was used to mask water bodies in the target area.

[0054] The ecological patch data was obtained from 2m resolution remote sensing images of the Gaofen-1 satellite in 2015 and 2022. After preprocessing the remote sensing data, including radiometric calibration, atmospheric correction, and geometric correction, computer-aided automatic identification and interactive visual interpretation were used to obtain five primary types and 23 secondary types of wetland ecosystems, shrubland ecosystems, grassland ecosystems, bare land, and urban ecosystems. The classification standard referenced the land use and cover classification system of the Chinese Academy of Sciences, resulting in approximately 70,000 ecological patches.

[0055] The multi-source data used for the analysis of ecological quality impact factors are shown in Table 1. These include climate factors such as annual average temperature, annual (ground) average relative humidity, annual average precipitation, potential evapotranspiration, and annual average wind speed; geographical factors such as elevation, slope, aspect, and distance to major water systems; and social factors such as distance to major roads, distance to settlements, population density, and land use type.

[0056] Table 1. Source of Data

[0057]

[0058] like Figure 2As shown, this embodiment includes four main steps: (1) using the optimized PCA process to extract features of LAI, GPP and FVC to form EQI for ecological quality assessment; (2) interpreting 2m remote sensing images to obtain high-precision ecological patches, and integrating the ecological quality assessment results and potential influencing factor data into the ecological patch unit; (3) based on the ecological patch unit, using Moran's I and the ecological quality change matrix, exploring the spatial autocorrelation and spatiotemporal changes of ecological quality; (4) using multiple regression methods such as OLS (least square regression), GWR (geographically weighted regression) and MGWR (multi-scale geographically weighted regression) to analyze the main influencing factors of ecological quality, providing a reference for resource allocation for ecological protection.

[0059] Ecological quality assessment based on EQI:

[0060] The Ecological Quality Index (EQI) is used to reflect the quality of the ecosystem (excluding water bodies) in the target area. According to the 2021 standard document from the Ministry of Ecology and Environment of China, three indicators—LAI, GPP, and FVC—are used to characterize the EQI. Principal component analysis (PCA), a commonly used method in constructing comprehensive ecological quality index systems such as RSEI, is adopted to objectively and maximize the reflection of the comprehensive characteristics of the original indicator data. Since the units and dimensions of the three initial indicators are inconsistent, the data is first normalized before PCA is used to construct the EQI.

[0061] It should be noted that this embodiment optimizes the process of using the PCA method (e.g., Figure 3 As shown). The existing ecological quality assessment method involves the following process: initial index normalization for each period - PCA - normalization (e.g., ...). Figure 3 As shown in Part a), the ecological quality data from the two periods, as their respective spatial relative values, cannot be compared over time. Starting from the step highlighted in the yellow box, the two indicators are no longer comparable. This study normalizes the same initial indicator data from both years (to the 0-1 interval) and then performs PCA to unify the dimensions and weights of the same indicators across different years. The new explanatory variables formed by PCA analysis can be represented as PC... n If n∈{1,2,3}, then EQI c This can be expressed as the following formula:

[0062]

[0063] in, a , b , c PC represents the weighting coefficient. n New explanatory variables formed during the PCA analysis, n∈{1,2,3}; retain the corresponding number of PCs based on the cumulative contribution rate (≥85%).

[0064] The final ecological quality result EQI for PCA c Normalize to obtain EQI, and extract the EQI for each year according to the year attribute.

[0065] Ecological quality data integration based on map features:

[0066] To explore the changes and driving factors of ecological quality and provide a basis for human understanding and intervention in ecological quality changes, this study selected potential influencing factors of ecological quality. Ecological patches interpreted from 2m high-resolution remote sensing imagery were used as spatial units. The EQI results and potential influencing factor data were mapped to vector patch units to prepare for subsequent precise analysis of the main influencing factors of the EQI of each ecological patch. First, all raster images were resampled to 2m to avoid errors caused by the cutting of large-scale rasters by vector unit boundaries. Then, each ecological patch was assigned a unique ID, and the EQI and mean influencing factor of the rasters contained within each ecological patch were calculated sequentially according to spatial location. The EQI of the patches was divided into five levels: best (0.8–1), good (0.6–0.8), medium (0.4–0.6), poor (0.2–0.4), and worst (0–0.2) for a direct comparison of changes over two years.

[0067] Spatiotemporal assessment of ecological quality:

[0068] (1) Spatial correlation of ecological quality

[0069] By studying the spatial correlation of ecological quality within the target area, namely spatial autocorrelation and spatial heterogeneity, this study explores the relationship between changes in ecological quality and spatial geographical location from both overall and local spatial perspectives, revealing the spatial variation patterns of ecological quality in the target area.

[0070] Moran's I index is an important indicator for analyzing the degree of spatial correlation. First, global Moran's I is used to perform spatial autocorrelation analysis of the overall ecological quality of the target area. The overall spatial autocorrelation pattern is determined by the Moran's I index, z-score (representing the multiple of the standard deviation), and p-value. Second, local Moran's I is used to analyze the local spatial ecological quality correlation and heterogeneity of each spatial unit, resulting in a LISA clustering diagram.

[0071] (2) Ecological quality change matrix

[0072] Use the ecological quality change matrix (e.g.) Figure 4 The area corresponding to the ecological quality level transition from 2015 to 2022 (as shown in the image) is used to understand the overall changes in the ecological quality of the target area. Furthermore, based on ecological patches, the degree of change in ecological quality levels over the two years is spatially visualized to explore changes in local spaces. Figure 4 In the diagram, B1-B5 represent the five ecological quality levels from worst to best in 2015, and L1-L5 represent the five ecological quality levels from worst to best in 2022. Then S... ij Let represent the total area from ecological quality level i in 2015 to ecological quality level j in 2022 (i∈[1,5], j∈[1,5]).

[0073] Analysis of factors influencing ecological quality:

[0074] Based on ecological quality and influencing factor data contained in ecological patches, this study explores the main influencing factors. First, the influencing factor data were normalized to eliminate the impact of dimensional differences on the model regression. Then, OLS was used to test for collinearity, removing highly collinear factors and retaining those with a variance inflation factor (VIF) > 7.5, obtaining the OLS-based regression analysis results. Subsequently, GWR and MGWR analyses were used to analyze the influencing factors of ecological quality under geospatial influences. The modified Akaike Information Content (AICc) index values ​​of the three regression models were compared to measure model performance; a smaller AICc indicates better model performance, and the model with the best regression fit was selected.

[0075] As a specific experimental result, the following conclusions are presented.

[0076] Ecological quality status based on EQI:

[0077] This embodiment performs principal component analysis (PCA) on the initial ecological index data from the two phases according to the optimized process to obtain EQI results. From the PCA results (such as...) Figure 5 As shown in the figure, the contribution rate of the first principal component (PC1) data reached 89.8%, which is greater than 85%, indicating that PC1 explains most of the characteristics of the target region's EQI and can be used to characterize the target region's EQI in 2015 and 2022.

[0078] In both 2015 and 2022, the ecological quality showed a spatial differentiation, with higher quality in the south, southwest, and southeast, and lower quality in the north. The areas with the worst ecological quality were generally concentrated from north to south in the northern part of Donggecuona Lake, the central area north of Zalinghu Township, the area around Madoi County, and the Yellow River basin extending southeast from the southeast corner of Eling Lake to Huanghe Township and then to the southeastern part of the target area. The overall ecological quality of the core protected area was relatively high, especially in the southeastern core protected area; however, some areas in the northern part of the core protected area still exhibited poor or severe ecological quality, requiring further attention and improvement.

[0079] Spatiotemporal variation of EQI in patch units:

[0080] Spatial heterogeneity of EQI: Moran's I for ecological quality in 2015 and 2022 were 0.681 and 0.791, respectively (as shown in Table 2). Meanwhile, the Z-value was much greater than 2.58 and the P-value was 0.001, indicating that the EQI in the target area exhibits spatial clustering and spatial heterogeneity at a 99% confidence level. To more intuitively characterize and further understand the spatial distribution characteristics of EQI in the target area, the spatial distribution of ecological quality of map units was reflected based on local Moran's I.

[0081] Table 2 Global Moran's I

[0082]

[0083] Using LISA clustering graphs to obtain and analyze local spatial association patterns of EQI (e.g.) Figure 6 (As shown). Insignificant areas are mainly distributed around and south of Lake Eling and south of Donggecuona Lake. Areas with poor LL ecological quality are mainly located in the northern part of the target area and around the town of Maduo County, distributed along the southeastern direction of the Yellow River. From 2015 to 2022, the area of ​​LL clusters in the southern part of Maduo County and the northern part of the town decreased significantly, changing from large-area clusters to a strip-like distribution, reflecting a substantial improvement in ecological quality. H=H clusters expanded into large areas in the southeastern part of the target area.

[0084] Two-period ecological quality change matrix: From 2015 to 2022, the overall ecological quality has significantly improved (e.g., Figure 7 In two squares located along diagonal 2 and equidistant from diagonal 1, the area values ​​of the green areas are almost always greater than those of the yellow areas. This indicates that during the transformation of ecological quality at the same level, the area of ​​quality improvement is far greater than the area of ​​quality decline. Clearly, the overall ecological quality of Maduo County in the Yellow River Source Area showed an upward trend from 2015 to 2022. The largest changes in area were observed from B2 to L3 and from B3 to L4. The proportions of EQI degradation, no change, and improvement were 13.6%, 62.0%, and 24.4%, respectively.

[0085] The spatial distribution of ecological quality grade changes reflects the spatial pattern of ecological quality changes from 2015 to 2022. In most areas, the ecological quality grade remained unchanged from 2015 to 2022, remaining stable. Most changes in ecological quality involved a single grade shift. Grade declines were mainly concentrated in the northern parts of Zaling Lake and Eling Lake, and the eastern part of Maduo County. Grade increases in ecological quality were almost randomly and evenly distributed, except in the eastern and northwestern areas of Maduo County.

[0086] The Importance of Ecological Quality Influencing Factors

[0087] Potential factors affecting ecological quality:

[0088] Considering that the target area is largely composed of natural ecosystems, potential factors for ecological quality changes were screened out. Factors with similar meanings and reflecting overall structural characteristics were removed. Based on data availability, 11 representative influencing factors were retained, and the EQI and influencing factor overview based on ecological patch units were statistically obtained.

[0089] The importance of impact factors based on OLS: First, multicollinearity factors were removed using the VIF value of the OLS regression results, and insignificant factors were removed based on the P-value, resulting in the impact factors (as shown in Table 3). Both years of OLS results showed significant K values ​​(Koenker (BP) statistic) and chi-square (joint F statistic). Therefore, the explanatory variables of this ecological quality regression model have inconsistent relationships with the dependent variable in both geographic and data spaces, requiring further spatial analysis using Geographically Weighted Regression (GWR) and Multiscale Geographically Weighted Regression (MGWR).

[0090] Table 3. Test results using the OLS method

[0091]

[0092] Spatial Impact Factor Importance: The importance of impact factors to ecological quality was analyzed using both GWR and MGWR methods, and the model performance was compared in conjunction with OLS results. The MGWR model had the lowest AICc, indicating the best model fit. The adjusted R-value of the MGWR model was also the highest. 2 The MGWR model has the largest sum of squared residuals (RSS), explaining 66.4% of the ecological quality in 2015 and 71.3% in 2022, indicating that it effectively reveals the different effects of climate, geography, and anthropogenic factors on the ecological quality of Maduo County. The MGWR model also has the smallest overall RSS, suggesting it obtains regression results closer to the true values. Therefore, using the MGWR model is optimal for studying the importance of ecological quality influencing factors. By comprehensively considering the influence of geospatial factors and the diversity of scales of various ecological quality influencing factors, it reduces noise and bias in the regression coefficients, making the ecological quality regression results closer to reality.

[0093] The influence of climate factors, geographical factors, and human activities on ecological quality exhibits significant spatial heterogeneity. The bandwidths used for different influencing factors in the MGWR vary considerably, directly reflecting the differentiated scales of their impact on the ecological quality of Maduo County. Specifically, in 2015, the bandwidths for PET, DIS_WA, RH, and WS were all greater than 14000, indicating a relatively large scale of influence on the ecological quality of Maduo County. Conversely, the bandwidths for LDU, DIS_RE, and SLO were all below the MGWR global bandwidth of 2939, indicating a relatively small scale of influence. LDU's scale was significantly the smallest, demonstrating that even with smaller-scale spatial variations, LDU can still impact ecological quality.

[0094] The MGWR model was used to obtain the regression coefficients of influencing factors. Overall, climate factors mostly had a positive impact on ecological quality, while geographical factors showed no significant positive or negative effects in this study. Human activity factors all showed a trend of decreasing ecological quality with increasing intensity of human activities. Specifically, in 2015, ecological quality had a relatively significant positive correlation with MAP, PET, SLO, and LDU, and a relatively significant negative correlation with WS. In 2022, ecological quality had a relatively significant positive correlation with MAP and LDU, and a relatively significant negative correlation with WS, DIS_RE, and POP.

[0095] Based on the combined results of the two years, the influencing factors MAP, POP, and LDU have the greatest impact on the ecological quality of Maduo County in the Yellow River source region. Compared to 2015, the influence of RH and LDU increased significantly in 2022. This indicates that the ecological quality of Maduo County is mainly affected by climate factors and human activities, and the impact was stronger in 2022 than in 2015.

[0096] By spatially visualizing the coefficients of the MGWR (Magnetic Metrics and Environmental Data Reduction) ecological quality impact factors in 2015 and 2022, we can further understand the spatial differentiation of the importance of different impact factors. Overall, the spatial distribution of the local coefficients of the above impact factors is non-random, exhibiting local spatial clustering characteristics. In the coefficients of the 2015 impact factors, the coefficients of each impact factor are affected by spatial distribution, with high-value coefficients showing local clustering characteristics, and the clustering areas are different. A large part of the MAP coefficient range is concentrated between -9.99 and -1, indicating that for every 1% increase in MAP, the ecological quality in this range decreases by 1-9.99, mainly reflected in Huanghe Township in the south of Madoi County, the western part of Zalinghu Township, and the northern part of Huashixia Town.

[0097] Based on the 2022 impact factor coefficients, most coefficient values ​​exhibit a spatially uniform distribution, indicating that different impact factors have varying effects on the same ecological patch unit. From these results, ecological quality data and preliminary improvement plans for each patch unit are obtained. A certain patch has an ecological quality of 0.614, classified as "good." Combining this with its ecological quality impact factor regression data, it can be seen that the requirements for impact factors vary depending on the patch's ecological quality. If the average annual rainfall is increased by 19.28 mm, considering the synergistic effect of rainfall, relative humidity, evapotranspiration, and other impact factors on ecological quality improvement, theoretically, the improvement in ecological quality will be greater than 1%.

[0098] Attribution of spatiotemporal differentiation:

[0099] Some influencing factors in 2015 and 2022 exhibited unique spatial distribution characteristics. In 2015, the RH coefficient and DEM coefficient showed similar high-value distribution areas, which may be related to the decrease in humidity with increasing altitude in plateau regions. Among the influencing factors in 2022, the PET coefficient showed a clustering of extremely low values ​​in the eastern part of Maduo County, with some areas having a PET value of zero, resulting in a very large absolute value for the coefficient. The POP coefficient showed a significant clustering of values ​​from -100 to 0 in the northern and central southern regions of Maduo County, and was also affected by the uninhabited nature of some areas, resulting in a very large absolute value for the coefficient.

[0100] Optimization of ecological quality assessment methods:

[0101] The method in this embodiment optimizes the ecological quality assessment method in two main aspects: firstly, it enables direct comparison of the relative ecological quality values ​​obtained from multiple periods of PCA; secondly, it optimizes the delineation of assessment units to reduce the spatial heterogeneity of ecological quality within units.

[0102] In constructing an ecological quality assessment system using multiple indicators, this study employs PCA (Potentially Constructing Analysis) to objectively represent the characteristics of the comprehensive indicators to the greatest extent possible. Unlike conventional studies that perform PCA independently on data from each period, this study performs unified PCA on a collection of multi-period indicator data, enabling precise comparisons between ecological quality data across different periods. Existing research indicates that traditional PCA methods, due to the dynamic weighting of normalized indicators, result in relative values ​​for ecological quality assessments across different periods. Their comparability is constrained by differences in the initial indicator value ranges for each period and the variation in PCA weight allocation with input data. To address these limitations, this study proposes an improved method: after unified normalization of multi-period datasets, unified PCA weight calculation is performed, followed by normalization of the results and data segmentation and grading according to time sequence. This method effectively solves the problem of inconsistent benchmarks in cross-period ecological quality index (EQI) comparisons, ensuring that EQI values ​​across different periods truly reflect the actual evolutionary trend of ecological quality.

[0103] Secondly, in terms of the spatial carrier of ecological quality assessment, the ecological quality assessment results are correlated with vectorized ecological patches based on 2m resolution image recognition, enabling a precise understanding of the ecological quality of every piece of land in Maduo County, the source of the Yellow River. Compared to the assessment units using fishing nets or grids in most studies, this method has the advantages of accurate and clear boundaries and easy correspondence with geographical entities. Ecological land use type in nature reserves is one of the key factors in ecological quality. Dividing assessment units based on ecological land use type and land feature boundaries, compared to artificially delineating idealized grids, can greatly reduce spatial heterogeneity interference within assessment units and improve the accuracy of ecological quality assessment.

[0104] Refined management of ecological quality:

[0105] The more refined spatiotemporal results of ecological quality obtained by the proposed scheme in this embodiment are of great value to policymakers and authorities in implementing more targeted ecological restoration efforts.

[0106] Ecological quality is a crucial foundation for ecological restoration and management. Ecological restoration and optimization measures should comprehensively consider ecological priorities, protection needs, and natural restoration methods, developing tailored strategies for each region. Current regional strategies suffer from a lack of diversity and a static structure, necessitating the exploration of innovative zoning methods. This embodiment studies the patterns of ecological quality change based on ecological patch units. It can concentrate resources on the refined management of land units with significant ecological quality deterioration and important value within a large spatial area of ​​different protection levels. Achieving high management accuracy at a relatively low cost using remote sensing imagery effectively improves regulatory efficiency and saves on the cost of implementing ecological measures. It can also be applied to assessing the effectiveness of ecological quality management, helping to assign specific regulatory responsibilities to local areas. Continuous ecological quality assessments can also more promptly identify and predict trends, promoting sustainable regional development.

[0107] In the specific refined management of ecological quality, the ecological quality dataset results of individual vector map patches can be further analyzed to obtain the annual changes in ecological quality and more effective improvement methods. This mainly consists of three steps: first, identifying map patches that urgently need quality improvement through changes in ecological quality levels; second, determining the importance of ecological protection by combining the watershed human settlement system with the delineation of protected areas; and finally, understanding the reasons for changes in the ecological quality of each land parcel based on regression coefficients, screening map patches based on the adjustability of dominant influencing factors, and classifying them according to dominant variable factors, thereby determining the priority and specific measures for ecological patch protection, truly achieving precise and site-specific implementation.

[0108] In summary, this embodiment utilizes the Google Earth Engine (GEE) platform and innovatively improves the ecological quality assessment method using EQI as the evaluation index by optimizing the application process of principal component analysis based on ecological patches. It supports refined analysis of long-term dynamic ecological quality evolution. This study employs least squares, geographic weighted regression, and multi-scale geographic weighted regression models to explore the importance and effects of potential influencing factors on ecological quality, both globally and precisely at the patch unit level. The research results provide a scientific analysis and theoretical basis for the rational management of ecological quality in Maduo County, the source region of the Yellow River, and have value for regional ecological restoration and resource protection. Specific conclusions are as follows.

[0109] (1) The overall ecological quality of Madoi County improved in 2022 compared to 2015. In terms of spatial distribution, the ecological quality in the south showed a greater upward trend, while the northwest and northeast showed regional clustering characteristics of declining ecological quality.

[0110] (2) The MGWR method was used to analyze the importance and local spatial regression coefficients of factors influencing ecological quality, and the results were better than those obtained by the OLS and GWR methods. Climate factors and human activities were the dominant factors in the changes in ecological quality in Maduo County from 2015 to 2022. Overall, MAP, POP, and LDU had the greatest impact on the ecological quality of Maduo County. The spatial distribution of the regression coefficients of each influencing factor was different, and the regression coefficients of the influencing factors for each map unit could be obtained.

[0111] (3) The study contributes to improving the accuracy of ecological quality assessment and management. Based on the assessment of ecological quality changes and the analysis of influencing factors, the study has the advantages of highlighting key points, low cost and sustainability in the formulation and implementation of ecological restoration strategies in Madoi County, and facilitates the targeted focus of resources on key areas that urgently need protection.

[0112] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0113] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for spatio-temporal assessment of ecological quality of a river basin human settlement system, characterized in that, The method comprises the following steps: acquiring target region data of a typical human settlement system in the Yellow River source area; the target region data is provided by a MODIS data set and a LANDSAT data set; calculating an ecological quality index EQI based on the target region data; using a 2m resolution polygon based on a real ecological boundary to proxy a large-scale grid as a target region evaluation unit, and integrating the ecological quality index EQI and collected potential influencing factor data into the target region evaluation unit to generate an ecological polygon unit; based on the ecological polygon unit, using a Moran's I index and an ecological quality change matrix to determine an ecological quality spatio-temporal evaluation result; the ecological quality spatio-temporal evaluation result is used to describe the spatial autocorrelation and spatio-temporal change of ecological quality; the calculation of the ecological quality index EQI based on the target region data specifically comprises: based on the MODIS data set, using MCD15A3H to acquire a leaf area index LAI, and using MOD17A2H to acquire a gross primary productivity GPP; using TI_L2 in the LANDSAT series satellite data to acquire an NDVI value, and calculating a vegetation coverage FVC; constructing initial index data according to the leaf area index LAI, the gross primary productivity GPP, and the vegetation coverage FVC; The same initial index data of different years are normalized in the 0-1 interval, and the dimension and weight of the same index of different years are unified by using a PCA analysis method to form an initial ecological quality index EQI c , which is expressed as: ， wherein a , b , c are weight coefficients; PC n are new explanatory variables formed during the PCA analysis, n e {1,2,3}; For the initial ecological quality index EQI c Normalization is performed to obtain the final ecological quality index EQI; the use of a 2m resolution polygon based on a real ecological boundary to proxy a large-scale grid as a target region evaluation unit, and the integration of the ecological quality index EQI and collected potential influencing factor data into the target region evaluation unit to generate an ecological polygon unit specifically comprises: selecting potential influencing factor data of ecological quality based on the target region data; firstly, resampling all grid images to 2m, then giving each ecological polygon a unique ID, integrating the ecological quality index EQI and the potential influencing factor data in each ecological polygon according to the spatial position, and calculating an influencing factor mean value based on the potential influencing factor data, and generating an ecological polygon unit according to the ecological quality index EQI and the influencing factor mean value; wherein the ecological quality index EQI in each polygon is divided into five levels; each level is: setting 0.8-1 as the best, setting 0.6-0.8 as good, setting 0.4-0.6 as medium, setting 0.2-0.4 as poor, and setting 0-0.2 as the worst; the use of a Moran's I index and an ecological quality change matrix to determine an ecological quality spatio-temporal evaluation result based on the ecological polygon unit specifically comprises: based on the ecological polygon unit, firstly, using global Moran's I to analyze the spatial autocorrelation of the overall ecological quality of the target region, and judging the spatial autocorrelation pattern by using a Moran's I index, a z-score, and a p value, and secondly, using local Moran's I to analyze the local spatial ecological quality correlation and heterogeneity of each spatial unit to obtain a LISA clustering map; constructing an ecological quality change matrix according to the ecological quality levels in the ecological polygon unit; the ecological quality change matrix is the corresponding area of the ecological quality level conversion.

2. The method according to claim 1, wherein, Also comprising: determining the main influencing factors of the ecological quality according to a plurality of regression methods and the spatio-temporal evaluation result of the ecological quality; the main influencing factors of the ecological quality are used to provide a reference for resource allocation for ecological protection; the plurality of regression methods include OLS, GWR and MGWR.

3. A system for spatio-temporal assessment of ecological quality of a catchment human settlement system, applying the method according to any one of claims 1-2, characterized in that, Comprising: a data acquisition unit configured to acquire target region data of the Yellow River source region; the target region data is provided by a MODIS data set and a LANDSAT data set; an index construction unit configured to calculate an ecological quality index EQI based on the target region data; an ecological map patch construction unit configured to use a 2m resolution patch based on a real ecological boundary to represent a large-scale grid as an evaluation unit, and integrate the ecological quality index EQI and potential influencing factor data into the ecological map patch unit; an evaluation unit configured to determine a spatio-temporal evaluation result of the ecological quality based on the ecological map patch unit, using a Moran's I index and an ecological quality change matrix; the spatio-temporal evaluation result of the ecological quality is used to describe the spatial autocorrelation and spatio-temporal change of the ecological quality.

4. An electronic device, comprising: An electronic device comprising a memory and a processor, the memory being configured to store a computer program, and the processor being configured to execute the computer program to cause the electronic device to perform the spatio-temporal evaluation method of the ecological quality of the catchment area human settlement system according to any one of claims 1-2.

5. A computer readable storage medium, characterized in that, The computer program is stored in the memory and executed by the processor to implement the spatio-temporal evaluation method of the ecological quality of the catchment area human settlement system according to any one of claims 1-2.

Images