Regional ecological geological environment quantitative coupling evaluation method and system

Abstract

The invention relates to the technical field of ecological geological environment treatment, and discloses a regional ecological geological environment quantitative coupling evaluation method, which comprises the following steps of S1, constructing a multi-source data set; s2, preprocessing the multi-source data set, eliminating errors and unifying formats; s3, constructing a regional comprehensive ecological system service evaluation model; s4, constructing a regional ecosystem vulnerability evaluation model; s5, constructing a regional geological environment sensitivity evaluation model; s6, constructing a regional ecological geological environment quantitative coupling evaluation model to obtain an ecological geological environment safety index; and S7, performing driving mechanism analysis and clustering partitioning, performing multi-scale driving mechanism analysis on the ecological geological environment safety index, and performing clustering partitioning based on an unsupervised neural network to obtain an evaluation result. According to the method, an analysis framework of a regional ecological geological environment system pattern is disclosed from different angles, and the requirement for sustainable development of a macroscopic ecological geological environment in actual planning application of a large urban agglomeration is met.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological geological environment governance, and particularly to a method and system for quantitatively coupling the assessment of regional ecological geological environment. Background Art

[0002] The research of ecological geology involves the interaction between the multi-layer structures of the Earth's surface and the mutual feedback change mechanism under the interference of natural environment and human activities. The various elements of the ecological geological environment contained therein are organically linked to each other and maintain a dynamic balance while maintaining their own resilience. However, with the change of climate and the development of human society, this balance process has been broken, bringing a series of ecological geological environment problems. At present, most of the regional ecological geological environment assessment studies are limited to the assessment of the external vegetation conditions in ecological quality analysis or the analysis of geological environment stability, and there is still a gap from the assessment requirements of the macroscopic regional ecological geological environment. In addition, it is not clear how the assessment results can be used to guide actual implementation.

[0003] Aiming at the above problems, the deficiencies of the existing technologies are mainly manifested in two aspects: one is the lack of an effective framework to realize the coupled quantitative assessment of the regional ecological geological environment; the other is that the existing assessment results mainly focus on the assessment of ecological security and cannot comprehensively and accurately reflect the ecological geological environment problems, especially in supporting the sustainable development and planning of regional macroscopic urban agglomerations. Summary of the Invention

[0004] In order to overcome or alleviate one or more of the above technical problems, the object of the present invention is to provide a method and system for quantitatively coupling the assessment of regional ecological geological environment based on ecosystem service function - ecosystem vulnerability - geological environment sensitivity, and to propose an analysis framework for revealing the pattern of the regional ecological geological environment system from different perspectives to meet the needs of sustainable development of the macroscopic ecological geological environment in the actual planning application of large urban agglomerations.

[0005] The present invention provides the following technical solutions:

[0006] In the first aspect, the present invention provides a method for quantitatively coupling the assessment of regional ecological geological environment, including the following steps:

[0007] S1: Construct a multi-source data set, collect multi-source and multi-type heterogeneous data, and adjust the spatio-temporal resolution and projection mode of the data to be consistent. The multi-source and multi-type heterogeneous data includes optical remote sensing images, climate and meteorology, topography and geomorphology, and vegetation and soil data;

[0008] S2: Preprocess the multi-source data set to eliminate the errors caused by uncertain pixels and unify the format. The preprocessing methods include de-clouding, de-hazing, histogram matching, median calculation, and spatial interpolation for the multi-source and multi-type heterogeneous data;

[0009] S3: Construct a regional comprehensive ecosystem service assessment model for quantitatively coupling the functions of multiple key ecosystems to achieve comprehensive analysis of regional ecosystem services;

[0010] S4: Construct a regional ecosystem vulnerability assessment model for comprehensively measuring the external structure and internal quality of the ecosystem to achieve a comprehensive quantitative characterization of the vulnerability of natural ecosystems;

[0011] S5: Construct a regional geological environment sensitivity evaluation model to comprehensively analyze the geological background affecting the regional eco-geological environment;

[0012] S6: According to the regional comprehensive ecosystem service assessment model, the regional ecosystem vulnerability assessment model, and the regional geological environment sensitivity evaluation model, construct a regional eco-geological environment quantitative coupling assessment model through the equal-weight summation method to obtain the eco-geological environment security index;

[0013] S7: Driving mechanism analysis and clustering partition, conduct multi-scale driving mechanism analysis on the eco-geological environment security index and perform clustering partition based on unsupervised neural network to obtain the evaluation results.

[0014] According to some embodiments, the S1 specifically includes:

[0015] S11: Use long-term optical remote sensing satellites to obtain optical image data of the target area through multi-spectral sensors, and record the acquisition time and information on the number of all available images;

[0016] S12: Synchronously collect climate and meteorological data such as precipitation and temperature during the same period for resampling processing, and adjust the data products to be consistent with the spatial resolution and geographical coordinates of the optical images;

[0017] S13: Collect digital elevation model products of the study area and calculate topographic and geomorphic parameters based on the digital elevation model products for resampling processing. The topographic and geomorphic parameters include topographic slope, aspect, and topographic relief, and adjust the data products to be consistent with the spatial resolution and geographical coordinates of the optical images;

[0018] S14: Conduct spatial interpolation and resampling processing on the biophysical parameters of vegetation and soil, and adjust the data products to be consistent with the spatial resolution and geographical coordinates of the optical images;

[0019] S15: Perform spatial registration on all optical images and their derived remote sensing data products, use geographical reference points for precise spatial alignment, and adjust the spatial consistency of images from different data sources;

[0020] S16: Convert the spatially registered optical images and images of each remote sensing data product into a unified standardized data format.

[0021] According to some embodiments, S2 specifically includes:

[0022] S21: Perform cloud and fog removal on the selected time-series optical remote sensing images, and use the pixel quality control band to eliminate unstable pixels;

[0023] S22: Perform linear interpolation on the time-series optical images after cloud and fog removal to fill in some missing pixels caused by cloud removal, and ensure the spatial continuity of the optical images;

[0024] S23: Perform geometric precise correction on the optical images after image cloud removal and missing pixel interpolation. Adopt a geometric correction algorithm based on georeference points to ensure the spatial consistency of images at different times;

[0025] S24: Perform coordinate transformation on the geometrically corrected images. Adopt a unified geographic coordinate system to ensure the precise alignment of multi-source data in space. Subsequently, use the histogram matching method to process the time-series images.

[0026] According to some embodiments, S3 specifically includes:

[0027] S31: Construct a framework for the regional comprehensive ecosystem service assessment model to separately construct service evaluation models and quantitatively evaluate the services of the key ecosystems respectively. The key ecosystems include carbon storage, habitat quality, soil conservation, and water yield;

[0028] S32: Input vegetation and soil-related physiological parameters to perform time-series assessment on the services of the key ecosystems;

[0029] S33: Perform maximum-minimum normalization on the services of the key ecosystems at the same time phase to eliminate the dimension difference, and use the equal-weight summation method for quantitative coupling to construct a regional comprehensive ecosystem service index.

[0030] According to some embodiments, S4 specifically includes:

[0031] S41: Construct a framework for the regional ecosystem vulnerability assessment model to comprehensively quantify the external structure and internal quality of the ecosystem;

[0032] S42: Divide the study area into several square grid units by using the equal-spacing sampling method, and then calculate the landscape ecological indicators for each grid unit and conduct regional landscape structure vulnerability evaluation through spatial interpolation;

[0033] S43: Construct a landscape ecological risk index LERI based on the landscape disturbance degree E i and the landscape sensitivity F i The calculation formula is:

[0034]

[0035] Among them, A ki is the area of landscape type i in the kth evaluation unit; A k is the area of the kth evaluation unit;

[0036] S44: Internal ecological quality vulnerability assessment is carried out through the ecological quality vulnerability index, which is calculated as follows:

[0037] EQVI=1-f(Greenness, Wetness, Dryness, Heat) (2)

[0038] Among them, Greenness represents indicators related to surface vegetation biomass and coverage, and is represented by the Normalized Difference Vegetation Index (NDVI); Wetness is used to reflect the moisture conditions of vegetation and soil, and is represented by the moisture component enhanced by the tasseled cap transformation; Dryness represents the dryness index, which is comprehensively reflected by the building index and bare soil index; Heat can be reflected by the surface temperature;

[0039] S45: The landscape ecological risk index reflecting the external structure of the ecosystem and the ecological quality vulnerability index reflecting the internal quality are added with equal weights. The sum reflects the comprehensive vulnerability of the regional ecosystem.

[0040] According to some implementation modes, S5 specifically includes:

[0041] S51: Based on the geological environment background, external environment induction and disaster occurrence results, a multi-level indicator system framework for regional geological environment sensitivity assessment model is constructed;

[0042] S52: Using an integrated machine learning method based on the Stacking strategy, taking historical geological disaster occurrence points, an equal number of non-disaster points and multi-level indicators as input, a grid search method was used to train and adjust the optimal parameters of the regional geological environment sensitivity assessment model, where the training data set accounted for 70% and the validation data set accounted for 30%, and the optimal model after training was obtained;

[0043] S53: Use the best performing model for geological environment sensitivity prediction and output the regional geological environment sensitivity index evaluation results;

[0044] S54: Use interpretable machine learning to analyze the optimal model and output the global and local impact results of different influencing indicators on the sensitivity of the geological environment;

[0045] According to some implementations, S61: construct a regional ecological geological environment quantitative coupling assessment model, and its coupling method is shown in the following formula:

[0046]

[0047] Among them, ESI is the comprehensive ecosystem service index; CEVI is the ecosystem vulnerability index; GESI is the geological environment sensitivity index. By adding these three indices with equal weights, the regional eco-geological environment security index EGESI can be obtained.

[0048] S62: Conduct trend analysis on the regional eco-geological environment security index of multiple time phases, and conduct comparative analysis with the actual regional situation to test the effectiveness of the eco-geological environment security index.

[0049] According to some embodiments, the specific steps of S7 include:

[0050] S71: Using multi-source driving factors and the eco-geological environment security index as inputs, and adopting the optimal parameter geographical detector to identify the dominant influencing factors and conduct interactive influence effect analysis. The multi-source driving factors include annual average temperature, annual precipitation, terrain relief degree, vegetation health index, Shannon diversity index, and per capita GDP.

[0051] S72: Introduce the structural equation model into the research of eco-geology, construct and test the conceptual model of causal changes in the eco-geological environment, and conduct causal relationship analysis on the eco-geological environment pattern involving multiple factors.

[0052] S73: Use multi-scale geographically weighted regression to further detect the spatial heterogeneity relationship between multiple influencing factors and the eco-geological environment security index, give the distribution pattern of heterogeneity in space and the multi-scale effects of different influencing factors.

[0053] S74: Mine the eco-geological environment security index of multiple time series, and use the clustering quality index (CQI) to analyze at two scales of pixel and county level respectively to determine the optimal number of clusters, as shown in the following formula:

[0054]

[0055] Among them, N is the number of input samples, k is the final number of clusters, p is the number of eco-geological environment characteristics to be analyzed, n and CV i j are respectively the number of samples in category i and its coefficient of variation in feature j, k max represents the maximum number of clusters under the classification scheme, and δ is a constant used to adjust the order of magnitude of the number of categories;

[0056] S75: Based on the determined optimal number of clusters, use the self-organizing mapping neural network to conduct unsupervised clustering partition of the eco-geological environment pattern to obtain the evaluation result.

[0057] In a second aspect, the present invention further provides a regional ecological geological environment quantitative coupling evaluation system, which includes the following modules:

[0058] A multi-source data set construction module, which is used to collect multi-source heterogeneous data, and adjust the spatio-temporal resolution and projection mode of the data to be consistent. The multi-source heterogeneous data includes optical remote sensing images, climate and meteorology, topography and geomorphology, and vegetation and soil data;

[0059] A data preprocessing module, which is used to preprocess the multi-source data set, eliminate the errors caused by uncertain pixels and further unify the format. The preprocessing methods include de-clouding and de-hazing, histogram matching, median calculation, and spatial interpolation of the multi-source heterogeneous data;

[0060] A module for constructing a regional comprehensive ecosystem service evaluation model, which is used to quantitatively couple the functions of multiple key ecosystems and realize the comprehensive analysis of regional ecosystem services;

[0061] A module for constructing a regional ecosystem vulnerability evaluation model, which is used to comprehensively measure the external structure and internal quality of the ecosystem and realize the comprehensive quantitative characterization of the vulnerability of the natural ecosystem;

[0062] A module for constructing a regional geological environment sensitivity evaluation model, which is used to comprehensively analyze the geological background affecting the regional ecological geological environment;

[0063] A module for constructing a regional ecological geological environment quantitative coupling evaluation model, which is used to construct the regional ecological geological environment quantitative coupling evaluation model by the equal-weight summation method according to the regional comprehensive ecosystem service evaluation model, the regional ecosystem vulnerability evaluation model, and the regional geological environment sensitivity evaluation model, and obtain the ecological geological environment security index;

[0064] A driving mechanism analysis and clustering partition module, which is used to perform multi-scale driving mechanism analysis on the ecological geological environment security index and perform clustering partition based on unsupervised neural network to obtain the evaluation result.

[0065] Compared with the prior art, the present invention has the following beneficial effects:

[0066] The regional ecological geological environment quantitative coupling evaluation method and system based on ecosystem services-ecosystem vulnerability and geological environment sensitivity provided by the present invention ensure the consistency and high-quality basis of data by collecting and calibrating multi-source remote sensing data and products; by preprocessing the calibrated multi-source heterogeneous data, the image data errors and noises are effectively suppressed, the reliability and accuracy of the data are improved, and the problems of incomplete and inaccurate reflection of the ecological geological environment in the prior art are solved.

[0067] The present invention realizes a comprehensive quantitative assessment of the eco-geological environment by constructing a regional comprehensive ecosystem service assessment model, a regional ecosystem vulnerability assessment model, and a regional geological environment sensitivity evaluation model. On this basis, the evaluation results of the sub-models are quantitatively coupled to construct a regional eco-geological environment quantitative coupling assessment model, and an eco-geological environment security index is developed. Through the fusion of multi-source data and quantitative coupling methods, the quantitative analysis ability of regional eco-geological environment assessment and the analytical ability of multi-factor causal associations are significantly improved.

[0068] The present invention ensures the high reliability and accuracy of the evaluation results in practical applications by analyzing the multi-scale driving mechanisms of the time-series eco-geological environment security index and combining the neural network unsupervised clustering partition method. By using multi-index quality assessment and actual verification, the effects and performance of the assessment model are comprehensively detected, providing strong support for high-precision applications in the fields of eco-geological environment protection and regional planning of large urban agglomerations in macro regions. Overall, the method of the present invention significantly improves the quantification, scientificity, and practicality of regional eco-geological environment assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is a flowchart of the regional eco-geological environment quantitative coupling assessment method provided by an embodiment of the present invention.

[0070] Figure 2 It is a structural diagram of the regional eco-geological environment quantitative coupling assessment system provided by an embodiment of the present invention.

[0071] Figure 3 It is a schematic diagram of the regional eco-geological environment coupling analysis of Beijing-Tianjin-Hebei provided by an embodiment of the present invention.

[0072] Figure 4 It is the regional comprehensive ecosystem service assessment process of Beijing-Tianjin-Hebei provided by an embodiment of the present invention.

[0073] Figure 5 It is the regional ecosystem vulnerability assessment process of Beijing-Tianjin-Hebei provided by an embodiment of the present invention.

[0074] Figure 6 It is the regional geological environment sensitivity assessment process of Beijing-Tianjin-Hebei provided by an embodiment of the present invention.

[0075] Figure 7 It is the regional eco-geological environment unsupervised clustering partition result of Beijing-Tianjin-Hebei provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0076] The present invention will be described in detail below in conjunction with embodiments and the accompanying drawings. However, it should be understood that the embodiments and the drawings are only used for exemplary description of the present invention, and do not constitute any limitation to the protection scope of the present invention. All reasonable transformations and combinations within the scope of the inventive concept of the present invention fall within the protection scope of the present invention.

[0077] The present invention will be further described below in conjunction with the accompanying drawings.

[0078] Embodiment 1

[0079] As Figure 1 , for the regional ecological geological environment quantitative coupling assessment method, this embodiment provides the whole process of quantitative coupling analysis of the regional ecological geological environment in Beijing-Tianjin-Hebei. The specific step flow is as Figure 3 , including the following main steps:

[0080] S1: Construct a multi-source data set, collect multi-source and heterogeneous data, including optical remote sensing images, climate and meteorology, topography and geomorphology, vegetation and soil and other physiological parameters with the Beijing-Tianjin-Hebei region as the target area, ensure the consistency of the spatio-temporal resolution and projection method of the data, and create a multi-source data set;

[0081] S2: Preprocess the multi-source data set, including image cloud and haze removal, histogram matching, median calculation, spatial interpolation and other processes to eliminate errors caused by uncertain pixels and further unify;

[0082] S3: Construct a regional comprehensive ecosystem service assessment model for quantitatively coupling multiple key ecosystem functions, so as to realize the comprehensive analysis of regional ecosystem services;

[0083] S4: Construct a regional ecosystem vulnerability assessment model for comprehensively measuring the external structure and internal quality of the ecosystem, and realizing the comprehensive quantitative characterization of the vulnerability of the natural ecosystem;

[0084] S5: Construct a regional geological environment sensitivity evaluation model to comprehensively analyze the geological background affecting the regional ecological geological environment;

[0085] S6: On the basis of S3, S4, and S5, construct a regional ecological geological environment quantitative coupling assessment model and propose an ecological geological environment security index to improve the time-space consistency of the assessment;

[0086] S7: Driving mechanism analysis and clustering partition, conduct multi-scale driving mechanism analysis on the temporal ecological geological environment security index after S6 processing and perform clustering partition based on unsupervised neural network to ensure the practical application effect of the assessment results.

[0087] More specifically, the multi-source and heterogeneous data collected in S1 specifically includes:

[0088] S11: Use a large number of long-time optical remote sensing satellite records to obtain optical image data of the target area through a multispectral sensor, and record the acquisition time and information on the number of all available images.

[0089] S12: Synchronously collect climate and meteorological data such as precipitation and temperature during the same period for resampling processing to ensure that the data products are consistent with the optical images in terms of spatial resolution and geographical coordinates.

[0090] S13: Collect digital elevation model products of the study area and calculate topographic and geomorphic parameters such as terrain slope, aspect, and terrain relief based on these products for resampling processing to ensure that the data products are consistent with the optical images in terms of spatial resolution and geographical coordinates.

[0091] S14: Perform spatial interpolation and resampling processing on biophysical parameters such as vegetation and soil to ensure that the data products are consistent with the optical images in terms of spatial resolution and geographical coordinates.

[0092] S15: Perform spatial registration on all optical images and their derived remote sensing data products, and use georeference points for precise spatial alignment to ensure strict spatial consistency of images from different data sources.

[0093] S16: Convert the spatially registered optical images and remote sensing data product images into a unified standardized data format and store them in the GeoTIFF format; through the above sub-steps, during the process of collecting and calibrating multi-source element data related to the ecological and geological environment, the temporal consistency of the data can be effectively ensured. By removing unreliable pixels and interpolating and filling missing values, the consistency and continuity of the data are significantly improved. Such a high-precision dataset provides a solid foundation for subsequent ecological system and geological environment sensitivity analysis, greatly enhancing the application value of the data, especially in the field of quantitative analysis of the ecological and geological environment.

[0094] S2 Preprocess the multi-source, multi-element, and heterogeneous dataset, specifically including:

[0095] S21: Remove clouds and fog from the selected time-series optical remote sensing images, and use the pixel quality control band to eliminate unstable pixels to ensure that only reliable optical images are used.

[0096] S22: Perform linear interpolation on the time-series optical images after cloud and fog removal to fill in some missing pixels caused by cloud removal and ensure the spatial continuity of the optical images.

[0097] S23: Perform geometric precision correction on the optical images after image cloud removal and missing pixel interpolation, and use a geometric correction algorithm based on georeference points to ensure the spatial consistency of images at different times.

[0098] S24: Perform coordinate transformation on the geometrically corrected image, adopt a unified geographic coordinate system to ensure the precise spatial alignment of multi-source data, and then process the time-series images using the histogram matching method; through the above sub-steps, in the preprocessing process of multi-source remote sensing data, the present invention effectively suppresses the errors of remote sensing images through cloud removal, and improves the data availability and reduces the sequence data noise through the time-series linear interpolation algorithm and SG filtering respectively, ensuring the data quality. Geometric correction, resampling, and reprojection techniques further ensure the precise spatial alignment of different images, and histogram matching further improves the consistency level of the quality of time-series images, laying a reliable foundation for subsequent ecological geological environment assessment, and significantly enhancing the application accuracy and reliability of ecological geological environment analysis.

[0099] S3 Construct a regional comprehensive ecosystem service evaluation model, and the technical process is as Figure 4 shown, specifically including:

[0100] S31: Construct the framework of the regional comprehensive ecosystem service evaluation model, which includes multiple sub-parts such as carbon storage, habitat quality, soil conservation, and water yield, etc., and are respectively used to quantify the above different key ecosystem services;

[0101] S32: Input the physiological parameters related to vegetation and soil, and conduct time-series evaluations on the four services of carbon storage, habitat quality, soil conservation, and water yield;

[0102] S33: Perform maximum-minimum normalization processing on the 4 services in the same time phase to eliminate the dimension difference, and use the equal-weight summation method for quantitative coupling to construct a comprehensive ecosystem service index; through the above sub-steps, a comprehensive ecosystem service index considering the key ecosystem service functions of the region is constructed. Through the analysis of the carbon storage, habitat quality, soil conservation, and water yield of natural ecosystems, a comprehensive and overall reflection of the regional ecosystem service functions is realized. This index will significantly improve the accuracy and reliability of the quantification of the regional ecosystem service capacity, providing a solid technical foundation for subsequent application planning.

[0103] S4 Construct a regional ecosystem vulnerability assessment model, and the technical process is as Figure 5 shown, specifically including:

[0104] S41: Construct the framework of the regional ecosystem vulnerability assessment model, which is used to comprehensively quantify the external structure and internal quality of the ecosystem; for example, in the framework, the vulnerability indicators are divided into landscape scale and regional scale. The landscape scale includes landscape vulnerability and landscape disturbance degree, and the regional scale includes greenness, humidity, dryness, and heat degree, etc.

[0105] S42: The study area was divided into several square grid cells using an equidistant sampling method, and then the landscape ecological indicators were calculated for each grid cell and the regional landscape structure vulnerability was evaluated by spatial interpolation;

[0106] S43: Use landscape disturbance level E i and landscape sensitivity F i The landscape ecological risk index LERI is constructed based on the formula:

[0107]

[0108] Among them, A ki is the area of landscape type i in the kth evaluation unit; A k is the area of the kth evaluation unit;

[0109] S44: Internal ecological quality vulnerability assessment is carried out through the ecological quality vulnerability index, which is calculated as follows:

[0110] EQVI=1-f(Greenness, Wetness, Dryness, Heat) (2)

[0111] Among them, Greenness represents indicators related to surface vegetation biomass and coverage, and is represented by the normalized difference vegetation index NDVI; Wetness is used to reflect the moisture conditions of vegetation and soil, and is represented by the moisture component enhanced by the tasseled cap transformation; Dryness represents the dryness index, which is comprehensively reflected by the building index and bare soil index; Heat can be reflected by the surface temperature.

[0112] S45: The landscape ecological risk index reflecting the external structure of the ecosystem and the ecological quality vulnerability index reflecting the internal quality are equally weighted and the sum is used to reflect the comprehensive vulnerability of the regional ecosystem;

[0113] Through the above sub-steps, in the natural ecosystem vulnerability assessment, an algorithm that combines the external structure and internal quality of the ecosystem is adopted. Through evaluation indicators such as landscape pattern vulnerability and ecological quality vulnerability, the internal and external vulnerability indexes of the natural ecosystem are normalized and summed with equal weights. This method effectively optimizes the shortcomings of the current ecosystem vulnerability assessment that only considers the external structure, ensures the accuracy and representativeness of the evaluation results, and improves the application effect in ecological geological environment analysis.

[0114] S5 builds a regional geological environment sensitivity assessment model. The technical process is as follows Figure 6 As shown, specifically including:

[0115] S51: Construct the index system framework of the regional geological environment sensitivity evaluation model. Based on the geological environment background, external environment induction, and disaster occurrence results, construct a multi-level index system framework of the regional geological environment sensitivity evaluation model, as shown in Table 1 below:

[0116] Table 1 Index System Framework of Geological Environment Sensitivity Evaluation Model

[0117]

[0118] S52: Adopt an integrated machine learning method based on the Stacking strategy. Use the historical geological disaster occurrence points, the same number of non-disaster points, and multi-level indicators as inputs, and use the grid search method to train and adjust the optimal parameters of the regional geological environment sensitivity evaluation model. Among them, the training data set accounts for 70%, and the validation data set accounts for 30% to obtain the optimal model after training.

[0119] S53: Use the optimal model to predict the geological environment sensitivity and output the evaluation results of the regional geological environment sensitivity index;

[0120] S54: Use interpretable machine learning to analyze the optimal model and output the global and local influence results of different influencing indicators on the geological environment sensitivity;

[0121] Through the above sub-steps, the construction process of the regional geological environment sensitivity evaluation model is described. Through integrated machine learning for quantitative evaluation of geological environment sensitivity, it is ensured that the evaluation model can integrate the advantages of different algorithms and improve the reliability of the results. Finally, through the interpretable machine learning method, the global and local interpretations of the integrated model are carried out, further revealing the differential driving mechanism behind the regional geological environment sensitivity and improving the understanding of geological environment sensitivity.

[0122] S6 Construct a quantitative coupling evaluation model of the regional ecological geological environment, specifically including:

[0123] S61: Construct the framework of the quantitative coupling evaluation model of the regional ecological geological environment. The coupling method in the framework is shown in the following formula:

[0124]

[0125] Among them, ESI is the comprehensive ecosystem service index; GEVI is the ecosystem vulnerability index; GESI is the geological environment sensitivity index. The three are added with equal weights to obtain the regional ecological geological environment security index EGESI;

[0126] S62: Conduct a trend analysis on the multi-temporal ecological geological environment security index and compare it with the actual situation of the region to test the effect of the proposed ecological geological environment security index to ensure its actual implementation and application;

[0127] Based on the evaluation results of different aspects of the regional eco-geological environment system through the above sub-steps, the quantitative coupling of the eco-geological environment can be realized to reflect the change degree and trend of the eco-geological environment. Furthermore, the regional eco-geological environment security index constructed accordingly also provides important scientific support for the long-time series and large-scale quantitative characterization of the regional eco-geological environment.

[0128] Analysis of the S7 driving mechanism and zoning clustering, specifically including:

[0129] S71: Using multi-source driving factors such as annual average temperature, annual precipitation, terrain relief degree, vegetation health index, Shannon diversity index, and per capita GDP, as well as the eco-geological environment security index as inputs, the optimal parameter geographical detector is used to identify the dominant influencing factors and conduct interactive influence effect analysis. The optimal parameter geographical detector is a set of parameter values that can make the model achieve the best performance or optimal effect found through certain algorithms and strategies during the solution process, and gives the influence degree of different factors on the regional eco-geological environment.

[0130] S72: Introduce the structural equation model used to reveal the causal change relationship between multi-factor variables into the research of eco-geology, construct and test the conceptual model of the causal change of the eco-geological environment, and conduct causal relationship analysis on the eco-geological environment pattern involving multiple factors; the structural equation model is an existing technology, which is mainly used to analyze and reveal the causal change relationship between multi-factor variables.

[0131] S73: Use multi-scale geographically weighted regression to further detect the spatial heterogeneity relationship between multiple influencing factors (independent variables) and the eco-geological environment security index (dependent variable), give the distribution pattern of heterogeneity in space and the multi-scale effects of different influencing factors;

[0132] S74: Mine the eco-geological environment security index of multi-time series, and use the clustering quality index (CQI) to analyze at two scales of pixel and county level respectively to determine the optimal number of clusters, as shown in the following formula:

[0133]

[0134] Among them, N is the number of input samples, k is the final number of clusters, p is the number of eco-geological environment characteristics to be analyzed, n and CV i j are the number of samples in category i and its coefficient of variation in feature j respectively. k max represents the maximum number of clusters under the classification scheme, and δ is a constant used to adjust the order of magnitude of the number of categories;

[0135] S75: Based on the determined optimal number of clusters, use the self-organizing mapping neural network for unsupervised clustering partitioning of the eco-geological environment pattern, improve the scientific nature of the partitioning and the practicality of the evaluation results, and obtain the final evaluation results. The eco-geological environment partitioning results of the Beijing-Tianjin-Hebei region are as Figure 7 shown.

[0136] Through the above sub-steps, the dominant factors of the eco-geological environment evolution are identified by the optimal parameter geographical detector, and the causal relationship analysis under multi-factor association is carried out using the structural equation model, which deepens the accurate understanding of the regional eco-geological environment evolution. The spatial heterogeneity relationship between the regional eco-geological environment status and various driving factors is intuitively shown through multi-scale geographically weighted regression, significantly improving the efficiency of eco-geological environment governance. Finally, the neural network unsupervised clustering is further used to automatically partition the regional eco-geological environment, and the final evaluation results are obtained, providing a practical application reference for the sustainable development of the eco-geological environment.

[0137] Example 2

[0138] As Figure 2 shown, this example provides a regional eco-geological environment quantitative coupling evaluation system, which can implement the regional eco-geological environment quantitative coupling evaluation method in Example 1. The evaluation system includes the following modules:

[0139] Construct a multi-source dataset module, which is used to collect multi-source and multi-type heterogeneous data, and adjust the spatio-temporal resolution and projection method of the data to be consistent. The multi-source and multi-type heterogeneous data includes optical remote sensing images, climate meteorology, topography and geomorphology, and vegetation and soil data;

[0140] A data preprocessing module, which is used to preprocess the multi-source dataset, eliminate the errors brought by uncertain pixels and further unify the format. The preprocessing methods include de-clouding and de-hazing, histogram matching, median calculation, and spatial interpolation of the multi-source and multi-type heterogeneous data;

[0141] Construct a regional comprehensive ecosystem service evaluation model module, which is used to quantitatively couple the functions of multiple key ecosystems and realize the comprehensive analysis of regional ecosystem services;

[0142] Construct a regional ecosystem vulnerability evaluation model module, which is used to comprehensively measure the external structure and internal quality of the ecosystem and realize the comprehensive quantitative characterization of the vulnerability of the natural ecosystem;

[0143] Construct a regional geological environment sensitivity evaluation model module, which is used to comprehensively analyze the geological background affecting the regional eco-geological environment;

[0144] Construct a quantitative coupling evaluation model module for the regional eco-geological environment, which is used to construct the quantitative coupling evaluation model for the regional eco-geological environment by the equal-weight summation method according to the regional comprehensive ecosystem service evaluation model, the regional ecosystem vulnerability evaluation model and the regional geological environment sensitivity evaluation model, and obtain the eco-geological environment security index;

[0145] The driving mechanism analysis and clustering partition module is used to conduct multi-scale driving mechanism analysis on the eco-geological environment security index and conduct clustering partition based on unsupervised neural network to obtain the evaluation result.

[0146] The above embodiments are only the preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art, improvements and refinements made without departing from the principle of the present invention should also be regarded as within the protection scope of the present invention.

Claims

1. A quantitative coupling evaluation method for regional ecological geological environment, characterized in that, It includes the following steps: S1: Construct a multi-source dataset, collect multi-source and multi-type heterogeneous data, and adjust the spatio-temporal resolution and projection method of the data to be consistent. The multi-source and multi-type heterogeneous data includes optical remote sensing images, climate meteorology, topography and geomorphology, and vegetation and soil data; S2: Preprocess the multi-source dataset to eliminate the errors caused by uncertain pixels and unify the format. The preprocessing methods include de-clouding and de-hazing, histogram matching, median calculation, and spatial interpolation of the multi-source and multi-type heterogeneous data; S3: Build a regional comprehensive ecosystem service assessment model for quantitatively coupling the functions of multiple key ecosystems and realizing the comprehensive analysis of regional ecosystem services; S4: Build a regional ecosystem vulnerability assessment model for comprehensively measuring the external structure and internal quality of the ecosystem and realizing the comprehensive quantitative characterization of the vulnerability of natural ecosystems; S5: Build a regional geological environment sensitivity evaluation model for comprehensively analyzing the geological background affecting the regional eco-geological environment; S6: According to the regional comprehensive ecosystem service assessment model, the regional ecosystem vulnerability assessment model, and the regional geological environment sensitivity evaluation model, build a regional eco-geological environment quantitative coupling assessment model through the equal-weight summation method to obtain the eco-geological environment security index; S7: Conduct driving mechanism analysis and clustering partition. Conduct multi-scale driving mechanism analysis on the eco-geological environment security index and conduct clustering partition based on unsupervised neural network to obtain the assessment results.

2. The quantitative coupling evaluation method for regional ecological geological environment according to claim 1, characterized in that, The specific content of S1 includes: S11: Use long-term optical remote sensing satellite records to obtain optical image data of the target area through multi-spectral sensors, and record the acquisition time and information on the number of all available images; S12: Synchronously collect climate meteorology data such as precipitation and temperature in the same period for resampling processing, and adjust the spatial resolution and geographic coordinates of the data products to be consistent with those of the optical images; S13: Collect digital elevation model products of the study area and calculate topographic and geomorphic parameters based on the digital elevation model products for resampling processing. The topographic and geomorphic parameters include topographic slope, aspect, and topographic relief, and adjust the spatial resolution and geographic coordinates of the data products to be consistent with those of the optical images; S14: Conduct spatial interpolation and resampling processing on the biophysical parameters of vegetation and soil, and adjust the spatial resolution and geographic coordinates of the data products to be consistent with those of the optical images; S15: Conduct spatial registration on all optical images and their derived remote sensing data products, and use georeference points for precise spatial alignment to adjust the spatial consistency of images from different data sources; S16: Convert the spatially registered optical images and each remote sensing data product image into a unified standardized data format.

3. The quantitative coupling evaluation method for regional ecological geological environment according to claim 2, characterized in that, The specific content of S2 includes: S21: Conduct de-clouding and de-hazing on the selected time-series optical remote sensing images, and use the pixel quality control band to eliminate unstable pixels; S22: Conduct linear interpolation processing on the time-series optical images after de-clouding and de-hazing processing to fill in the missing pixels caused by de-clouding and ensure the spatial continuity of the optical images; S23: Perform geometric correction on the optical image after image cloud removal and missing pixel interpolation, using a geometric correction algorithm based on geographic reference points to ensure the spatial consistency of images at different times; S24: coordinate transformation is performed on the geometrically corrected images, and a unified geographic coordinate system is used to ensure accurate spatial alignment of multi-source data. Subsequently, the time series images are processed using a histogram matching method.

4. The quantitative coupling assessment method for regional eco-geological environment according to claim 3, wherein The S3 specifically includes: S31: A framework for constructing a regional integrated ecosystem service assessment model to construct service assessment models to quantify the services of key ecosystems, including carbon storage, habitat quality, soil conservation, and water production; S32: Input vegetation and soil related physiological parameters to conduct time series assessment of the services of the key ecosystem; S33: The services of the key ecosystems in the same phase are normalized to the maximum and minimum to eliminate dimensional differences, and the equal-weighted summation method is used for quantitative coupling to construct a regional comprehensive ecosystem service index.

5. The quantitative coupling evaluation method for regional ecological geological environment according to claim 4, characterized in that The S4 specifically includes: S41: Construct a framework for regional ecosystem vulnerability assessment models to comprehensively quantify the external structure and internal quality of ecosystems; S42: The study area was divided into several square grid cells using an equidistant sampling method, and then the landscape ecological indicators were calculated for each grid cell and the regional landscape structure vulnerability was evaluated by spatial interpolation; S43: Use the landscape interference degree E i and the landscape sensitivity F i as the basis to construct the landscape ecological risk index LERI, and the calculation formula is: Among them, A ki is the area of landscape type i in the k-th evaluation unit; A k is the area of the k-th evaluation unit; S44: Internal ecological quality vulnerability assessment is carried out through the ecological quality vulnerability index, which is calculated as follows: EQVI=1-f(Greenness, Wetness, Dryness, Heat) (2) Among them, Greenness represents indicators related to surface vegetation biomass and coverage, and is represented by the Normalized Difference Vegetation Index (NDVI); Wetness is used to reflect the moisture conditions of vegetation and soil, and is represented by the moisture component enhanced by the tasseled cap transformation; Dryness represents the dryness index, which is comprehensively reflected by the building index and bare soil index; Heat can be reflected by the surface temperature; S45: The landscape ecological risk index reflecting the external structure of the ecosystem and the ecological quality vulnerability index reflecting the internal quality are added with equal weights. The sum reflects the comprehensive vulnerability of the regional ecosystem.

6. The quantitative coupling evaluation method for regional ecological geological environment according to claim 5, characterized in that The S5 specifically includes: S51: Based on the geological environment background, external environment induction and disaster occurrence results, a multi-level indicator system framework for regional geological environment sensitivity assessment model is constructed; S52: Using an integrated machine learning method based on the Stacking strategy, taking historical geological disaster occurrence points, an equal number of non-disaster points and multi-level indicators as input, a grid search method was used to train and adjust the optimal parameters of the regional geological environment sensitivity assessment model, where the training data set accounted for 70% and the validation data set accounted for 30%, and the optimal model after training was obtained; S53: Use the best performing model for geological environment sensitivity prediction and output the regional geological environment sensitivity index evaluation results; S54: Analyze the optimal model using interpretable machine learning and output the global and local impact results of different impact indicators on the sensitivity of the geological environment.

7. The quantitative coupling evaluation method for regional eco-geological environment according to claim 1, characterized in that The specific steps of S6 are as follows: S61: Construct a quantitative coupling evaluation model for the regional ecological geological environment. The coupling method is shown in the following formula: Among them, ESI is the comprehensive ecosystem service index; CEVI is the ecosystem vulnerability index; GESI is the geological environment sensitivity index. The three are summed with equal weights to obtain the regional ecological geological environment security index EGESI. S62: Conduct a trend analysis on the regional ecological geological environment security index of multiple time phases and compare it with the actual regional situation to test the effectiveness of the ecological geological environment security index.

8. The quantitative coupling evaluation method for regional eco-geological environment according to claim 1, characterized in that The specific steps of S7 are as follows: S71: Using multi-source driving factors and the ecological geological environment security index as inputs, identify the dominant influencing factors using the optimal parameter geographical detector and conduct an interactive influence effect analysis. The multi-source driving factors include annual average temperature, annual precipitation, terrain relief degree, vegetation health index, Shannon diversity index, and per capita GDP. S72: Introduce the structural equation model into the research of ecological geology, construct and test the conceptual model of causal changes in the ecological geological environment, and conduct a causal relationship analysis on the ecological geological environment pattern involving multiple factors. S73: Use multi-scale geographically weighted regression to further detect the spatial heterogeneity relationship between multiple influencing factors and the ecological geological environment security index, give the distribution pattern of heterogeneity in space and the multi-scale effects of different influencing factors. S74: Mine the ecological geological environment security index of multiple time series, and use the clustering quality index (CQI) to analyze at two scales of pixel and county level respectively to determine the optimal number of clusters, as shown in the following formula: Among them, N is the number of input samples, k is the number of final clusters, p is the number of eco-geological environment characteristics to be analyzed, n and CV i j are the number of samples in class i and its coefficient of variation in feature j, respectively, k max represents the maximum number of clusters under the classification scheme, and δ is a constant used to adjust the order of magnitude of the number of classes; S75: Based on the determined optimal number of clusters, use the self-organizing mapping neural network to conduct unsupervised clustering partition of the ecological geological environment pattern to obtain the evaluation result.

9. A quantitative coupling evaluation system for regional ecological geological environment, characterized in that, It includes the following modules: Construct a multi-source dataset module, which is used to collect multi-source and multi-variety heterogeneous data, and adjust the spatio-temporal resolution and projection method of the data to be consistent. The multi-source and multi-variety heterogeneous data includes optical remote sensing images, climate meteorology, landform, and vegetation soil data. Data preprocessing module, which is used to preprocess the multi-source dataset, eliminate the errors caused by uncertain pixels and further unify the format. The preprocessing methods include cloud and fog removal, histogram matching, median calculation, and spatial interpolation for the multi-source and multi-variety heterogeneous data. Construct a regional comprehensive ecosystem service evaluation model module, which is used to quantitatively couple the functions of multiple key ecosystems and realize the comprehensive analysis of regional ecosystem services. Construct a regional ecosystem vulnerability evaluation model module, which is used to comprehensively measure the external structure and internal quality of the ecosystem and realize the comprehensive quantitative characterization of the vulnerability of the natural ecosystem. Construct a regional geological environment sensitivity evaluation model module, which is used to comprehensively analyze the geological background affecting the regional ecological geological environment. Construct a quantitative coupling evaluation model module for the regional eco-geological environment, which is used to construct the quantitative coupling evaluation model for the regional eco-geological environment by the equal-weight summation method according to the regional comprehensive ecosystem service evaluation model, the regional ecosystem vulnerability evaluation model and the regional geological environment sensitivity evaluation model, and obtain the eco-geological environment security index; The driving mechanism analysis and clustering partition module is used to conduct multi-scale driving mechanism analysis on the eco-geological environment security index and perform clustering partition based on unsupervised neural network to obtain the evaluation results.

Images