Ecological degradation condition identification method for multi-dimensional data collaborative processing

Abstract

The application discloses an ecological degradation condition identification method for multi-dimensional data cooperative processing, which comprises the following steps: step 1, a satellite image processing system interconnected with a local and a cloud is constructed, and a multi-dimensional metadata index mechanism is adopted to obtain cloud image data; step 2, the obtained satellite image is subjected to cloud removal processing to obtain a satellite image with a small cloud content; step 3, a processing model for mutual communication between satellite images and ecological indexes is constructed by using a combination of a plurality of vegetation index fusion and satellite inversion indexes and surface elements; step 4, a processing model for mutual communication between land utilization and ecological elements is constructed; conversion rules from land utilization to an ecological system type in the same period and ecological system succession in different periods are formulated, and the processing of attribute and spatial information is intelligently completed; and step 5, an ecological function and structure regional ecological degradation identification model is constructed by using a GIS platform algorithm. The method is simple, and the working efficiency and the accuracy are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of ecological protection and restoration, and particularly relates to a multi-dimensional data collaborative processing ecological degradation condition identification method. BACKGROUND

[0002] Land use / cover change research has become an important direction in the fields of environmental science, geography and land resource management, wherein analysis of land ecological conditions has gradually become a primary task. Land ecological condition evaluation can systematically measure and comprehensively evaluate the influence of human activities on land resources and ecological effects, so as to achieve sustainable development of land resources. Through ecological condition degradation evaluation of land space, the regional land distribution pattern and change characteristics can be qualitatively and quantitatively grasped, and the mutual influence law between nature, economy and society in the region is clarified, thereby providing a scientific reference for the construction of regional ecological civilization and the development of economic society.

[0003] At present, the identification of land ecological degradation condition mainly adopts the traditional method of manual division by operating personnel according to different types of land data, and the ecological condition is evaluated by combining GIS tool experience and traditional method. However, the existing method has low processing efficiency and low evaluation accuracy. Therefore, it is urgent to build a comprehensive intelligent ecological identification and evaluation method which can automatically identify and scientifically evaluate the land ecological condition. SUMMARY

[0004] The purpose of the present application is to provide a multi-dimensional data collaborative processing ecological degradation condition identification method, which automatically identifies by combining satellite resources and surface element data, thereby improving work efficiency and accuracy.

[0005] In order to achieve the above purpose, the present application adopts the following technical scheme:

[0006] A multi-dimensional data collaborative processing ecological degradation condition identification method comprises the following steps:

[0007] Step 1: a satellite image processing system interconnected by local and cloud is constructed, and a multi-dimensional metadata index mechanism is adopted to obtain cloud image data;

[0008] Step 2: the satellite image obtained in step 1 is subjected to cloud removal processing to obtain a satellite image with less cloud content;

[0009] Step 3: a satellite image and ecological index intercommunication processing model is constructed by using a combination of multiple vegetation index fusion and satellite inversion index and surface elements;

[0010] Step 4: a land use and ecological element intercommunication processing model is constructed; conversion rules of land use to ecological system type in the same period and ecological system succession in different periods are formulated, and the processing of attribute and spatial information is intelligently completed.

[0011] Step 5, using GIS platform algorithm, constructing ecological function, structural regional ecological degradation identification model.

[0012] Further, it further comprises step 6, setting multi-source format mixed calculation algorithm, constructing ecological degradation condition evaluation model, and evaluating ecological degradation condition from the perspective of spatial distribution data and text data.

[0013] Further, the step 1 specifically comprises:

[0014] Step 1.1: Constructing a satellite image processing system connected with local and cloud, which is a "star-ground" data processing system;

[0015] Step 1.2: Establishing a multi-dimensional metadata index in the satellite image processing system, the multi-dimensional metadata index mechanism comprises:

[0016] ① Based on B+Tree, establish pixel scale index, obtain satellite image set satisfying spatial resolution requirement, denoted as SC;

[0017] ② Based on B+Tree, establish time node index, obtain image timestamp, obtain satellite image set satisfying time resolution requirement, denoted as T;

[0018] ③ Invert the metadata index with spectral characteristics as key value, obtain satellite image set satisfying spectral resolution requirement, denoted as R;

[0019] ④ Based on R-tree, establish spatial index, adopt pruning algorithm to locate demand area, obtain satellite image set satisfying spatial range requirement, denoted as SP;

[0020] Step 1.3: Calculate the intersection of the four sets of indexes SP, T, R, SC, to obtain satellite image data IMa which satisfies spatial resolution, time resolution, spectral resolution and spatial range.

[0021] Further, step 2 specifically comprises:

[0022] Step 2.1: Using QA band method to preliminarily process cloud-containing image in the satellite image data IMa, dividing the image into four categories of no cloud, cloud, cloud shadow and other pixels; constructing cloud mask extraction algorithm to extract the pixels of each category to obtain cloud layer mask Cloud and non-cloud layer pixels Non-Cloud;

[0023] Step 2.2: According to the time sequence rainfall extreme metadata matching method, according to the cloud layer mask Cloud, match the spatial range image corresponding to the minimum rainfall value of the ground meteorological monitoring station on the same day, and screen out the satellite image;

[0024] Step 2.3: Use the median composite method to obtain satellite images with low cloud cover.

[0025] Further, step 3 includes:

[0026] Step 3.1: Using the satellite imagery data obtained in Step 2 as the identification base map, vegetation information is extracted based on the characteristics of the satellite imagery bands and the ratio calculation between bands. This enhances the difference between vegetation and other land features, allowing for the acquisition of vegetation information for any area of ​​the national territory.

[0027]

[0028] In the formula, Normalized Difference Vegetation Index (NDVI) Represents reflectivity in the near-infrared band. Represents reflectivity in the red light band;

[0029] Step 3.2: Based on the distribution characteristics of the satellite image histogram, a pixel-based binary classification method is adopted to divide the vegetation information obtained in Step 3.1 into pixels with and without vegetation cover. The vegetation cover index is then calculated from the divided satellite pixels. As a component of regional forest and grassland ecosystem biomass:

[0030]

[0031] Pixel division principle: Based on the histogram distribution information of remote sensing satellite imagery and the NDVI range, NDVI values ​​of 5% and below can be well represented as no-vegetation pixels, denoted as NDVI value range. ; 95% or more of the NDVI value is taken as the NDVI value of pure vegetation pixels, and is denoted as . ;

[0032] Step 3.3: Water body in green light band High reflectivity, in the mid-infrared band Strong absorption; while vegetation and soil have high reflectance in the mid-infrared. The difference between green light and mid-infrared can suppress vegetation interference. The green light band and mid-infrared band in the image are extracted, and the improved normalized water index is used. As a component of regional wetland ecosystem biomass:

[0033] .

[0034] Further, step 4 includes:

[0035] Step 4.1: Establish rules for the conversion of land use data into ecological elements:

[0036] Various land use type N land class coding -> mapping rule Q -> various ecosystem type M (M1, M2, M3, …, M8), wherein M1-M8 are forest, shrub, wetland, grassland, farm garden, farmland, town, and other ecosystems respectively;

[0037] Step 4.2: Based on the mapping rule Q specified in the above step 4.1, the attribute index of B+Tree is established, the retrieval balanced tree is constructed for the attribute field of the N land class coding elements, and the elements of each land class coding are obtained;

[0038] Step 4.3: Based on the cursor Cursor, the retrieved elements in the step 4.2 are modified in a streaming manner, the attribute field of the retrieved elements is assigned a value, the new value is stored in the new field STLX under the same map spot according to the mapping rule Q, the name is M1, M2, M3, …, M8, until the value in the new field STLX is not empty, and the spatial element Spatial-Feature1 with complete attribute value is obtained;

[0039] Step 4.4: Based on the above steps 4.1, 4.2, and 4.3, the spatial constraint is used for recursive fusion, the classification system of M1, M2, M3, …, M8 under each new field STLX is grouped and fused, and the spatial element Spatial-Feature2 is obtained.

[0040] Further, the step 5 specifically includes:

[0041] Step 5.1: Ecological function degradation condition identification

[0042] ①An ecological system function degradation identification model is constructed, the forest-grass-wetland ecological system is taken as an identification unit, the vegetation coverage index and the improved normalized water body index obtained in the step 3 are used as the surface area acquisition sources of the regional pixel ecosystem biomass and the pixel ecosystem top community biomass in the natural reserve, and the regional ecosystem degradation index is obtained by combining the weighted calculation of the adjustment coefficient:

[0043]

[0044] In the formula, is the ecosystem degradation index, , are respectively the regional ecosystem biomass expression; , are respectively the biomass expression of the regional ecosystem top community, and a and b are adjustment coefficients;

[0045] ② Extract the forest, grassland, and wetland ecosystems from the Spatial-Feature2 feature to obtain the vector spatial extent of the corresponding ecosystems. Using the vector spatial extent of the forest, grassland, and wetland ecosystems defined in step 4, and based on the GIS platform raster data, use the ExtractByMask function to obtain the actual biomass of the pixels of the forest, grassland, and wetland ecosystems within the region. , By overlaying the vector spatial extent of the nature reserve, the biomass of the climax communities of the forest, grassland, and wetland ecosystems within the nature reserve is obtained. , ;

[0046] ③ Based on the same value retention attribute of the cell center point of the raster data in the GIS platform, the raster data is converted into spatial point data using the RasterToPoint function;

[0047] ④ Using Python, create a new field in the pixel attribute table. M , , A new field is created to store the ecosystem biomass of a pixel and the biomass of the ecosystem's climax community. , EDI, and based on the cursor, flow cytometry calculations were performed to obtain the ratio of ecosystem biomass to climax community biomass of the corresponding pixel points and the EDI related to the regulation coefficient, and finally the vector point data of ecological function degradation, Feature-STGN, was obtained.

[0048] ⑤ Based on the PointToRaster function, a pixel point format conversion algorithm of the GIS platform, spatial point data is converted into raster data to obtain raster-STGN data on the degree of ecological function degradation;

[0049] Step 5.2: Identification of Ecological Structure Degradation

[0050] ① Based on the laws of urban development and ecological protection, the standards for constructing the transfer matrix are determined from the perspective of the ecosystem composition structure;

[0051] ②Assign values ​​to the transition matrices of different levels of change according to different types of succession;

[0052] a. If the land cover patches in years X and Y have spatial variations, the geometric boundaries of the land cover patches in years X and Y are traversed based on the Line Sweep algorithm, the intersection points of line segments are detected, the boundaries are cut along the intersection points based on the Weiler-Atherton algorithm, and the patches are reorganized into independent sub-faces. The attribute values ​​of the output sub-faces are directly inherited.

[0053] If the spatial range of the plot S of the land type map in X and Y years does not change, the spatial range surface is reserved, and the output sub-surface attribute value is directly inherited;

[0054] b. Based on the cursor, the elements meeting the corresponding standards of the transfer matrix are retrieved in turn, and dynamic assignment (0-5) is performed;

[0055] c. Using the boundary algebra method, the polygon boundary is traversed and the grid value is dynamically assigned, and the spatial surface domain data is converted into grid data to obtain the ecological structure degradation degree grid data Raster-STJG.

[0056] Further, the step 6 specifically comprises:

[0057] Step 6.1: Reclassify function is used to perform grid attribute value gradient reorganization on the data of functional degradation and structural degradation, Jenks natural breakpoint classification method is used to establish a value range mapping rule, and discrete pixel values are divided into continuous pixel value intervals;

[0058] Update the ecological function degradation degree grid data Raster-STGN and the ecological structure degradation degree grid data Raster-STJG field and rebuild the attribute table to obtain the degradation degree gradient attribute value, and realize the spatial distribution indication function of different degradation degrees;

[0059] Step 6.2: In combination with Python language, the degradation degree gradient attribute value obtained in the above step 6.1 is decoded in dbf binary structure, reconstructed into excel data, and the final result data is classified and counted in area, so that the proportion of different grades in the attribute table of the above grid data is determined, to assist the management department in precise policy-making.

[0060] The beneficial effects of the present application are:

[0061] (1) A multi-dimensional metadata retrieval mechanism is established to efficiently retrieve and obtain cloud image data;

[0062] (2) The image cloud removal mechanism is optimized, the image screening process with the best mapping quality is established, the combination of satellite image screening and ground meteorological monitoring station measured information is strengthened, the objective data verification support is enhanced, and the image precision is improved;

[0063] (3) A variety of vegetation index fusion, satellite inversion index combined with surface elements (nature reserves) are formulated to support the evaluation method of ecological function degradation, and the index evaluation precision is improved;

[0064] (4) Formulate the same period land use to the ecosystem type, the conversion rule of the ecosystem succession in different periods, intelligently complete the processing of attribute, spatial information, improve the data processing efficiency;

[0065] (5) Set the GIS platform algorithm, package the ecological function, structure biomass degradation calculation model, intelligently complete the processing of attribute, spatial information, improve the work module connection;

[0066] (6) Set the multi-source format mixed calculation algorithm, support the evaluation of ecological degradation condition from the angle of spatial distribution data and text data. BRIEF DESCRIPTION OF DRAWINGS

[0067] Figure 1 The flow chart of embodiment 1 of the present application;

[0068] Figure 2 The flow chart of embodiment 1 of the present application for obtaining satellite images of best quality;

[0069] Figure 3 The flow chart of embodiment 1 of the present application for constructing the processing model flow chart of satellite images and ecological index intercommunication Figure 1 ,

[0070] Figure 4 The flow chart of embodiment 1 of the present application for constructing the processing model flow chart of satellite images and ecological index intercommunication Figure 2 ;

[0071] Figure 5 The flow chart of embodiment 1 of the present application for constructing the regional ecological degradation condition identification model;

[0072] Figure 6 The flow chart of embodiment 1 of the present application for evaluating the ecological degradation condition. DETAILED DESCRIPTION

[0073] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Embodiment 1

[0075] Referring to Figures 1 to 6 The ecological degradation condition identification method of multi-dimensional data collaborative processing comprises the following steps:

[0076] Step 1: Construct a satellite image processing system interconnected with local and cloud, and acquire cloud image data by adopting a multi-dimensional metadata index mechanism; specifically comprising:

[0077] Step 1.1: Construct a satellite image processing system connected with local and cloud, which is a "star-ground" data processing system, specifically including:

[0078] The GEE platform permission is obtained by using the ee.Authenticate function and the ee.Initialize function, the satellite image resource is accessed by using the ee.ImageCollection function, and the cloud-local storage function is realized by using the ee.batch.Export.image.toDrive function. The system needs to have cloud data resource pool access module, cloud data resource pool retrieval module and cloud image data download-local storage module functions. A local database is established, and the database stores multi-source satellite images and surface elements. The satellite images include: Sentinel, Landsat, MODIS and other satellite sensor images, and the surface elements include land use data and natural protection data.

[0079] Step 1.2: Establish a multi-dimensional metadata index in the satellite image processing system, the multi-dimensional metadata index mechanism includes:

[0080] ①Based on B+Tree to establish pixel scale index, get satellite image set that meets spatial resolution requirements, denoted as Scale, abbreviated as SC;

[0081] ②Based on B+Tree to establish time node index, get image timestamp, get satellite image set that meets time resolution requirements, denoted as Time, abbreviated as T;

[0082] ③Inverted index of metadata with spectral features as key values, get satellite image set that meets spectral resolution requirements, denoted as Resolution, abbreviated as R; In this embodiment, the required wavebands are: red waveband, green waveband, near-infrared waveband, and mid-infrared waveband.

[0083] ④Based on R-tree to establish spatial index, use pruning algorithm to locate the required area, get satellite image set that meets spatial range requirements, denoted as Space, abbreviated as SP;

[0084] Step 1.3: Calculate the intersection of the four sets of indexes SP, T, R, SC, SP ∩ T ∩ R ∩ SC, get satellite image data Image that meets spatial resolution, time resolution, spectral resolution and spatial range, abbreviated as IMa, the image data is TIF format data, IMa represents the image obtained by filtering through the multi-dimensional metadata index mechanism.

[0085] Step 2, cloud removal processing is performed on the satellite image obtained in step 1 to obtain a satellite image with less cloud content, which is the best quality satellite image; specifically including:

[0086] Step 2.1: Preliminary processing of cloud-containing image in satellite image data IMa using QA band method, dividing the image into four categories of no cloud, cloud, cloud shadow and other pixels; constructing a cloud mask extraction algorithm to extract the pixels of each category to obtain cloud layer mask Cloud and non-cloud layer pixel Non-Cloud;

[0087] In view of the problem that the satellite image data IMa obtained in step 1 has multiple clouds and cloud shadows, affecting the mapping quality, a regional band recognition algorithm is constructed to decompose the satellite image into multi-level band data, recognize the QA band and perform mask operation, and divide the recognized results into four categories of no cloud, cloud, cloud shadow and other pixels; a cloud mask extraction algorithm is constructed to extract the pixels of each category to obtain cloud layer mask Cloud and non-cloud layer pixel Non-Cloud;

[0088] Step 2.2: Time sequence rainfall extreme value data matching method, according to the cloud layer mask Cloud, matching the spatial range image corresponding to the minimum rainfall value of the ground meteorological monitoring station on the same day to filter out the satellite image; including:

[0089] Step 2.2.1: Retrieve the ground meteorological monitoring station information Position in the area where the cloud layer mask Cloud is located, abbreviated as P.

[0090] Step 2.2.2: Retrieve the time sequence arranged rainfall data to obtain the minimum rainfall value Ri: Ri=min{R1,R2,……,Rn}, obtain the ground station number i corresponding to the minimum rainfall value and the date corresponding to the station number i and return the value: return{Di}, where i is the station number of the minimum rainfall value, D is the date corresponding to the station, and R is the rainfall corresponding to the station.

[0091] Step 2.2.3: Obtain the time retrieval value Di, return the B+Tree time node index with the time retrieval value Di, and match the time stamp of the image data. If the satellite image corresponding to the time retrieval value Di can completely cover the range of the cloud layer mask Cloud, stop retrieving; if it cannot completely cover, continue to retrieve the next low value, i.e. repeat the above steps 2.2.1, 2.2.2, 2.2.3, retrieve Rj, Rj>Ri, return{Dj}, and so on until the returned images Di, Dj, Dk,……,Dn can completely cover the cloud layer mask Cloud, then stop retrieving, wherein Di, Dj,……,Dn correspond to the returned retrieval satellite images CIi, CIj,……,CIn.

[0092] Step 2.3: obtain a satellite image with low cloud content by using the median synthesis method, in the embodiment: the satellite image with low cloud content is a satellite image with a cloud content of at most 15%, specifically including:

[0093] Step 2.3.1: according to the satellite image CIi, CIj, …, CIn returned in step 2.2.3, use the median synthesis method to splice, that is, CI (CI = CIi + CIj + … + CIn);

[0094] Step 2.3.2: replace the cloud mask Cloud with the synthesized CI to obtain the replaced cloud-containing pixel;

[0095] Step 2.3.3: splice the replaced cloud-containing pixel and the non-cloud layer pixel Non-Cloud to obtain the final satellite image with low cloud content IMb = Non-cloud + CI.

[0096] Referring to Figure 3 and Figure 4 , step 3, a processing model of satellite image and ecological index intercommunication is constructed by using a combination of multiple vegetation index fusion and satellite inversion index and surface elements, the surface elements including land use data and nature reserves; specifically including:

[0097] Step 3.1: using the satellite image data obtained in step 2 as the identification base map, according to the band characteristics of the satellite image, using the ratio operation between bands to extract vegetation information, enhancing the difference between vegetation and other ground objects, obtaining the vegetation information of any region of the land space,

[0098]

[0099] In the formula, is the normalized vegetation index, represents the reflectivity of the near-infrared band, represents the reflectivity of the red light band. Using satellite image data to judge ecological conditions is more economical and saves labor costs compared to existing large-scale and high-cost field measurements.

[0100] Band characteristics: satellite remote sensing images have multiple band information such as red light, green light, blue light, and infrared band. Based on the cell structure inside the vegetation leaf, the near-infrared light is scattered multiple times between cells, thereby having a high reflectivity, and in the red light band, the chlorophyll has strong absorption of red light, and the physical properties of the reflectivity of the vegetation are low.

[0101] Step 3.2: Based on the distribution characteristics of the satellite image histogram, a pixel-based binary classification method is adopted to divide the vegetation information obtained in Step 3.1 into pixels with and without vegetation cover. The vegetation cover index is then calculated from the divided satellite pixels. As a component of regional forest and grassland ecosystem biomass:

[0102]

[0103] Pixel division principle: Based on the histogram distribution information of remote sensing satellite imagery and the NDVI range, NDVI values ​​of 5% and below can be well represented as no-vegetation pixels, denoted as NDVI value range. ; 95% or more of the NDVI value is taken as the NDVI value of pure vegetation pixels, and is denoted as . ;

[0104] Step 3.3: Water body in green light band High reflectivity, in the mid-infrared band Strong absorption; while vegetation and soil have high reflectance in the mid-infrared. The difference between green light and mid-infrared can suppress vegetation interference. The green light band and mid-infrared band in the image are extracted, and the improved normalized water index is used. As a biomass of regional wetland ecosystems:

[0105] .

[0106] Step 4: Construct a processing model for the interconnection of land use and ecological elements; formulate conversion rules for land use to ecosystem type in the same period and ecosystem succession in different periods, and intelligently complete the processing of attribute and spatial information. That is, land use to ecosystem type in the same period: cultivated land in year X → cultivated land ecosystem in year X; ecosystem succession in different periods: cultivated land ecosystem in year X → forest ecosystem in year Y.

[0107] Step 4.1: Establish rules for the conversion of land use data into ecological elements:

[0108] Various land use type N land class code -> mapping rule Q -> various ecosystem types M (M1, M2, M3, …, M8), wherein the mapping rule Q follows the classification standard of the third national land survey work, and the first level land use type and part of the second level land use type are correspondingly converted into the ecosystem type of the method. The wetland, grassland, plantation, farmland, other land in the first level land use type are respectively converted into the wetland, grassland, farmland, farmland, other ecosystem of the ecosystem. The commercial service land, industrial and mining land, residential land, public management and public service land, special land, transportation land in the first level land use type are converted into the town of the ecosystem. The arbor forest land in the second level land use type is converted into the forest of the ecosystem. The shrub land in the second level land use type is converted into the shrub of the ecosystem. M1-M8 are forest, shrub, wetland, grassland, farmland, farmland, town, and other ecosystems.

[0109] Step 4.2: Based on the mapping rule Q specified in the above step 4.1, the attribute index of B+Tree is established, the retrieval balanced tree is constructed for the element attribute field of N land class codes, and the elements of each land class code are obtained.

[0110] Step 4.3: Based on the cursor Cursor, the retrieved elements in the step 4.2 are modified in a streaming manner, the attribute field of the retrieved elements is assigned a value, the new value is stored in the new field STLX under the same map spot according to the mapping rule Q, the name is M1, M2, M3, …, M8, until the value in the new field STLX is not empty, and the spatial element Spatial-Feature1 with complete attribute value is obtained.

[0111] Step 4.4: Based on the above steps 4.1, 4.2 and 4.3, the spatial constraint is used for recursive fusion, the classification system of M1, M2, M3, …, M8 under each new field STLX is grouped and fused, and the spatial element Spatial-Feature2 is obtained.

[0112] Step 5, using the algorithm of GIS platform, an ecological function and structure regional ecological degradation identification model is constructed, the processing of attribute and spatial information is intelligently completed, and the repeated manual work is reduced; the specific operation is as follows:

[0113] Step 5.1: Ecological function degradation condition identification

[0114]

[0115]

[0116] wherein, is the ecosystem degradation index, , are the regional ecosystem biomass expression respectively; , are the biomass expression of the regional ecosystem top community respectively, a, b are the adjustment coefficients, the interval range of the adjustment coefficient a is 0.7-0.9, the interval range of the adjustment coefficient b is 0.1-0.3, to ensure the relative importance of the balance of vegetation and water factors, to meet the value interval of the ratio of the adjustment coefficient a to the adjustment coefficient b is 2:1-9:1.

[0117] ②extract the forest, grassland, wetland ecosystem in the spatial element Spatial-Feature2 to obtain the corresponding ecosystem vector space range, utilize the vector space range of the forest, grassland, wetland ecosystem delineated in step 4, based on the GIS platform grid data, utilize the ExtractByMask function to obtain the actual biomass of the forest, grassland, wetland ecosystem in the region , , superimpose the vector space range of the nature reserve to obtain the biomass of the forest, grassland, wetland ecosystem top community in the nature reserve , .

[0118] ③based on the GIS platform grid data pixel center point same value retention attribute, utilize the RasterToPoint function to convert the raster data into spatial point data.

[0119] ④combine Python language, create a new field , , , in the pixel point attribute table, used to store the ecosystem biomass and the ecosystem top community biomass of the pixel point, create a new field , , EDI, and based on the cursor Cursor to carry out flow calculation to obtain the corresponding pixel point ecosystem biomass ratio and the ecosystem top community biomass ratio ​EDI related to the adjustment coefficient, and finally get the ecological function degradation degree vector point data Feature-STGN;

[0120] ⑤Based on the GIS platform pixel point format conversion algorithm PointToRaster function, the spatial point data is converted into raster data, and the ecological function degradation degree raster data Raster-STGN is obtained.

[0121] Step 5.2: Ecological structure degradation condition identification

[0122] ①Focusing on the law of urban development and ecological protection, the transfer matrix construction standard is determined from the composition structure of the ecological system,

[0123] : X year { }-Y year { }(i{1-8})

[0124] That is : X year { }-Y year { }, : X year { }-Y year { }, …, : X year { }-Y year { }

[0125] Forest, shrub, wetland, grassland, farmland, farmland, town, other ecological system set, i…8 for specific ecological system, Si represents the matrix number that can be listed, X is the base period year, Y represents the end period year.

[0126] Form the following changes:

[0127] Forest-forest, forest-shrub, forest-wetland, forest-grassland, forest-farmland, forest-farmland, forest-town, forest-other;

[0128] Shrub-forest, shrub-shrub, shrub-wetland, shrub-grassland, shrub-farmland, shrub-farmland, shrub-town, shrub-other;

[0129] Wetland-forest, wetland-shrub, wetland-wetland, wetland-grassland, wetland-farmland, wetland-farmland, wetland-town, wetland-other;

[0130] Grassland-forest, grassland-shrub, grassland-wetland, grassland-grassland, grassland-farmland, grassland-farmland, grassland-town, grassland-other;

[0131] Forest-forest, forest-shrub, forest-wetland, forest-grassland, forest-farmland, forest-farmland, forest-town, forest-other;

[0132] Farmland-forest, farmland-shrub, farmland-wetland, farmland-grassland, farmland-farmland, farmland-farmland, farmland-town, farmland-other;

[0133] Town-forest, town-shrub, town-wetland, town-grassland, town-farmland, town-farmland, town-town, town-other;

[0134] Other-forest, other-shrub, other-wetland, other-grassland, other-farmland, other-farmland, other-town, other-other.

[0135] 2. According to different types of succession mode, the transfer matrix of different change levels is valued.

[0136] A. If the X, Y year land type map spot has changed in the spatial range, based on the Line Sweep algorithm, the geometric boundary of the X, Y year land type map spot is traversed, the intersection points of line segments are detected, the boundary is cut along the intersection points based on the Weiler-Atherton algorithm, and the independent sub-surface is reorganized. The attribute value of the sub-surface is directly inherited;

[0137] If the X, Y year land type map spot S has no change in the spatial range, the spatial range surface is reserved, and the attribute value of the sub-surface is directly inherited;

[0138] B. Based on the cursor, the elements meeting the S1, S2, …, Si transfer matrix corresponding standard are retrieved in turn, and the dynamic value of 0-5 is given;

[0139] C. Using the boundary algebra method, the polygon boundary is traversed and the grid value is dynamically given, and the spatial surface domain data is converted into grid data to obtain the ecological structure degradation degree grid data Raster-STJG; as shown in the following table:

[0140]

[0141] Possibly, it also includes step 6, setting a multi-source format mixed calculation algorithm, constructing an ecological degradation condition evaluation model, and evaluating the ecological degradation condition from the perspective of spatial distribution data and text data; specifically including:

[0142] Step 6.1: Reclassify function is used to reorganize the grid attribute value of the functional degradation and the structural degradation degree grid data, and the Jenks natural breakpoint classification method is used to establish the value domain mapping rule, and the discrete pixel value is divided into continuous pixel value interval;

[0143] The value range mapping rule is to divide discrete numerical values into certain numerical intervals by using natural breakpoints, and to reorganize the numerical interval steps into different levels, such as low, lower, medium, higher, and high. The numerical interval is set according to the evaluation requirements. In the specific embodiment, 0-0.2 is reorganized and divided into a low level, 0.2-0.4 is reorganized and divided into a lower level, 0.4-0.6 is reorganized and divided into a medium level, 0.6-0.8 is reorganized and divided into a higher level, and 0.8-1 is reorganized and divided into a high level. See Figure 6

[0144] The Raster-STGN and Raster-STJG fields of the updated ecological function degradation degree raster data and the ecological structure degradation degree raster data are reconstructed to obtain the degradation degree step attribute value, and the spatial distribution of different degradation degrees is realized.

[0145] Step 6.2: In combination with the Python language, the degradation degree step attribute value obtained in the above step 6.1 is decoded in a dbf binary structure, reconstructed into an excel data, and various types of area classification statistics are performed on the final result data to determine the proportion of different levels, such as low, lower, medium, higher, and high, in the attribute table of the updated and reconstructed raster data, to assist the management department in making precise decisions.

[0146] The technical solutions provided by the present application are described in detail above. The principles and implementation methods of the present application are described by applying specific examples. The above examples are only used to help understand the method of the present application and its core idea. It should be noted that for ordinary skilled persons in the technical field, without departing from the principles of the present application, the present application can be improved and modified in several ways, and these improvements and modifications also fall within the protection scope of the claims of the present application.​

Claims

1. An ecological degradation condition identification method of multi-dimensional data collaborative processing, characterized in that: It comprises the following steps: Step 1: Construct a satellite image processing system connected with local and cloud, and obtain cloud image data by using multi-dimensional metadata indexing mechanism; Step 2: Perform cloud removal processing on the satellite image obtained in step 1 to obtain a satellite image with less cloud content; Step 3, the vegetation coverage index calculated by combining the satellite-retrieved index with the ground surface elements , as the regional forest-grassland ecosystem biomass and improved normalized water body index , as the regional wetland ecosystem biomass; Step 4: Formulate a conversion rule from land use data to ecosystem type, and complete the processing of attribute spatial information; Step 5: Use GIS platform algorithm to construct an ecological function and structure regional ecological degradation identification model, and step 5.1: ecological function degradation condition identification An ecological system function degradation identification model is constructed, taking a forest-grass wetland ecosystem as an identification unit, using the vegetation coverage index and the improved normalized water body index obtained in step 3 as the surface area acquisition source of the regional pixel ecosystem biomass and the pixel ecosystem top community biomass in the nature reserve, and combining the weighted calculation of the adjustment coefficient to obtain the regional ecosystem degradation index: wherein, is an ecosystem degradation index, , are biomass expressions of regional ecosystems, respectively; , are biomass expressions of top communities of regional ecosystems, respectively, a and b are adjustment coefficients, the interval range of the adjustment coefficient a is 0.7-0.9, the interval range of the adjustment coefficient b is 0.1-0.3, and the ratio of the adjustment coefficient a to the adjustment coefficient b is 2:1-9:

1. Step 5.2: Ecological structure degradation condition identification ①Focusing on the laws of urban development and ecological protection, the transfer matrix construction standard is determined from the composition structure of the ecosystem; ②According to different types of succession modes, the transfer matrix is valued for different change levels; a. If the X, Y year land class polygon has changed in spatial range, based on the Line Sweep algorithm, the geometric boundaries of the X, Y year land class polygon are traversed, the intersection points of the line segments are detected, the boundaries are cut along the intersection points based on the Weiler-Atherton algorithm, and the independent sub-surfaces are reorganized, and the attribute values of the sub-surfaces are directly inherited; If the X, Y year land class polygon S has no change in spatial range, the spatial range surface is retained, and the attribute values of the sub-surfaces are directly inherited; b. Based on the cursor, the elements meeting the corresponding standard of the transfer matrix are retrieved in turn, and dynamic assignment is performed (0-5); c. Using the boundary algebra method, the polygon boundaries are traversed and the grid values are dynamically assigned, and the spatial surface data is converted into grid data to obtain the ecological structure degradation degree grid data Raster-STJG.

2. The method of claim 1, wherein: It also comprises step 6: setting a multi-source format mixed calculation algorithm to construct an ecological degradation condition evaluation model to evaluate the ecological degradation condition from the perspective of spatial distribution data and text data; Step 6.1: Reclassify function is used to reorganize the attribute values of the function degradation and structure degradation degree grid data, and the Jenks natural break classification method is used to establish a value domain mapping rule to divide the discrete pixel values into continuous pixel value intervals; Update the ecological function degradation degree grid data Raster-STGN and the ecological structure degradation degree grid data Raster-STJG field and rebuild the attribute table to obtain the degradation degree gradient attribute value, realize the spatial distribution indication function of different degradation degrees; Step 6.2: In combination with the Python language, the degradation level attribute values obtained in the above step 6.1 are decoded in the dbf binary structure, reconstructed into excel data, and various types of area classification statistics are performed on the final result data to determine the proportion of different levels that have been divided in the attribute table of the preceding raster data, thereby assisting the management department in making precise decisions.

3. The method of claim 1, wherein: The step 1 specifically comprises: Step 1.1: constructing a satellite image processing system connected locally and in the cloud; Step 1.2: establishing a multi-dimensional metadata index in the satellite image processing system, and the multi-dimensional metadata index mechanism comprises: ①establishing a pixel scale index based on B+Tree to obtain a satellite image set satisfying the spatial resolution requirement, denoted as SC; ②establishing a time node index based on B+Tree to obtain an image timestamp and obtain a satellite image set satisfying the time resolution requirement, denoted as T; ③performing metadata inverted index with spectral features as key values to obtain a satellite image set satisfying the spectral resolution requirement, denoted as R; ④establishing a spatial index based on R-tree, positioning the required area by using a pruning algorithm, and obtaining a satellite image set satisfying the spatial range requirement, denoted as SP; Step 1.3: calculating the intersection of the four sets of indexes SP, T, R, and SC to obtain satellite image data IMa that simultaneously satisfies the spatial resolution, time resolution, spectral resolution, and spatial range.

4. The method of claim 3, wherein: Step 2 specifically comprises: Step 2.1: preliminarily processing cloud-containing image data in the satellite image data IMa by using a QA band method to divide the image into four categories of no cloud, cloud, cloud shadow, and other pixels; and constructing a cloud mask extraction algorithm to extract the cloud layer mask Cloud and the non-cloud layer pixel Non-Cloud from the pixels of the categories; Step 2.2: according to a time sequence rainfall extreme value metadata matching method, matching the cloud layer mask Cloud with the spatial range image corresponding to the minimum rainfall value of the ground meteorological monitoring station on the same day to screen out satellite image data; Step 2.3: obtaining satellite image data with low cloud content by using a median synthesis method.

5. The method of claim 1, wherein: The step 3 comprises: Step 3.1: using the satellite image data obtained in the step 2 as a recognition base map, extracting vegetation information according to the band characteristics of the satellite image, enhancing the difference between the vegetation and other ground objects, and obtaining vegetation information of any region of the national space, wherein, NDVI is the normalized difference vegetation index, Rnir represents reflectance in the near infrared band, Rred represents reflectance in the red band. Step 3.2: According to the distribution characteristics of the satellite image histogram, the pixel dichotomy is adopted to divide the vegetation information obtained in step 3.1 into vegetation-covered and non-vegetation-covered pixels, and the vegetation coverage index calculated from the divided satellite pixels is taken as the regional forest-grassland ecosystem biomass: , as the regional forest-grassland ecosystem biomass: wherein, is the value of a pixel for pure vegetation, is the value of a pixel for complete absence of vegetation cover; Pixel division principle: remote sensing satellite image histogram distribution information, according to the interval of the degree of location, for 5% and below NDVI value as no vegetation pixel can well represent the NDVI value range, recorded as ; 95% and above NDVI value as pure vegetation pixel NDVI value, recorded as ; Step 3.3: Water body in green band Reflectance is higher, in the mid-infrared band Strong absorption; while vegetation and soil reflect higher in the mid-infrared, the difference between green and mid-infrared can suppress vegetation interference, extract the green band and the mid-infrared band in the image, and use the improved normalized water body index , as the regional wetland ecosystem biomass: 。 6. The method of claim 5, wherein: The step 4 comprises: Step 4.1: specifying a rule for converting land use data into ecological elements: Various land use types N -> mapping rule Q -> various ecological system types M (M1, M2, M3, …, M8), wherein M1-M8 are forest, shrub, wetland, grassland, farm, farmland, town, and other ecological systems; Step 4.2: referring to the mapping rule Q specified in the above step 4.1, establishing an attribute index based on B+Tree, constructing a retrieval balanced tree for the element attribute fields of the N land class codes, and obtaining elements of each land class code; Step 4.3: Based on the cursor, the retrieved elements in step 4.2 are modified in a streaming manner, the retrieved elements are assigned attribute fields, and the new values are stored in the new field STLX under the same plot according to the reference mapping rule Q, named M1, M2, M3, …, M8, until the values in the new field STLX are not empty, obtaining the spatial element Spatial-Feature1 with complete attribute values; Step 4.4: Based on the above steps 4.1, 4.2, and 4.3, the spatial constraints are used for recursive fusion, and the classification system of references M1, M2, M3, …, M8 in each new field STLX is grouped and fused, obtaining the spatial element Spatial-Feature2.

7. The method of claim 6, wherein: The step 5 specifically comprises: Step 5.1: Ecological function degradation condition identification ① Construct an ecological system function degradation identification model, take the forest-grass wet ecological system as the identification unit, use the vegetation coverage index and the improved normalized water body index obtained in step 3 as the area domain acquisition source of the regional pixel ecosystem biomass and the pixel ecosystem top community biomass in the nature reserve, and combine the weighted calculation of the adjustment coefficient to obtain the regional ecosystem degradation index: wherein, is an index of ecosystem degradation, , are biomass expressions of regional ecosystems, respectively; , are biomass expressions of top communities of regional ecosystems, respectively, a, b are adjustment coefficients; ② Extract the forest, grassland, and wetland ecosystems in the spatial element Spatial-Feature2, obtain the vector space range of the corresponding ecosystems, use the vector space range of the forest, grassland, and wetland ecosystems delineated in step 4, based on the GIS platform grid data, use the ExtractByMask function to obtain the actual biomass of the forest, grassland, and wetland ecosystem pixels in the region 、 , superimpose the vector space range of the nature reserve, and obtain the biomass of the top community of the forest, grassland, and wetland ecosystems in the nature reserve 、 ; ③ Based on the same value retention attribute of the grid data pixel center point on the GIS platform, the grid data is converted into spatial point data using the RasterToPoint function; (4) Combined with Python language, a new field is created in the pixel point attribute table , M , , , for storing the ecosystem biomass and the biomass of the top community of the pixel point, a new field is created , , EDI, and based on the cursor, the corresponding pixel point ecosystem biomass ratio and the top community biomass ratio are calculated and the EDI related to the adjustment coefficient , and finally the ecological function degradation degree vector point data Feature-STGN is obtained ⑤ Based on the PointToRaster function of the pixel point format conversion algorithm on the GIS platform, the spatial point data is converted into grid data, obtaining the ecological function degradation degree grid data Raster-STGN; Step 5.2: Ecological structure degradation condition identification ① Face the city development and ecological protection law, and determine the transfer matrix construction standard from the composition structure of the ecological system; ② According to different types of succession modes, the transfer matrix is valued for different change levels; a. If the X, Y year land class plot has changed in spatial range, based on the Line Sweep algorithm, the geometric boundaries of the X, Y year land class plot are traversed, the intersection points of line segments are detected, and the boundaries are cut along the intersection points based on the Weiler-Atherton algorithm, and the independent subfaces are reorganized, and the subface attribute values are directly inherited; If the X, Y year land class plot S has no change in spatial range, the spatial range surface is retained, and the subface attribute values are directly inherited; b. Based on the cursor, the elements meeting the corresponding standard of the transfer matrix are retrieved in turn, and dynamic assignment is performed (0-5); c. Using the boundary algebra method, the polygon boundaries are traversed and dynamically assigned grid values, and the spatial surface data is converted into grid data, obtaining the ecological structure degradation degree grid data Raster-STJG.

Images