A dynamic monitoring method for the spatial distribution of urban biodiversity

Abstract

The present invention discloses a dynamic monitoring method for the spatial distribution of urban biodiversity, which relates to the technical field of biodiversity distribution surveys. The method includes using multi-source data to extract the boundary of the urban built-up area; and generating a spatial distribution map of urban biodiversity by using public big data and extrapolation method. By adopting the open big data of biological distribution records on the network platform and combining with the extrapolation method to generate a biodiversity map of urban areas with fine resolution, the present invention develops a general analysis framework, which can reduce the labor and time costs in the acquisition stage of basic data on urban biodiversity. On the other hand, by comparing remote sensing image data, it is possible to obtain data on animals and plants under the canopy cover, which is convenient for evaluating the species diversity under the canopy, and can capture complex information on species composition, which is beneficial to improving the adaptability to the complex artificial environment and highly fragmented habitat patches in urban areas.

Description

Technical Field

[0001] The present invention relates to the technical field of biodiversity distribution surveys, and specifically to a dynamic monitoring method for the spatial distribution of urban biodiversity. Background Art

[0002] Urban plant diversity is of great significance in maintaining the dynamic balance of urban ecosystems, improving social service levels, enhancing the living environment, and promoting the physical and mental health of urban residents. It is also an important indicator for measuring the sustainable development level of the urban environment. Efficient and detailed measurement of urban biodiversity is the basis for investigating the spatial distribution of biodiversity in urban areas, dynamically monitoring its changes, and formulating biodiversity plans, and plays an important role in evaluating the construction level of urban ecosystems.

[0003] Currently, there are mainly two types of evaluation methods for urban biodiversity: one is the traditional method of calculating diversity indices based on field survey sampling data or literature records; the other is to evaluate the vegetation type, coverage, forest degradation, etc. on a large scale by means of remote sensing (RS) and geographic information system (GIS) technologies through texture and green quantity characteristics.

[0004] In the prior art, the Chinese invention patent with the publication number CN116502800A discloses a method for evaluating urban biodiversity. This method uses a combination of remote sensing aerial photography technology and manual observation to obtain data on the types and numbers of organisms, and based on this biological annotation feature data, trains feature extraction models for aerial photography images of various sampling areas, thereby constructing a regional biodiversity database. This technology can meet the simulation of biodiversity on a relatively large scale, but the plant species distribution data therein needs to be obtained by relying on manual field survey sampling locations, which requires a large amount of human and time costs, and requires the investigators to have relatively high professional knowledge. Another Chinese invention patent with the publication number CN116485275A also proposes a method for evaluating the state of urban biodiversity. Based on a variety of basic materials and data, it establishes an evaluation data set and comprehensively evaluates urban biodiversity from three aspects: ecological security pattern, species diversity, and biodiversity management. However, in the data set construction stage, this method also relies on traditional data sources such as remote sensing images, field surveys, questionnaire visits, and government department document materials, which requires a relatively high level of data integrity, and further limits the estimation of urban biodiversity to the concept of administrative regions.

[0005] The above analysis of existing technologies shows that the current biodiversity assessment methods for urban areas are still mainly based on traditional field sampling surveys in the basic data acquisition stage, which not only has a certain professional threshold for operators, but also requires a lot of manpower and time. However, due to the particularity of the urban environment, remote sensing data has obvious deficiencies in data integrity and accuracy. For example, since remote sensing data is aerial images, it is impossible to assess species diversity under the canopy and it is difficult to capture complex information about species composition. To this end, we propose a dynamic monitoring method for the spatial distribution of urban biodiversity. Summary of the invention

[0006] The purpose of the present invention is to provide a dynamic monitoring method for the spatial distribution of urban biodiversity to solve the problems that the existing technology proposed in the above background technology requires a lot of manpower and time and is unable to evaluate the species diversity under the canopy and has difficulty in capturing the complex information of species composition.

[0007] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a method for dynamic monitoring of the spatial distribution of urban biodiversity, comprising extracting the boundaries of urban built-up areas using multi-source data; generating a spatial distribution map of biodiversity in urban areas using public big data and extrapolation method, wherein generating a spatial distribution map of biodiversity in urban areas comprises the following steps:

[0008] S1: Obtain and screen the distribution data of plants and animals in the target area on the public platform;

[0009] S2: Determine the size of the grid unit in the urban built-up area, divide the urban built-up area into square grids, and identify the grid units for saturated sampling based on the sample coverage of each grid unit calculated after expanding the reference sample size to twice;

[0010] S3: Based on the Hill number index of biodiversity, the abundance and relative richness of species are integrated into a type of diversity measurement index. The estimated biodiversity values ​​of saturated sampling grid cells based on the Hill number framework are obtained by programming in Python. The estimated biodiversity values ​​include species richness, Shannon diversity index, and Simpson diversity index.

[0011] S4: Collect data to build an extrapolation spatial model, evaluate the performance of the extrapolation spatial model, confirm the robustness of the extrapolation spatial model and determine the most reasonable extrapolation method, and extrapolate the diversity of non-saturated sampling units;

[0012] S5: Draw and refine the diversity distribution map, retaining the extrapolated values ​​in unsaturated grid cells and replacing the indicators in saturated grid cells with observed estimates.

[0013] As a further solution of the present invention: the extraction of the urban built-up area boundary includes:

[0014] Obtain remote sensing image data of the target city and perform preprocessing;

[0015] Calculate the preliminary normalized building index of the target city;

[0016] Combine the night-time light data with the normalized difference vegetation index to calculate the vegetation-corrected urban night-time light index;

[0017] Use the corrected urban night-time light index data to correct the preliminary normalized building index to obtain the building-corrected urban night-time light index;

[0018] Adopt the watershed algorithm. Consider the building-corrected urban night-time light index data as the terrain surface. The highlands with high pixel brightness are the high values of the regional night-time light data, and the lowlands with low pixel brightness are the low values of the regional night-time light data. The calculated terrain watershed is the vector boundary of the urban built-up area.

[0019] As a further solution of the present invention: The data for calculating the preliminary normalized building index of the target city is based on the data of satellite Sentinel-2A.

[0020] As a further solution of the present invention: When obtaining and screening the distribution data of animals and plants in the target area on the public platform, use the R interface package of the Global Biodiversity Information Facility.

[0021] As a further solution of the present invention: The size of the grid unit of the urban built-up area is based on the resolution of the acquired data, and the acquired data is multi-source data.

[0022] As a further solution of the present invention: The calculation formula of the diversity measure index is:

[0023]

[0024]

[0025] In the above formula, i represents the i-th species, S is the total number of species, and p represents the relative abundance of the i-th species;

[0026] When q = 0, 0D refers to species richness, and this richness does not consider the relative abundance of species;

[0027] When q = 1, formula 1 is meaningless, but when q approaches 1, the limit of this formula is the exponent of the Shannon entropy, representing Shannon diversity;

[0028] When q = 1, individuals are counted equally, so the number of species calculated is proportional to their abundance. Therefore, 1D can be interpreted as the effective number of common species in the community;

[0029] When q = 2, it represents the Simpson index. The roles of species other than the dominant species are underestimated, which is the effective number of dominant species in the community. Compared with other diversity measurement indicators, Hill numbers are more intuitive and statistically more rigorous.

[0030] As a further solution of the present invention: when q = 1, the number of species obtained by equally counting individuals is proportional to their abundances. That is to say, the more abundant a species is, the easier it is to be found and the more common it is. Among them, 1D is the effective number of common species in the community.

[0031] When q = 2, it represents the Simpson index. Since the relative abundances of each species are squared, the weights of other species in the calculation are reduced, and this parameter is the effective number of dominant species in the community.

[0032] As a further solution of the present invention: in the step S3, the programming includes:

[0033] A11_Plants02<-A11_Plants[,colsums(A11_Plants02)>0]

[0034] result<-iNEXT(A11_Plants02,q=c(0,1.2),datatype="abundance"

[0035] # Display my results. Here, it means to extract the sampling sufficiency percentage SC value. Among them, DataInfo is used to summarize data information, Asy

[0036] result$DataInfo

[0037] result$iNextEst

[0038] result$AsyEst

[0039] ASyESt A1]<-result$AsyESt

[0040] DataInfo A11<-result$DataInfo

[0041] # Calculate the diversity estimate value when the sample completeness is unified to 0.9. estimateD(A11 Plants02,datatype"abundance",base_"coverage"level=NULL,conf0.95)

[0042] # Export to the SC.CSV file

[0043] write.csv(DataInfo A11, "DataInfo A1l.csv", row.names = FALSE)

[0044] # Create a new table variable Species_richness_a11 and add species richness data

[0045] Species_richness_a11 <- AsyEst_A1l[seq(1, nrow(ASyEst_A11), by = 3), ]

[0046] # Export as a.CSV file

[0047] write.csv(species_richness_a11, "species_richness_a1l.csv", row.names = FALSE)

[0048] As a further aspect of the present invention: The collected data includes one or a combination of several of precipitation data, temperature data, terrain elevation data, actual evapotranspiration data, and soil survey geographic database data.

[0049] As a further aspect of the present invention: The extrapolation spatial model includes an ordinary Kriging model and a regression model with covariates.

[0050] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0051] 1. By using the open big data of biological distribution records on the network platform in combination with the extrapolation method to generate a fine-resolution biodiversity map of urban areas, the present invention develops a general analysis framework, which can reduce the labor and time costs in the acquisition stage of basic data on urban area biodiversity. On the other hand, by comparing remote sensing image data, it is possible to obtain data on animals and plants under the canopy cover, which is convenient for evaluating the species diversity under the canopy and can capture complex information on species composition, which is beneficial to improving the adaptability to the complex artificial environment and highly fragmented habitat patches in urban areas.

[0052] 2. By introducing multi-source data in the process of collecting biological distribution data in urban areas, the present invention can truly reflect the scope of urban built-up areas and densely populated areas, is not restricted by administrative unit boundaries, and is beneficial to improving the authenticity and accuracy of biological distribution data in urban areas.

[0053] 3. By comprehensively utilizing multi-source data such as nighttime light data and normalized vegetation data, the present invention improves the resolution of the map depicting the biodiversity distribution in urban areas. At the same time, by extracting the fine vector boundary of the urban built-up area, it can more realistically reflect the biodiversity distribution characteristics of the urban built-up area and densely populated areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a flowchart of the method for extracting the boundary of the urban built-up area of the present invention;

[0055] Figure 2 is a flowchart of the method for mapping the spatial distribution of biodiversity in urban areas of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0057] As Figure 1 shown is a flowchart of the method for extracting the boundary of the urban built-up area of the present invention, which provides a dynamic monitoring method for the spatial distribution of urban biodiversity, including the following steps:

[0058] Step 1: Obtain the remote sensing data of the target city, including remote sensing image data and nighttime light data, and perform preprocessing on the data.

[0059] Among them, the acquisition methods of satellite image data include the China Resources Satellite Application Center, the Geospatial Data Cloud Platform of the Computer Network Information Center of the Chinese Academy of Sciences, Copernicus Open Access Hub, LAADS, and USGS. The nighttime light data mainly includes two types: PMSP-OLS and NPP-VIIRS.

[0060] In this embodiment, the Sentinel-2A Level-2A data is selected for the remote sensing image data, and the NPP-VIIRS data with high-resolution guarantee is selected for the nighttime light data to meet the identification requirements of the fine built-up area range.

[0061] The data preprocessing includes image cropping, geometric correction, radiometric correction, and atmospheric correction. The preprocessing of the satellite image in this embodiment is implemented through ENVI software.

[0062] Step 2: Based on the Sentinel-2A data, extract the preliminary normalized building index of the target city;

[0063] By combining different groups of the 12 bands included in Sentinel-2A data, representative pixels of various land types in the urban built-up area are selected, and the band combination with the best land use recognition effect is chosen.

[0064] After comparison, in this embodiment, the 8th band and the 11th band are selected to calculate the preliminary normalized difference built-up index. The formula for the preliminary normalized difference built-up index is as follows:

[0065]

[0066] Among them, NDBI is the English abbreviation symbol for the preliminary normalized difference built-up index, and S8 and S 11 respectively represent the reflectance of the 8th band and the 11th band in Sentinel-2A data.

[0067] Step 3: Combine the nightlight data of NPP-VIIRS with the normalized difference vegetation index to calculate the vegetation-corrected urban nightlight index.

[0068] In this embodiment, MODIS sensor data is selected to calculate the normalized difference vegetation index. The formula for the vegetation-corrected urban nightlight index is as follows:

[0069] VANUI = (1 - NDVI) × NTL;

[0070] Among them, VANUI is the English abbreviation symbol for the vegetation-corrected urban nightlight index;

[0071] Step 4: Use the VANUI data and the NPP-VIIRS data to correct the preliminary normalized difference built-up index, and calculate the more accurate building-corrected urban nightlight index BANUI. This index will be used as the main basis for identifying the boundary of the urban built-up area. The formula for the building-corrected urban nightlight index BANUI is as follows:

[0072] BANUI = (1 + NDBI) × V NTL

[0073] Among them, V NTL represents the brightness value of the NPP-VIIRS data.

[0074] Step 5: Apply the watershed algorithm, regard the building-corrected urban nightlight index BANUI data as the terrain surface, and extract the vector boundary of the urban built-up area.

[0075] In this embodiment, in order to remove small patches at the built-up area boundary and smooth the patches adjacent to the aggregated boundary, the adaptive weighted median filtering method is used to remove noise from the BANUI image of the urban night light index based on building correction, and the image is morphologically reconstructed. The local maximum value of the reconstructed image is obtained as the foreground, and the background is marked by the inverse Euclidean distance method. Finally, the watershed segmentation algorithm is used to extract the vector boundary of the urban built-up area.

[0076] As Figure 2 shown in the flowchart of the method for drawing the spatial distribution map of biodiversity in urban areas of the present invention, the method in this stage mainly includes the following steps:

[0077] Step 6: Use the R interface package rgbif 3.7.9 of the Global Biodiversity Information Facility (GBIF) to obtain and screen the distribution data of animals and plants in the target area on the public platform. It is necessary to select the target species and public organizations or individuals with sufficient data samples according to the actual needs of biodiversity distribution assessment.

[0078] In this embodiment, taking the drawing of the spatial distribution of vascular plant species diversity as an example, all the recorded data of vascular plant species in the target area uploaded to the iNaturalist platform since 2020 are downloaded. In the RStudio software, the screening conditions are set as query = "Tracheophyta", rank = "PHYLUM", pred_gte("year", 2020).

[0079] During the data screening process, first eliminate the data entries in the downloaded original plant distribution record data that have incomplete species name information, missing longitude and latitude data, or the recorded geographical location falling into the ocean. Then, combined with the longitude and latitude values of the species upload records, the data is imported into the ArcGIS platform, and finally the records within the urban built-up area are screened out.

[0080] Step 7: Identify the grid cells with saturated sampling in the built-up area; determine the size of the grid cells in the urban built-up area according to the resolution of various data, divide the urban built-up area into square grids, and number them in sequence;

[0081] Connect the grid cells with the screened vascular plant distribution record data according to the spatial position relationship to create an occurrence matrix of various plants in each grid;

[0082] Use Python programming to measure the integrity of the samples based on the sample coverage rate of each grid cell calculated after expanding the reference sample size to twice, and use it to identify the grid cells with saturated sampling.

[0083] In this embodiment, the size of the grid unit in the urban built-up area is selected as 5 km × 5 km, and it is stipulated that the sample with a sample coverage rate ≥ 90% is complete, that is, the grid unit is identified as a saturated sampling grid unit.

[0084] Step 8: Calculate the biodiversity estimate of the saturated sampling grid unit;

[0085] In this embodiment, Python programming is used to calculate the α-diversity index based on the Hill number framework, including species richness, Shannon diversity index, and Simpson diversity index. For the convenience of unified comparison of the final calculation results, the sample coverage rate is uniformly stipulated as 90% for diversity extrapolation.

[0086] Step 9: Diversity extrapolation of unsaturated sampling units. Using precipitation data, temperature data, terrain elevation data, actual evapotranspiration data, and soil survey geographic database data, multiple diversity extrapolation spatial models are constructed, including ordinary Kriging model and regression model with covariates, and through 10-fold cross-validation, 90% of the training data of each model is randomly selected for model development and optimal variogram fitting, and the mean absolute error (MAE), relative mean absolute error (RRMSE), root mean square error (RMSE), and relative root mean square error (RRMSE) are calculated to evaluate the performance of each model, confirm the robustness of the model and determine the most reasonable extrapolation method. After comparison in this embodiment, the generalized orthogonal model with ordinary Kriging method (GAM OK) is selected as the final diversity extrapolation spatial model for unsaturated sampling units.

[0087] Step 10: Draw and optimize the diversity distribution map, retain the extrapolated values in the unsaturated grid units, and replace the indicators in the saturated grid units with the observed estimates.

[0088] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0089] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for dynamic monitoring of spatial distribution of urban biodiversity, characterized in that: The dynamic monitoring method comprises: Extract urban built-up area boundaries using multi-source data; Generate a spatial distribution map of biodiversity in urban areas by using public big data and extrapolation method, wherein generating a spatial distribution map of biodiversity in urban areas comprises the following steps: S1: Obtain and screen the distribution data of plants and animals in the target area on the public platform; S2: Determine the size of the grid unit in the urban built-up area, divide the urban built-up area into square grids, and identify the grid units for saturated sampling based on the sample coverage of each grid unit calculated after expanding the reference sample size to twice; S3: Based on the Hill number index of biodiversity, the abundance and relative richness of species are integrated into a type of diversity measurement index, and the estimated biodiversity of saturated sampling grid cells based on the Hill number framework is obtained by programming in Python; S4: Collect data to build an extrapolation spatial model, evaluate the performance of the extrapolation spatial model, confirm the robustness of the extrapolation spatial model and determine the most reasonable extrapolation method, and extrapolate the diversity of non-saturated sampling units; S5: Draw and refine the diversity distribution map, retaining the extrapolated values ​​in unsaturated grid cells and replacing the indicators in saturated grid cells with observed estimates.

2. The method for dynamic monitoring of spatial distribution of urban biodiversity according to claim 1, characterized in that: The extraction of the urban built-up area boundary includes: Obtain remote sensing image data of the target city and perform preprocessing; Calculate the initial normalized building index of the target city; The night light data is combined with the normalized vegetation index to calculate the urban night light index based on vegetation correction; The preliminary normalized building index is corrected using the corrected urban night light index data to obtain an urban night light index based on building correction; Using the watershed algorithm, the urban night light index data based on building correction is regarded as the terrain surface. The highland with high pixel brightness is the high value of the regional night light data, and the lowland with low pixel brightness is the low value of the regional night light data. The calculated terrain watershed is the vector boundary of the urban built-up area.

3. A method for dynamic monitoring of spatial distribution of urban biodiversity according to claim 2, characterized in that: The data for calculating the preliminary normalized building index of the target cities are based on data from the satellite Sentinel-2A.

4. The method for dynamic monitoring of spatial distribution of urban biodiversity according to claim 1, characterized in that: The R interface package of the Global Biodiversity Information Platform was used to obtain and screen the distribution data of plants and animals in the target area on the public platform.

5. The method for dynamic monitoring of spatial distribution of urban biodiversity according to claim 1, characterized in that: The size of the grid unit in the urban built-up area is based on the resolution of the acquired data, which is multi-source data.

6. The method for dynamic monitoring of spatial distribution of urban biodiversity according to claim 1, characterized in that: The calculation formula of the diversity measurement index is: In the above formula, i represents the i-th species, S is the total number of species, and p represents the relative abundance of the i-th species; When q = 0, 0D refers to species richness, which does not take into account the relative abundance of species; When q = 1, formula (1-1) is meaningless, but when q approaches 1, the limit of the formula is the exponent of Shannon entropy, which represents Shannon diversity; When q = 1, individuals are counted equally, so the number of species calculated is proportional to their abundance, so 1D is the effective number of common species in the community; When q=2, it represents the Simpson index, which is the effective number of dominant species in the community.

7. A method for dynamic monitoring of spatial distribution of urban biodiversity according to claim 6, characterized in that: When q = 1, the number of species obtained by equally counting individuals is proportional to their abundance, where 1D is the effective number of common species in the community; When q=2, it represents the Simpson index, which is the effective number of dominant species in the community.

8. The method for dynamic monitoring of spatial distribution of urban biodiversity according to claim 1, characterized in that: The collected data includes one or a combination of precipitation data, temperature data, terrain elevation data, actual evaporation data, and soil survey geographic database data.

9. A method for dynamic monitoring of spatial distribution of urban biodiversity according to claim 8, characterized in that: The extrapolated spatial models include ordinary Kriging models and regression models with covariates, and the biodiversity estimates include species richness, Shannon diversity index, and Simpson diversity index.

Images