Coastal zone planning method and system based on maXent and hotspot analysis

Abstract

The application discloses a coastal zone planning method and system based on MaXent and hotspot analysis, and relates to the technical field of coastal zone planning. The method comprises the following steps: S1, collecting environmental variable data, species distribution data and development intensity data, constructing a unified space-time grid system, and realizing time series expression of the data through data preprocessing; S2, constructing a quarterly time series maximum entropy model, time series evaluating habitat states of each grid, and forming habitat evolution types; S3, identifying time series significant hotspots and hotspot persistence and intensity characteristics, constructing a comprehensive representation of ecological development relationship, and identifying ecological development conflict types; and S4, spatially superimposing the habitat evolution types and the ecological development conflict types, forming adaptive access rules, and generating differentiated planning suggestions according to regions. The problems that the long-term dependence of coastal zone habitat suitability and development intensity on static time point data and the inability to identify time series evolution rules result in the lack of dynamic adaptability of planning are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coastal zone planning, in particular to a coastal zone planning method and system based on MaXent and hotspot analysis. BACKGROUND

[0002] As a key ecological-economic composite region where land and sea meet, the coastal zone integrates important ecological habitats, fishery breeding areas, transportation channels, and tourism and leisure spaces, and is the core strategic resource for regional sustainable development. With the advancement of urbanization in coastal areas, the strengthening of industrial activities, and the increasing demand for ecological protection, the coastal zone presents characteristics such as a high concentration of ecologically sensitive areas, rapid accumulation of development pressure, and highly dynamic natural processes. The environmental state and human activities both exhibit significant spatiotemporal variability, making the resource utilization, ecological security, and spatial control of the coastal zone increasingly require scientific, refined, and dynamic planning. Under this background, building a planning system that reflects the temporal characteristics of natural system evolution and human activities in the coastal zone has become an important foundation and practical need for regional governance and ecological protection.

[0003] For example, the invention patent with publication number CN118982313A discloses an intelligent system for coastal zone planning and protection, which includes an environmental protection zoning module, a spatial function zoning module, and a disaster prevention zoning module. The environmental protection zoning module is a functional area divided according to the coastal zone ecological protection red line as the control line. The spatial function zoning module is used to comprehensively consider the natural and social economic conditions of the coastal zone area, and to evaluate the resource environment carrying capacity, development density, and development potential of the coastal zone using an index evaluation method. The invention has the following advantages: the coastal zone area is regarded as a unified spatial unit, and a spatial function zoning system, an environmental protection zoning system, and a disaster prevention zoning system are constructed, corresponding to spatial planning, disaster prevention planning, and environmental protection planning of the coastal zone area respectively. An information linkage platform for multi-rule integration is constructed based on spatial planning, disaster prevention planning, and environmental protection planning, and finally a comprehensive planning system for the coastal zone is formed.

[0004] For example, the invention patent with publication number CN117933555A discloses a method for optimizing the use of coastal zone land, which relates to the technical field of land use optimization. The method calculates the growth or reduction trend of coastal zone land, which not only accurately positions the boundaries of coastal zone land, but also reasonably utilizes coastal zone land resources in advance according to the trend. The method then calculates the economic value generated in each region according to the regional planning of coastal zone land use, and determines whether the proportion of coastal zone land use is reasonable. The method further optimizes the allocation of coastal zone land use and analyzes the proportion of plant survival rate environmental factors to determine whether human intervention is needed. The method provides a systematic solution for the process of coastal zone land use.

[0005] However, the coastal environmental parameters such as sea temperature, salinity, suspended matter, pollutants change significantly with seasons and years, and the port transportation, tourist flow and fishery operation intensity also have strong seasonality. The existing planning is mostly based on static point data for Maxent modeling and hotspot analysis, lacking the identification of the time sequence evolution law of habitat suitability and development hotspot, and unable to reflect the long-term trend, mutation or seasonal distribution characteristics, resulting in that the planning is difficult to meet the adaptability demand under the future change scenario.

[0006] Therefore, in view of the above problems, there is an urgent need for a coastal belt planning method and system based on Maxent and hotspot analysis. SUMMARY

[0007] Technical problems to be solved

[0008] In view of the deficiencies of the prior art, the present application provides a coastal belt planning method and system based on Maxent and hotspot analysis, which solves the problem that the long-term habitat suitability and development intensity of the coastal belt rely on static point data and cannot identify the time sequence evolution law, resulting in that the planning lacks dynamic adaptability.

[0009] Technical scheme

[0010] To achieve the above purpose, the present application realizes the following technical scheme: a coastal belt planning method based on Maxent and hotspot analysis, comprising the following steps: S1, collecting environmental variable data, species distribution data and development intensity data, constructing a unified spatio-temporal grid system, generating structured time sequence data through coordinate correction, abnormal elimination, interpolation completion, trajectory denoising and intensity integration; S2, constructing a quarterly time sequence maximum entropy model according to the environmental variable data and the species distribution data, and combining the quarterly time sequence maximum entropy model to evaluate the time sequence of each grid habitat state, and forming a habitat evolution type according to the time sequence evaluation result; S3, identifying the time sequence significant development hotspot based on the development intensity data, and calculating the significant hotspot persistence and development hotspot intensity characteristics of each grid; combining the time sequence evaluation result, the significant hotspot persistence and the development hotspot intensity characteristics, constructing an ecological development relationship comprehensive representation, and identifying an ecological development conflict type according to the ecological development relationship comprehensive representation; S4, spatially superimposing the habitat evolution type and the ecological development conflict type to form an adaptive access rule, generating a planning layer based on the adaptive access rule, and generating differentiated planning suggestions according to different partitions of the planning layer.

[0011] Further, the environmental variable data, species distribution data and development intensity data are collected to construct a unified space-time grid system. Through coordinate correction, abnormality elimination, interpolation completion, trajectory denoising and intensity integration, the specific process of generating structured time series data is as follows: collecting environmental variable data, species distribution data and development intensity data; the environmental variable data includes sea surface temperature, salinity, turbidity, seawater monitoring indicators, water depth, slope and bottom type; the species distribution data includes species patrol records, species sample points and typical habitat distribution; the development intensity data includes ship AIS trajectory, fishing boat operation trajectory, operation days, scenic spot passenger flow and human flow density; the research area is divided into grid units according to fixed spatial resolution, and quarterly time slices are constructed according to uniform time step; the environmental variable data, species distribution data and development intensity data are mapped to the corresponding grid units and time slices; for the environmental variable data, image projection transformation and grid resampling are performed based on the unified coordinate system; the median absolute deviation method and quantile threshold are used to eliminate outliers for sea surface temperature, salinity, turbidity and seawater monitoring indicators; linear interpolation and moving average are used to complete the missing data completion and time series smoothing; for the species distribution data, coordinate projection and timestamp alignment are performed; patrol records and species sample points are mapped to the corresponding grid units by nearest neighbor interpolation, and repeated records and drift points are eliminated; for the development intensity data, trajectory denoising and speed threshold filtering are performed on ship AIS trajectory and fishing boat operation trajectory; the ship activity intensity and fishing operation intensity are generated by aggregating the points in the grid unit according to the point frequency; time resampling and spatial projection unified processing are performed on scenic spot passenger flow and human flow density to generate tourism activity intensity and shoreline human flow intensity; the normalized ship activity intensity, fishing operation intensity, tourism activity intensity and shoreline human flow intensity are calculated in each grid unit and quarterly time slice to obtain the comprehensive development intensity value of the grid unit in the quarter; the environmental variable data, species distribution data and development intensity data are subjected to z-score normalization processing; the coastal zone planning database is established, and the original and preprocessed environmental variable data, species distribution data and development intensity data are written into the coastal zone planning database.

[0012] Further, the specific process of constructing a quarterly time series maximum entropy model according to the environmental variable data and species distribution data is as follows: taking the preprocessed environmental variable data and species distribution data in the quarterly time slice as input, taking the environmental variable data in each time slice as continuous features, and taking the current species distribution data as positive samples, a contrast sample set of occurrence points and background points is constructed; based on the contrast sample set, a quarterly time series maximum entropy model is trained by maximum entropy algorithm to obtain the habitat suitability probability of each grid unit in the corresponding time slice, and the habitat suitability probability sequence is output.

[0013] Further, the specific process of time series evaluation of each grid habitat state combined with the quarterly time series maximum entropy model is as follows: taking the habitat suitability probability sequence of each grid cell as input, the average value and standard deviation are calculated to obtain the time series average suitability value and suitability standard deviation of the grid cell; by performing a linear regression in the coordinate system of the quarterly number and habitat suitability probability, the trend of habitat suitability probability over time is fitted, and the slope of the regression line is recorded as the habitat suitability trend slope; the relative volatility is obtained by dividing the suitability standard deviation by the sum of the time series average suitability value and a minimum constant, and the constant is subtracted from the relative volatility, and multiplied by the time series average suitability value to obtain the stable suitability intensity value; the inverse tangent value of the habitat suitability trend slope is calculated and divided by half of the constant to obtain the normalized trend value; the normalized trend value is added to the constant one, and multiplied by the stable suitability intensity value to obtain the time series habitat stability evolution value.

[0014] Further, the specific process of forming habitat evolution type according to the time series evaluation result is as follows: calculating the time series habitat stability evolution value of each grid cell, taking the time series habitat stability evolution value as the stability axis and the normalized trend value as the change trend axis, mapping each grid cell to the stability and trend two-dimensional feature space; comparing the time series habitat stability evolution value with the multi-level stability threshold H1 and H2, comparing the normalized trend value with the multi-level trend threshold T1 and T2, and determining the habitat evolution type: when ≥H2 and ≥T1, it is determined as a steady-state gain suitable area; when H1≤ <H2 and T1≤ <T2, it is determined as a seasonal steady-state suitable area; when H1≤ <H2 and <T1, or <H1 and T1≤ <T2, it is determined as a suitable fluctuation area; when <H1 and <T1, it is determined as a suitable degradation area; the time series habitat stability evolution value and the corresponding habitat evolution type are written into the coastal zone planning database.

[0015] Further, the specific process of identifying temporal significant development hotspots based on development intensity data and calculating the significant hotspot persistence and development hotspot intensity characteristics of each grid cell is as follows: Taking the ship activity intensity, fishery operation intensity, tourism activity intensity, and shoreline pedestrian flow intensity under the preprocessed quarterly time slices as inputs, organizing them into a development intensity sequence of grid cells according to the quarterly time serial numbers, and under the constraint of a unified spatial weight matrix, performing hotspot analysis on the development intensity sequence of each quarter using the Getis-Ord Gi* local spatial autocorrelation statistical method respectively, calculating the development hotspot significance values of each grid cell in each quarter, and marking the time slices with development hotspot significance values greater than the significance critical threshold as significant development hotspot moments; Taking the development hotspot significance value sequence of each grid cell as input, in all quarterly time steps, dividing the number of significant development hotspot moments by the total number of time steps to obtain the ratio of the significant hotspot persistence duration of the grid cell; For each grid cell, within the corresponding all time steps, adding the development hotspot significance values corresponding to the significant development hotspot moments to the hotspot intensity accumulation value, and at the same time counting the significant development hotspot moments into the significant hotspot duration count. After traversing all time steps, dividing the hotspot intensity accumulation value by the sum of the significant hotspot duration count and a very small constant to obtain the average hotspot intensity value of the grid cell; Adding a constant one to the average hotspot intensity value, taking the natural logarithm, and then multiplying by the ratio of the significant hotspot persistence duration to obtain the significant hotspot persistence intensity value of the grid cell.

[0016] Further, the specific process of constructing a comprehensive characterization of the ecological development relationship by combining the temporal evaluation results, significant hotspot persistence, and development hotspot intensity characteristics is as follows: Receiving the time series of the comprehensive development intensity values of each grid cell, performing unary linear regression in the coordinate system of the quarterly serial number and the comprehensive development intensity value, and fitting the regression slope representing the change trend of the comprehensive development intensity value over time, denoted as the comprehensive development intensity time trend slope; Obtaining the temporal habitat stability evolution value, habitat suitability trend slope, and significant hotspot persistence intensity value; Multiplying the habitat suitability trend slope by the comprehensive development intensity time trend slope, substituting it into the hyperbolic tangent function, calculating the trend coupling value, and performing non-negativity processing on the trend coupling value; Multiplying the trend coupling value by the non-negatively processed temporal habitat stability evolution value and the non-negatively processed significant hotspot persistence intensity value to obtain the comprehensive ecological development conflict value of the grid cell.

[0017] Further, the specific process of identifying the ecological development conflict type based on the comprehensive characterization of the ecological development relationship is as follows: Calculating the comprehensive ecological development conflict value of each grid cell and comparing it with the multi-level conflict thresholds C1 and C2 to identify the ecological development conflict type; When < C1, it is determined as a conventional development available area; When C1 ≤ < C2, it is determined as an ecological sensitive regulation area; When When ≥C2, it is determined as an ecological conflict control area; the comprehensive ecological development conflict value and the corresponding ecological development conflict type are written into the coastal zone planning database.

[0018] Further, the habitat evolution type and the ecological development conflict type are spatially superimposed to form adaptive access rules, and the planning layer is generated based on the adaptive access rules. The specific process of generating differentiated planning suggestions according to different partitions of the planning layer is as follows: based on the habitat evolution type and the ecological development conflict type, spatial superposition is performed to construct a time-series adaptive access rule system, including: ecological protection area, ecological fine regulation area, ecological development balanced utilization area, seasonal special utilization area and development suitable area; based on the time-series adaptive access rule system, spatial clustering analysis is performed to identify the spatial continuous region of the same type of planning unit; the region connectivity analysis and raster vectorization processing are adopted to convert the raster result into a spatial domain; the boundary smoothing algorithm is used to generate a continuous and smooth planning partition boundary to obtain a coastal zone planning partition layer; and according to different types of partitions, corresponding regional planning suggestions are generated, including: for the ecological protection area, suggestions of restriction development, ecological restoration, strict protection and enclosure management are generated; for the ecological fine regulation area, suggestions of development quota, activity frequency control and ecological monitoring intensification are generated; for the ecological development balanced utilization area, suggestions of utilization intensity restriction, ecological friendly facility layout and development total amount control are generated; for the seasonal special utilization area, suggestions of seasonal opening, restriction and adjustment are generated; for the development suitable area, suggestions of allowing conventional development and infrastructure layout are generated, and the environmental constraint conditions that need to be met during development are prompted; the regional planning suggestions are written into the coastal zone planning database together with the coastal zone planning partition layer.

[0019] The second aspect of the present application provides a coastal zone planning system based on MaXent and hotspot analysis, comprising: a multi-source time series data acquisition and preprocessing module for acquiring environmental variable data, species distribution data and development intensity data, constructing a unified spatio-temporal grid system, and generating structured time series data through coordinate correction, abnormality elimination, interpolation completion, trajectory denoising and intensity integration; a time series habitat suitability evolution analysis module for constructing a quarterly time series maximum entropy model based on environmental variable data and species distribution data, and conducting time series evaluation of the habitat state of each grid based on the quarterly time series maximum entropy model, and forming habitat evolution types based on the time series evaluation results; a time series hotspot identification and ecological conflict analysis module for identifying time series significant development hotspots based on development intensity data, and calculating significant hotspot persistence and development hotspot intensity characteristics of each grid; combining the time series evaluation results, significant hotspot persistence and development hotspot intensity characteristics, constructing a comprehensive representation of ecological development relationship, and identifying ecological development conflict types based on the comprehensive representation of ecological development relationship; a coastal zone planning generation and update module for spatially superimposing habitat evolution types and ecological development conflict types to form adaptive admission rules, generating planning layers based on the adaptive admission rules, and generating differentiated planning suggestions according to different partitions of the planning layers.

[0020] Advantages

[0021] The present application has the following advantages:

[0022] (1) The present application, by constructing a unified spatio-temporal grid system and quarterly time series slices, coordinates the environmental variables, species distribution and development intensity for coordinate correction, abnormality elimination, interpolation completion and trajectory denoising, realizes the structured time series expression of multi-source heterogeneous data, breaks through the dependence on traditional static data, and enables the coastal zone ecology and development process to be described long-term and continuously, providing reliable data basis for dynamic planning.

[0023] (2) The present application, by combining the time series maximum entropy model with habitat trend, stability and fluctuation analysis, quantitatively constructing time series habitat stability evolution value and identifying multiple habitat evolution patterns, realizes the systematic expression of habitat change trend, periodicity and degradation characteristics, and makes up for the deficiency of the prior art that cannot reveal the dynamic evolution law of habitat.

[0024] (3) The present application, by quantifying development pressure from the significance, persistence and intensity of development hotspots, and constructing a nonlinear coupling conflict index with habitat evolution trend, realizes fine identification of the degree of ecological and development contradiction in different regions, and effectively identifies potential conflict enhancement areas, sensitive risk areas and sustainable utilization areas.

[0025] (4) The present application forms self-adaptive access rules by superimposing habitat evolution types and ecological conflict types, automatically generates planning partitions and boundaries, and outputs differentiated planning suggestions according to regional characteristics, so as to realize the transformation of the coastal belt planning from static partition to dynamic, renewable, executable and intelligent planning mode.

[0026] Of course, implementing any product of the present application does not necessarily require all the advantages described above to be achieved at the same time. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 It is a flow chart of the coastal belt planning method based on MaXent and hotspot analysis;

[0028] Figure 2 It is a system structure diagram of the coastal belt planning based on MaXent and hotspot analysis;

[0029] Figure 3 It is a trend chart of habitat suitability change in four seasons;

[0030] Figure 4 It is a column chart of habitat stability evolution value in time sequence of a grid unit. DETAILED DESCRIPTION

[0031] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. As understood by those skilled in the art, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0032] Please refer to Figures 1-4 The embodiments of the present application provide a technical solution: a coastal belt planning method and system based on MaXent and hotspot analysis, as shown in Figure 1As shown, comprising the following steps: S1, collecting environmental variable data, species distribution data and development intensity data, constructing a unified spatio-temporal grid system, generating structured time series data through coordinate correction, abnormal rejection, interpolation completion, trajectory denoising and intensity integration; S2, constructing a quarterly time series maximum entropy model according to the environmental variable data and the species distribution data, and combining the quarterly time series maximum entropy model to evaluate the habitat state of each grid in time sequence, and forming the habitat evolution type according to the time sequence evaluation result; S3, identifying the time sequence significant development hotspot based on the development intensity data, and calculating the significant hotspot persistence and development hotspot intensity characteristics of each grid; combining the time sequence evaluation result, the significant hotspot persistence and the development hotspot intensity characteristics, constructing the comprehensive representation of ecological development relationship, and identifying the ecological development conflict type according to the comprehensive representation of ecological development relationship; S4, superimposing the habitat evolution type and the ecological development conflict type in space to form the adaptive access rule, generating the planning layer based on the adaptive access rule, and generating the differentiated planning suggestion according to the different partitions of the planning layer.

[0033] Specifically, environmental variable data, species distribution data and development intensity data are collected, a unified spatio-temporal grid system is constructed, and structured time series data is generated through coordinate correction, abnormality elimination, interpolation completion, trajectory denoising and intensity integration. The specific process is as follows: environmental variable data, species distribution data and development intensity data are collected; environmental variable data includes sea surface temperature, salinity, turbidity, seawater monitoring indicators, water depth, slope and bottom type; species distribution data includes species patrol records, species sample points and typical habitat distribution; development intensity data includes ship AIS trajectory, fishing boat operation trajectory, operation days, scenic spot passenger flow and human flow density; the research area is divided into grid units according to a fixed spatial resolution, and quarterly time slices are constructed according to a unified time step; environmental variable data, species distribution data and development intensity data are mapped to corresponding grid units and time slices; in the data collection link, various environmental variable data are obtained by connecting the data collected by national marine observation stations, satellite remote sensing products and coastal monitoring buoys, species distribution data are obtained by real-time recording of patrol personnel mobile terminals and existing ecological survey points, and development intensity data are obtained by AIS system, fishing boat monitoring system and scenic spot human flow monitoring interface, ensuring the authenticity and continuity of the data source; in the coordinate unification and image projection transformation process, the WGS84 unified geodetic coordinate system is adopted to avoid spatial deviation when superimposing multiple data sources. For environmental variable data, image projection transformation and grid resampling are performed based on the unified coordinate system; for sea surface temperature, salinity, turbidity and seawater monitoring indicators, the median absolute deviation method and quantile threshold are used to eliminate abnormal values by setting the MAD multiple threshold and quantile limit, so that the abnormal point elimination standard is reproducible; linear interpolation and moving average are used to complete the missing data completion and time series smoothing; for species distribution data, coordinate projection and timestamp alignment are performed; patrol records and species sample points are mapped to corresponding grid units by nearest neighbor interpolation, and repeated records and drift points are eliminated; linear interpolation, moving average and nearest neighbor interpolation are all based on a fixed algorithm window to ensure the consistency and stability of data completion. For development intensity data, trajectory denoising and speed threshold filtering are performed on ship AIS trajectory and fishing boat operation trajectory; in the trajectory denoising process, the joint filtering rule of speed threshold and heading change amplitude is used for AIS and fishing boat trajectory to eliminate drift points and invalid points; and point frequency is aggregated in the grid unit to generate ship activity intensity and fishing operation intensity, wherein ship activity intensity is obtained by counting the frequency of AIS trajectory points in the grid unit, and the frequency calculation considers a fixed time window, such as hourly aggregation; fishing operation intensity is identified based on points in the fishing boat operation trajectory that meet the operation conditions such as speed threshold and heading stability, and frequency aggregation is performed at the same grid scale, so that ship activity intensity and fishing operation intensity have comparability at the same spatial scale.The time resampling and spatial projection unified processing are performed on the scenic spot passenger flow and human flow density to generate tourism activity intensity and shoreline human flow intensity. The tourism activity intensity is resampled by the scenic spot passenger flow on a quarterly basis, and is projected in space according to the spatial relationship between the scenic spot boundary and the grid cell and the area proportion mapping mode. The shoreline human flow intensity is based on the spatial superposition of human flow density data and shoreline vectors. The monitoring points are assigned to the shoreline grid cells according to the distance attenuation, to ensure the unified expression in the same grid system. The arithmetic mean of the normalized ship activity intensity, fishery operation intensity, tourism activity intensity and shoreline human flow intensity is calculated in each grid cell and quarterly time slice to obtain the comprehensive development intensity value of the grid cell in the quarter. The environmental variable data, species distribution data and development intensity data are subjected to z-score normalization processing to ensure that all kinds of indexes have a uniform scale in the subsequent model input. The coastal zone planning database is established, and the original and preprocessed environmental variable data, species distribution data and development intensity data are written into the coastal zone planning database. The coastal zone planning database is established in the form of a structured data table, which supports the storage of environmental, species and development data respectively, and records the time slice label to ensure the traceability and auditability of data query and model calling.

[0034] In the embodiment, by constructing a unified space-time grid and time slice system and performing projection correction, anomaly removal, interpolation completion, trajectory denoising and intensity aggregation on multi-source environmental, species and development intensity data, the unification and high-quality expression of the three types of data in spatial scale and time scale are realized. The generated comprehensive development intensity index and normalization result ensure the consistency and comparability of the data, and the traceability and callability are realized through the structured database, which provides a reliable data basis for subsequent habitat modeling and conflict analysis.

[0035] Specifically, the specific process of constructing a seasonal time-series maximum entropy model according to environmental variable data and species distribution data is as follows: the pre-processed environmental variable data and species distribution data in the quarterly time slice are taken as input, wherein each quarterly time slice is generated by uniform time steps, and has completed projection correction, resampling and abnormality removal through the preprocessing step, ensuring the consistency of the input data in spatial coordinates and time scale; the environmental variable data in each time slice is taken as a continuous feature, including sea surface temperature, salinity, turbidity, seawater monitoring indicators, water depth and slope, which have been expressed as numerical variables in grid scale, to ensure that the input format of the maximum entropy model meets the requirements of continuous features; the current species distribution data is taken as a positive sample, which is composed of species patrol records, species sample points and typical habitat distribution, and has completed coordinate alignment and repeated point removal in the preprocessing stage to ensure that the sample is real, effective and corresponds to a specific quarter; a contrast sample set of occurrence points and background points is constructed, the background points are generated in the grid of the study area according to random sampling, and the number of background points can be adaptively set according to the size of the study area and the number of occurrence points to avoid affecting the stability of the model due to unbalanced sample proportion, and the same spatial constraint range as the occurrence points is set to ensure the comparability and stability of the model training; based on the contrast sample set, the seasonal time-series maximum entropy model is trained by the maximum entropy algorithm, L1 / L2 regularization is used to avoid overfitting, and each quarter is trained independently to form a time-sequenced model set, obtaining the habitat suitability probability of each grid cell in the corresponding time slice, the suitability probability is output as a continuous value from 0 to 1, representing the relative possibility of the grid meeting the habitat conditions of the target species, which conforms to the probability output characteristics of the maximum entropy model, and outputs the habitat suitability probability sequence, which is stored in the order of quarters.

[0036] In the present embodiment, by constructing a seasonal time-series maximum entropy model, the dynamic relationship between environmental variables and species distribution is realized, which can reflect the change rule of habitat suitability over time; by constructing a contrast sample set of occurrence points and background points and using a unified coordinate, a unified grid and a unified time slice data system, the accuracy and time sequence consistency of the model input data are ensured; at the same time, by outputting the probability sequence, the habitat state can be expressed in the form of continuous time sequence, providing quantifiable and traceable basis for subsequent trend fitting, stability evaluation and habitat evolution type identification, thereby significantly improving the scientificity and operability of the coastal habitat dynamic evaluation.

[0037] Specifically, the specific process of time series evaluation of each grid habitat state combined with the quarterly time series maximum entropy model is: taking the habitat suitability probability sequence of each grid cell as input, calculating the mean and standard deviation to obtain the time series average suitability value and suitability standard deviation of the grid cell; wherein, the mean is used to represent the overall suitability level, and the standard deviation is used to represent the fluctuation amplitude between quarters. To ensure consistency of calculation, all statistics are completed on a unified decimal precision and a unified time axis. By performing a linear regression in the coordinate system of the quarter number and the habitat suitability probability, the trend of the habitat suitability probability over time is fitted, and the slope of the regression line is recorded as the habitat suitability trend slope; wherein, the linear regression is based on the least squares method, the quarter number is taken as the independent variable, and the habitat suitability probability is taken as the dependent variable, which ensures the reproducibility of the trend slope and directly reflects the direction and speed of the suitability change with the quarter. Divide the suitability standard deviation by the sum of the time series average suitability value and a minimum constant to obtain the relative fluctuation degree, subtract the constant one from the relative fluctuation degree, and multiply it by the time series average suitability value to obtain the stable suitability intensity value; wherein, the relative fluctuation degree is used to quantify the stability of the suitability between seasons, and the minimum constant is used to prevent division by zero error when the average suitability value is too small; the stable suitability intensity value is used to represent the comprehensive stability level of the grid in the time series scale. Calculate the inverse tangent value of the habitat suitability trend slope and divide it by half of the circumference to obtain the normalized trend value, the inverse tangent function compresses the habitat suitability trend slope to the interval [-1, 1], which makes the dimension unified and facilitates subsequent combination calculation with other indicators; add the constant one to the normalized trend value and multiply it by the stable suitability intensity value to obtain the time series habitat stability evolution value, which comprehensively considers the trend direction, trend strength and seasonal stability, and can be used for quantitative and comparable time series discrimination of habitat evolution state.

[0038] wherein, the specific formula of the time series habitat stability evolution value is:

[0039] ;

[0040] In the formula, represents the time series habitat stability evolution value, which is used to comprehensively measure the habitat stability evolution state of each grid cell in the quarterly time series; represents the time series average suitability value, which is the average value of the suitability probability of the grid cell in all quarters, and is multiplied as a basic factor to determine the overall magnitude of the evolution value; represents the suitability standard deviation, which describes the stability of habitat suitability between quarters; represents the minimum constant, which takes the value of ; represents the habitat suitability trend slope, which describes the direction and speed of habitat evolution over time.

[0041] In this embodiment, Table 1 is a time sequence habitat stability evolution value data table. The table details the Q1 habitat suitability probability, Q2 habitat suitability probability, Q3 habitat suitability probability, Q4 habitat suitability probability, time sequence average suitability value, suitability standard deviation, habitat suitability trend slope and time sequence habitat stability evolution value of five different grid cells. Among them, the Q1 habitat suitability probability corresponding to grid cell 1 is 0.72, the Q2 habitat suitability probability is 0.75, the Q3 habitat suitability probability is 0.78, the Q4 habitat suitability probability is 0.80, the time sequence average suitability value is 0.762, the suitability standard deviation is 0.030, the habitat suitability trend slope is 0.027, and the time sequence habitat stability evolution value is 0.7446; the Q1 habitat suitability probability corresponding to grid cell 2 is 0.55, the Q2 habitat suitability probability is 0.57, the Q3 habitat suitability probability is 0.56, the Q4 habitat suitability probability is 0.58, the time sequence average suitability value is 0.565, the suitability standard deviation is 0.011, the habitat suitability trend slope is 0.008, and the time sequence habitat stability evolution value is 0.5568; the Q1 habitat suitability probability corresponding to grid cell 3 is 0.40, the Q2 habitat suitability probability is 0.36, the Q3 habitat suitability probability is 0.32, the Q4 habitat suitability probability is 0.30, the time sequence average suitability value is 0.345, the suitability standard deviation is 0.038, the habitat suitability trend slope is -0.034, and the time sequence habitat stability evolution value is 0.3005; the Q1 habitat suitability probability corresponding to grid cell 4 is 0.20, the Q2 habitat suitability probability is 0.25, the Q3 habitat suitability probability is 0.35, the Q4 habitat suitability probability is 0.45, the time sequence average suitability value is 0.312, the suitability standard deviation is 0.096, the habitat suitability trend slope is 0.085, and the time sequence habitat stability evolution value is 0.2280; the Q1 habitat suitability probability corresponding to grid cell 5 is 0.60, the Q2 habitat suitability probability is 0.50, the Q3 habitat suitability probability is 0.40, the Q4 habitat suitability probability is 0.30, the time sequence average suitability value is 0.450, the suitability standard deviation is 0.112, the habitat suitability trend slope is -0.100, and the time sequence habitat stability evolution value is 0.3168.

[0042] Table 1 Time sequence habitat stability evolution value data table

[0043]

[0044] As shown in FIG. 5, it is a four-season habitat suitability change trend chart. It shows the habitat suitability probability change trend of five grid cells in Q1 to Q4 four seasons. It intuitively reflects the ecological suitability fluctuation of different spatial units within the year, and is convenient for identifying its long-term maintenance, seasonal fluctuation or continuous decline trend. According to Table 1 and FIG. 5, Figure 3 Figure 3 ​It can be seen that the habitat suitability probability of grid cell 1 is relatively high throughout the four seasons, maintaining a steady increase, which is a typical steady-state gain. Grid cell 1 is generally stable between 0.55 and 0.58, with small fluctuations, showing seasonal steady-state. Grid cell 3 declines seasonally, reflecting a possible slow degradation trend. Grid cell 4 continues to rise, reflecting emerging improvements. Grid cell 5 shows a more significant decline throughout the four seasons, indicating a potential risk of degradation.

[0045] like Figure 4 The image shows a bar chart of temporal habitat stability evolution values ​​for five grid cells. It illustrates the combined stability, volatility, and trend of each cell on a year-round scale, used to identify different types of regions. (Based on Table 1 and...) Figure 4 It can be seen that grid cell 1 has the highest temporal habitat stability evolution value, exhibiting the best stability throughout the year and a positive trend; grid cell 2 has good overall stability with small seasonal fluctuations; grid cell 3 has weak stability and shows a continuous downward trend; grid cell 4, although showing some upward trend, has large historical fluctuations and insufficient overall stability; grid cell 5 has a stability evolution value of 0.3168, with a moderately low level of stability and a significant annual decline. Overall, this indicates significant differences in stability and trend direction among different regions, providing a basis for subsequent classification management and planning decisions.

[0046] In this implementation plan, by simultaneously incorporating the average level of habitat suitability probability, quarterly fluctuation characteristics, and long-term trend changes into a unified quantitative system, a temporal habitat stability evolution characterization that comprehensively reflects both stability and trend dimensions is constructed. This method utilizes repeatable linear regression, standard deviation statistics, and normalization transformation to clearly characterize the trend direction, rate of change, and degree of stability, achieving precise quantification of the temporal evolution state of habitats. This facilitates the accurate identification of habitat gains, declines, and fluctuation characteristics under seasonal fluctuations, multi-year variations, and abrupt changes, providing a reliable numerical basis for subsequent habitat evolution classification and ecological risk identification.

[0047] Specifically, the process of forming habitat evolution types based on time-series assessment results is as follows: The time-series habitat stability evolution value of each grid cell is calculated. Using the time-series habitat stability evolution value as the stability axis and the normalized trend value as the change trend axis, each grid cell is mapped to a two-dimensional feature space of stability and trend. The stability axis characterizes the suitability stability of the grid cell at multiple seasonal scales, while the trend axis characterizes the direction and rate of change of suitability over time, ensuring that different habitat change patterns are distinguishable within the same coordinate system. The two-dimensional feature space is regularly divided according to threshold intervals to ensure that the classification boundaries are interpretable and repeatable. The time-series habitat stability evolution value... The normalized trend value is compared with the multi-level stability thresholds H1 and H2. Compare with the multi-level trend thresholds T1 and T2 to determine the type of habitat evolution: When ≥H2 and ≥T1, it indicates that high habitat stability is maintained on the multi-quarter scale, and the suitability shows a continuous improvement trend. The habitat state is stable and reliable and has the ability of long-term positive evolution, and it is determined as a steady-state gain suitable area; When H1≤ <H2 and T1≤ <T2, it indicates that the habitat stability is at a medium level and the change trend of suitability is relatively gentle, reflecting that the habitat state is affected by seasonal factors but remains relatively stable as a whole, and it is determined as a seasonal steady-state suitable area; When H1≤ <H2 and <T1, or 41><H1 and T1≤ <T2, it indicates that there are obvious inconsistencies in the stability or trend dimension, the habitat state is sensitive to external disturbances, showing periodic fluctuations or uncertain evolution characteristics, and it is determined as a suitable fluctuation area; When <H1 and <T1, it indicates that the habitat stability is low and the suitability shows a continuous downward trend, indicating that the habitat state has undergone obvious degradation, and its ecological risks need to be focused on, and it is determined as a suitable degradation area; To ensure that all types of combination situations in the two-dimensional feature space have clear type attribution rules, for the boundary states not directly covered by the typical threshold combinations, a transition determination strategy is adopted for supplementary explanation: When ≥H2 and <T1, it indicates that the current habitat state of the grid cell is overall stable but shows a potential degradation trend, and it is determined as a transitional situation of steady-state suitability but with change risks, and it is included in the suitable fluctuation area for management; When <H2 and ≥T2, it indicates that the current habitat stability of the grid cell is weak but the suitability shows a significant improvement trend, and it is determined as a transitional situation with recovery potential, and it is also included in the suitable fluctuation area for unified control. Among them, the multi-level stability thresholds can be set by the quantile method according to the distribution of the temporal habitat stability evolution values of the grid cells, and the multi-level trend thresholds can be set by the quantile method according to the distribution characteristics of the normalized trend values, so that the classification criteria for different regions have objectivity and reproducibility. The classification logics of each type correspond to different habitat evolution mechanisms. For example, the steady-state gain suitable area represents that the suitability continuously rises while ensuring stability, and the suitable degradation area represents low stability and a continuous downward trend, which is convenient for implementing differential management measures in subsequent planning. Write the temporal habitat stability evolution values and the corresponding habitat evolution types into the coastal zone planning database, and store the classification results with the grid cell number and quarterly time label as indexes.

[0048] In this embodiment, by constructing a two-dimensional characterization system of stability axis and trend axis, the fluctuation characteristics and long-term trend of habitat suitability on the quarterly scale are quantified synchronously, and the habitat state is upgraded from "static suitability" to "time evolution characteristics". Through the combination of multi-level stability threshold and trend threshold, the habitat state of different regions can be accurately distinguished according to its stability level and trend direction, forming a habitat evolution type with interpretability, reproducibility and landability. It provides a structured and differentiated input basis for subsequent ecological conflict identification and spatial planning, and significantly improves the scientificity and fine management ability of coastal habitat assessment.

[0049] Specifically, the process of identifying significant development hotspots based on development intensity data and calculating the persistence and intensity characteristics of significant hotspots in each grid is as follows: Using preprocessed quarterly time slices of vessel activity intensity, fishery operation intensity, tourism activity intensity, and shoreline pedestrian flow intensity as inputs, development intensity sequences are organized into grid units according to quarterly time sequence numbers. Under the constraint of a unified spatial weight matrix, the Getis-Ord Gi* local spatial autocorrelation statistical method is used to analyze hotspots for each quarter's development intensity sequence. The spatial weight matrix is ​​constructed based on the K-neighborhood method to constrain the spatial relationships during hotspot statistics, avoiding isolated points from affecting the significance calculation results and ensuring that hotspot identification has spatial statistical significance. The significance values ​​of development hotspots for each grid cell in each quarter were calculated, and time slices with development hotspot significance values ​​greater than the significance threshold were marked as significant development hotspot moments. The significance threshold adopted the significance level corresponding to the Gi statistic, such as 1.96 of the 95% confidence interval, as the judgment standard to ensure the statistical reliability of the hotspot identification results. Using the development hotspot significance value sequence of each grid cell as input, the number of significant development hotspot moments was divided by the total number of time steps in all quarters to obtain the significant hotspot duration ratio of the grid cell, which reflects the long-term hotspot stability of the grid cell within the study period and can be used to measure the continuity and sustained pressure level of development intensity. For each grid cell, within the corresponding time steps, the development hotspot significance value corresponding to the significant development hotspot moment was included in the cumulative hotspot intensity value, and the significant development hotspot moment was also included in the significant hotspot duration count. The accumulation process of development hotspot significance value was traversed step by step according to the time steps, and each significant development hotspot moment was accumulated once to ensure that the calculation of the cumulative hotspot intensity value had a strict time correspondence. After traversing all time steps, the average hotspot intensity value of the grid cell is obtained by dividing the cumulative hotspot intensity value by the sum of the significant hotspot duration count and the minimum constant. The minimum constant is used to avoid the abnormal situation of the denominator being zero, thereby ensuring the stability and continuity of the calculation process. The average hotspot intensity value is added to the constant and the natural logarithm is taken. Then, it is multiplied by the ratio of significant hotspot duration to obtain the significant hotspot persistence intensity value of the grid cell. By combining the characterization of hotspot intensity and persistence, the peak characteristics and long-term cumulative effects of development pressure can be reflected simultaneously, thus forming a more comprehensive quantitative description of development risk.

[0050] The specific formula for the persistence intensity value of significant hotspots is as follows:

[0051] ;

[0052] In the formula, The value represents the persistence intensity of significant hotspots, comprehensively reflecting the persistence and intensity level of significant development hotspots in the grid cells throughout the entire time series. represents the hotspot intensity accumulation value, which represents the total intensity contribution of the hotspot; represents the significance critical threshold value; represents the significant hotspot duration ratio value, which represents the proportion of time grids in the significant development hotspot state in all time steps; represents the hotspot intensity accumulation value, which represents the total intensity contribution of the hotspot; represents the significant hotspot duration count value, which represents the number of occurrences of the significant development hotspot; represents the minimum constant, which is .

[0053] In this embodiment, by introducing the grid representation, unified time step, and spatial statistical mechanism of multiple types of development intensity, the comparability of development activities on the quarterly scale is realized, making the hotspot identification have a stable statistical basis; by combining the Getis-Ord Gi* significance judgment, the joint quantification of the hotspot duration ratio and the hotspot intensity, the peak value and the persistence of the development pressure are formed, which can accurately distinguish the short-term abnormal fluctuations and the long-term cumulative high-pressure areas; at the same time, by the minimum constant, the calculation anomaly is avoided, making the average hotspot intensity have robustness, so as to ensure that the calculation process of the significant hotspot duration intensity value is reproducible and traceable, and the accuracy and reliability of the coastal belt development pressure identification are significantly improved.

[0054] Specifically, the specific process of constructing the comprehensive representation of the ecological development relationship is as follows: receiving the time series of the comprehensive development intensity value of each grid cell, wherein the comprehensive development intensity value is obtained by normalizing and arithmetically averaging the ship activity intensity, fishery operation intensity, tourism activity intensity and shoreline flow intensity at the quarterly scale; performing a linear regression in the coordinate system of the quarterly number and the comprehensive development intensity value, and fitting to obtain the regression slope representing the trend of the comprehensive development intensity value over time, denoted as the comprehensive development intensity time trend slope, the regression process is based on the least square method, which can quantify the direction and amplitude of the change of the development pressure over the quarters, and avoid the interference caused by the single time step anomaly; obtaining the time series habitat stability evolution value, habitat suitability trend slope and significant hotspot persistence intensity value, which correspond to the ecological carrying capacity, ecological trend change and development disturbance intensity respectively; multiplying the habitat suitability trend slope and the comprehensive development intensity time trend slope, and substituting into the hyperbolic tangent function to calculate the trend coupling value, and performing non-negative processing on the trend coupling value, the hyperbolic tangent function as a bounded mapping function can compress the trend coupling value to the interval [-1, 1], avoiding the unbounded amplification of the trend coupling value and improving the numerical stability of the subsequent comprehensive calculation; multiplying the trend coupling value and the non-negative processed time series habitat stability evolution value and the non-negative processed significant hotspot persistence intensity value to obtain the comprehensive ecological development conflict value of the grid cell; wherein the non-negative processing is a fixed mathematical transformation of the index value, which is limited in the non-negative interval: if the index value may appear a small negative value, the negative value truncation method is adopted, that is, the index value is compared with zero and the larger one is taken, so that the negative value is set to zero; avoiding the direction misjudgment of the conflict value caused by the negative value, and improving the stability and interpretability of the calculation of the comprehensive ecological development conflict value.

[0055] wherein the specific formula of the comprehensive ecological development conflict value is:

[0056] ;

[0057] in the formula, represents the comprehensive ecological development conflict value, which comprehensively measures the conflict degree of the grid cell in the aspects of long-term ecological evolution, development pressure accumulation and ecological development trend coupling, and is a composite index integrating the change of ecological carrying capacity, the change trend of development intensity and the persistence characteristics of significant development hotspots; represents the non-negative processed time series habitat stability evolution value, which measures the long-term stability and change trend of the habitat state, and is a basic magnitude of the ecological bottom carrying capacity; represents the non-negative processed significant hotspot persistence intensity value, which measures whether the development activities are spatially significant and long-term persistent, and reflects the long-term accumulation of existing development pressure; represents the habitat suitability trend slope; The development intensity trend slope represents the comprehensive development intensity trend, which is used to describe whether the development is growing rapidly or decaying.

[0058] In this embodiment, by constructing the multi-dimensional coupling relationship between the ecological carrying capacity trend, the habitat change trend and the development disturbance intensity, the quantitative description of the ecological development conflict in the coastal zone is realized. Through the regression analysis of the quarterly change of the comprehensive development intensity, the long-term trend of the development pressure can be identified; the introduction of the habitat suitability trend slope and the time sequence habitat stable evolution value can reflect the ecological vulnerability from the two aspects of the stability and evolution direction of the ecological system; at the same time, the cumulative effect of the development disturbance is combined with the significant hot spot continuous intensity. After the multi-index is unified and non-negative, it is involved in the conflict value calculation, which avoids the direction misjudgment caused by the negative value, so that the comprehensive ecological development conflict value has good stability, interpretability and comparability. This method can effectively identify the key areas of the development and ecological contradiction, and provide reliable data support and decision basis for scientific zoning, ecological red line delimitation and differentiated development management.

[0059] Specifically, the specific process of identifying the type of ecological development conflict according to the comprehensive representation of ecological development relationship is as follows: calculating the comprehensive ecological development conflict value of each grid cell , and comparing it with the multi-level conflict threshold C1 and C2 to identify the type of ecological development conflict. The multi-level conflict threshold can be dynamically adjusted according to the ecological sensitivity of different regions to ensure that the determination standard has regional adaptability; when C1, it is determined as a regular development available area, indicating that the ecological carrying pressure is low, the habitat state is stable, and the development activities have little impact on the ecological system, so regular construction and industrial layout are allowed, but the basic environmental constraints still need to be met; when C1≤ C2, it is determined as an ecological sensitive control area, indicating that the ecological suitability exists stage fluctuation or the development activities have a certain impact on the ecological system, so the total amount control, limited intensity development and access conditions need to be implemented within the developable range, and the ecological monitoring and risk warning need to be strengthened; when C2, it is determined as an ecological conflict control area, indicating that the ecological degradation trend is obvious and the development disturbance intensity is high, which belongs to high-risk space, so the newly added development needs to be strictly limited or suspended, and the ecological restoration and buffer zone construction measures need to be implemented to restore the ecological function; the comprehensive ecological development conflict value and the corresponding ecological development conflict type are written into the coastal zone planning database, and the grid cell number and time slice label are recorded synchronously.

[0060] In this embodiment, by grading the comprehensive ecological development conflict value, the quantifiable and interpretable conflict identification of different spatial units in the coastal zone under a unified standard is realized. The setting of multi-level conflict threshold can distinguish between the conventional developable area, the ecologically sensitive area and the high conflict control area, thereby providing a basis for matching differentiated management intensity for different types of areas. At the same time, combined with the database writing mechanism, it ensures that the conflict results are traceable, queryable and directly used for planning partition generation. This method not only improves the accuracy and consistency of conflict identification, but also greatly enhances the executability and decision support ability of coastal zone planning.

[0061] Specifically, the habitat evolution type and the ecological development conflict type are spatially overlaid to form adaptive access rules, and the planning layer is generated based on the adaptive access rules. According to the different partitions of the planning layer, the specific process of generating differentiated planning suggestions is as follows: Based on the habitat evolution type and the ecological development conflict type, spatial overlay is performed to construct a time-series adaptive access rule system, including: ecological protection zone, ecological fine regulation zone, ecological development balanced utilization zone, seasonal special utilization zone and development suitable zone; The spatial overlay process is based on a unified grid coordinate system and uses a grid-by-grid reclassification method to convert the combination of stability and conflict level into regional access categories, ensuring that the overlay logic is repeatable and traceable. Based on the time-series adaptive access rule system, spatial clustering analysis is performed to identify spatially continuous regions of the same planning unit. Spatial clustering uses eight-neighbor connectivity analysis to identify contiguous areas, and spatial fusion processing is performed on isolated grids with small areas to incorporate them into surrounding areas, improving the integrity and continuity of the planning blocks. Using regional connectivity analysis and raster vectorization processing, the raster results are converted into spatial domains. In the vectorization process, a polygon generation algorithm is used and a topological check is performed on the boundary to avoid topological errors such as fragmented domains and hanging nodes. The boundary smoothing algorithm is used to generate a continuous and smooth planning partition boundary. The Bezier curve smoothing algorithm is used to make the planning partition boundary conform to the natural morphological characteristics of the coastal zone and facilitate implementation by management departments. The coastal zone planning partition layer is obtained, and corresponding regional planning suggestions are generated for different types of partitions, including: For the ecological protection zone, suggestions for limiting development, ecological restoration, strict protection, and enclosure management are generated. No new construction activities are allowed, and regular ecological monitoring is required, with priority given to implementing degraded habitat restoration projects. For the ecological fine regulation zone, suggestions for development quota, activity frequency control, and intensified ecological monitoring are generated. Limited development can be implemented, but an activity approval mechanism and real-time monitoring system must be established. For the ecological development balanced utilization zone, suggestions for utilization intensity limitation, ecological-friendly facility layout, and total development control are generated. While allowing development, the intensity, layout, and rhythm must be controlled to ensure that ecological pressure is controllable. For the seasonal special utilization zone, suggestions for seasonal opening, restriction, and adjustment are generated. Seasonal closure and restriction measures must be implemented during fishing seasons, breeding seasons, and other ecologically sensitive periods. For the development suitable zone, suggestions for allowing regular development and infrastructure layout are generated, and environmental constraint conditions that must be met during development are suggested, such as construction period emission reduction measures and specific requirements for ecological buffer zone settings, to ensure that development behavior is compliant and controllable. The regional planning suggestions are written into the coastal zone planning database along with the coastal zone planning partition layer. At the same time, an update mechanism is supported: After new quarterly data is written into the coastal zone planning database, the model and indicators are automatically updated, including the habitat suitability probability sequence, the time-series habitat stability evolution value, the significant hotspot persistence intensity value, and the comprehensive ecological development conflict value. The habitat evolution type and the ecological development conflict type are also re-determined.The grid cells with changed categories are subjected to local spatial clustering, vectorization and differential updating of boundary smoothing to generate new planning partitions and corresponding planning suggestions, and a new version is written into the database.

[0062] In the embodiment, by spatially superimposing the habitat evolution type and the ecological development conflict type, and cooperating with connectivity analysis, raster vectorization and boundary smoothing, a complete conversion from the time-series ecological process to the spatial planning partition is realized, a clear structure, continuous boundary and logical consistent coastal belt planning layer is generated, and differentiated and executable planning suggestions are made for the five types of regions, so that the planning result has a clear management direction and landing operability, and the adaptability, implementability and management accuracy of the coastal belt spatial partition access rules are significantly improved.

[0063] Referring to Figure 2 The second aspect of the present application provides a coastal belt planning system based on MaXent and hotspot analysis, which is applied to the above-mentioned coastal belt planning method based on MaXent and hotspot analysis, and comprises: a multi-source time-series data acquisition and preprocessing module for acquiring environmental variable data, species distribution data and development intensity data, constructing a unified spatio-temporal grid system, and generating structured time-series data through coordinate correction, abnormality elimination, interpolation completion, trajectory denoising and intensity integration; a time-series habitat suitability evolution analysis module for constructing a quarterly time-series maximum entropy model according to the environmental variable data and the species distribution data, and conducting time-series evaluation of the habitat state of each grid based on the quarterly time-series maximum entropy model, and forming a habitat evolution type according to the time-series evaluation result; a time-series hotspot identification and ecological conflict analysis module for identifying time-series significant development hotspots based on the development intensity data, and calculating significant hotspot persistence and development hotspot intensity characteristics of each grid; combining the time-series evaluation result, the significant hotspot persistence and the development hotspot intensity characteristics, constructing a comprehensive representation of ecological development relationship, and identifying an ecological development conflict type according to the comprehensive representation of ecological development relationship; and a coastal belt planning generation and updating module for spatially superimposing the habitat evolution type and the ecological development conflict type to form an adaptive access rule, generating a planning layer based on the adaptive access rule, and generating differentiated planning suggestions according to different partitions of the planning layer.

[0064] In the embodiment, by integrating the four types of core capabilities of multi-source time-series data processing, habitat suitability evolution analysis, development hotspot identification and ecological conflict representation, and planning partition generation into the same platform, a complete closed-loop process of data, model, evaluation and planning is realized. Not only can the planning result be automatically generated in time-series, space and integration, but also the differentiated management suggestions can be output according to the partition, so that the coastal belt planning is transformed from a static and empirical decision-making system to a dynamic decision-making system based on data driving and ecological process description, and the scientificity, accuracy and adaptability of the planning are greatly improved.

[0065] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting; it is not intended to exclude myriad other embodiments of the present application that other present or future devices, platforms, technologies, and methodologies can utilize as appropriate for particular situations. For example, the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items, as well as the lack of combinations when interpreted in the alternative ("or"). Similarly, the terms "comprises," "comprising," "includes," "including," and the like can be used herein, and each means inclusion without restriction, but also permit additional unspecified elements within the stated collection. The use of "including" and "comprising" and variations thereof is meant to encompass the items listed thereafter and any subsequent replacement of the word "comprising" with the word "consisting" is not intended to be a limitation. The terms "consisting essentially of and "consisting of" are used herein to indicate that the scope of the invention includes the listed elements, and does not exclude the presence of additional elements that do not materially affect the character or function of the invention.

[0066] The preferred embodiments of the application disclosed above are only to help explain the principles of the present application. The preferred embodiments do not limit the present application to only the specific embodiments described. As those skilled in the art understand from the content of the present specification, many modifications and variations of the present application are possible. The present specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present application, so that those skilled in the art can well understand and utilize the present application. The present application is limited only by the claims and their full scope and equivalents.

Claims

1. A coastal zone planning method based on MaXent and hotspot analysis, characterized in that, Includes the following steps: S1 collects environmental variable data, species distribution data, and development intensity data to construct a unified spatiotemporal grid system. Through coordinate correction, anomaly removal, interpolation completion, trajectory denoising, and intensity integration, it generates structured time-series data. S2, based on environmental variable data and species distribution data, construct a quarterly time-series maximum entropy model, and combine the quarterly time-series maximum entropy model to conduct a time-series assessment of the habitat status of each grid, and form a habitat evolution type based on the time-series assessment results; S3, based on the development intensity data, identify significant development hotspots in time series, and calculate the persistence and intensity characteristics of significant hotspots in each grid. Combining the time-series assessment results, the persistence of the significant hotspots, and the intensity characteristics of the development hotspots, a comprehensive characterization of ecological development relationships is constructed, and the types of ecological development conflicts are identified based on the comprehensive characterization of ecological development relationships. S4. Spatially overlay habitat evolution types and ecological development conflict types to form adaptive access rules. Generate a planning layer based on the adaptive access rules, and generate differentiated planning suggestions according to different partitions of the planning layer. The specific process of spatially overlaying habitat evolution types and ecological development conflict types to form adaptive access rules, generating planning layers based on these adaptive access rules, and generating differentiated planning suggestions according to different partitions of the planning layers is as follows: Based on habitat evolution types and ecological development conflict types, spatial overlay is performed to construct a temporally adaptive access rule system, which includes: ecological protection zones, ecological fine regulation zones, ecological development and balanced utilization zones, seasonal special utilization zones, and development-suitable zones. Based on a time-adaptive admission rule system, spatial clustering analysis is performed to identify spatially continuous regions of similar planning units. Regional connectivity analysis and raster vectorization are used to transform the raster results into spatial areas. A boundary smoothing algorithm is then used to generate continuous and smooth planning zoning boundaries, resulting in a coastal zone planning zoning layer. Corresponding regional planning recommendations are generated based on different zoning types, including: for ecological protection zones, recommendations on restricted development, ecological restoration, strict protection, and enclosure management; for ecological fine-control zones, recommendations on development limits, activity frequency control, and enhanced ecological monitoring; for ecological development and balanced utilization zones, recommendations on utilization intensity limits, eco-friendly facility deployment, and total development control; for seasonal special utilization zones, recommendations on seasonal opening, restrictions, and regulation; and for suitable development zones, recommendations on allowing regular development and infrastructure deployment, along with environmental constraints to be met during development. These regional planning recommendations are then written into the coastal zone planning database along with the coastal zone planning zoning layer.

2. The coastal zone planning method based on MaXent and hotspot analysis according to claim 1, characterized in that, The specific process of collecting environmental variable data, species distribution data, and development intensity data, constructing a unified spatiotemporal raster system, and generating structured time-series data through coordinate correction, anomaly removal, interpolation completion, trajectory denoising, and intensity integration is as follows: Collect environmental variable data, species distribution data, and development intensity data; Environmental variable data include: sea surface temperature, salinity, turbidity, seawater monitoring indicators, water depth, slope, and bottom sediment type; species distribution data include: species patrol records, species sampling points, and typical habitat distribution; development intensity data include: ship AIS tracks, fishing vessel operation tracks, number of operation days, tourist flow in scenic areas, and population density. The study area was divided into grid cells with a fixed spatial resolution, and quarterly time slices were constructed with a uniform time step. Environmental variable data, species distribution data, and development intensity data were mapped to the corresponding grid cells and time slices. For environmental variable data, image projection transformation and raster resampling are performed based on a unified coordinate system. Outliers are removed for sea surface temperature, salinity, turbidity and seawater monitoring indicators using the median absolute deviation method and quantile threshold. Linear interpolation and moving average are used to complete missing data and smooth the time series. For species distribution data, coordinate projection and timestamp alignment are performed, patrol records and species samples are mapped to corresponding raster cells through nearest neighbor interpolation, and duplicate records and drift points are removed; For development intensity data, trajectory denoising and speed threshold filtering are performed on ship AIS trajectories and fishing vessel operation trajectories. Ship activity intensity and fishing operation intensity are aggregated in grid cells according to point frequency. Temporal resampling and spatial projection are uniformly processed for scenic area visitor flow and population density to generate tourism activity intensity and shoreline population intensity. The arithmetic mean of the normalized ship activity intensity, fishing operation intensity, tourism activity intensity and shoreline population intensity is calculated in each grid cell and quarterly time slice to obtain the comprehensive development intensity value of the grid cell in the quarter. z-score normalization was performed on environmental variable data, species distribution data, and development intensity data; a coastal zone planning database was established, and the original and preprocessed environmental variable data, species distribution data, and development intensity data were written into the coastal zone planning database.

3. The coastal zone planning method based on MaXent and hotspot analysis according to claim 1, characterized in that, The specific process of constructing the quarterly time-series maximum entropy model based on environmental variable data and species distribution data is as follows: Using preprocessed environmental variable data and species distribution data from quarterly time slices as input, environmental variable data in each time slice are used as continuous features, and current species distribution data are used as positive samples to construct a comparison sample set between occurrence points and background points. Based on the comparison sample set, a quarterly time-series maximum entropy model is trained using the maximum entropy algorithm to obtain the habitat suitability probability of each grid cell in the corresponding time slice, and outputs the habitat suitability probability sequence.

4. The coastal zone planning method based on MaXent and hotspot analysis according to claim 1, characterized in that, The specific process of using the quarterly time-series maximum entropy model to perform time-series assessment of the habitat status of each grid is as follows: Using the habitat suitability probability sequence of each grid cell as input, the mean and standard deviation are calculated to obtain the time-series average suitability value and suitability standard deviation of the grid cell; by performing univariate linear regression in the coordinate system of quarterly sequence and habitat suitability probability, the trend of habitat suitability probability over time is fitted, and the slope of the regression line is recorded as the habitat suitability trend slope. The relative volatility is obtained by dividing the standard deviation of suitability by the sum of the time-series average suitability value and the minimum constant. The relative volatility is then subtracted from the constant and multiplied by the time-series average suitability value to obtain the stability suitability strength value. Calculate the arctangent of the slope of habitat suitability trend and divide it by half of pi to obtain the normalized trend value; Add a constant one to the normalized trend value and multiply it by the stability suitability strength value to obtain the temporal habitat stability evolution value.

5. The coastal zone planning method based on MaXent and hotspot analysis according to claim 1, characterized in that, The specific process for forming habitat evolution types based on time-series assessment results is as follows: Calculate the temporal habitat stability evolution value for each grid cell, and map each grid cell to a two-dimensional feature space of stability and trend using the temporal habitat stability evolution value as the stability axis and the normalized trend value as the change trend axis. Compare the time-series habitat stable evolution value with the multi-level stability thresholds H1 and H2, and compare the normalized trend value with the multi-level trend thresholds T1 and T2 to determine the habitat evolution type: When ≥H2 and ≥T1, it is determined as a suitable area for steady-state gain; When H1 ≤ <H2 and T1 ≤ <T2, it is determined as a suitable area for seasonal steady state; When H1 ≤ <H2 and <T1, or <H1 and T1 ≤ <T2, it is determined as a suitable fluctuation area; When <H1 and <T1, it is determined as a suitable degradation area; Write the time-series habitat stable evolution value and the corresponding habitat evolution type into the coastal zone planning database.

6. The coastal zone planning method based on MaXent and hotspot analysis according to claim 1, characterized in that, The specific process of identifying significant development hotspots in time series based on the development intensity data, and calculating the persistence and intensity characteristics of significant hotspots in each grid is as follows: Using the intensity of ship activity, fishing operations, tourism activities, and shoreline pedestrian flow under the preprocessed quarterly time slices as input, the development intensity sequence is organized into grid cells according to the quarterly time sequence. Under the constraint of a unified spatial weight matrix, the Getis-Ord Gi* local spatial autocorrelation statistical method is used to perform hotspot analysis on the development intensity sequence of each quarter. The significance value of development hotspots in each grid cell in each quarter is calculated, and the time slices with development hotspot significance values ​​greater than the significance threshold are marked as significant development hotspot moments. Using the development hotspot significance value sequence of each grid cell as input, the ratio of significant hotspot duration of the grid cell is obtained by dividing the number of significant development hotspot moments by the total number of time steps in all quarterly time steps. For each grid cell, within the corresponding time steps, the significance value of the development hotspot corresponding to the moment of significant development hotspot is included in the cumulative value of hotspot intensity, and the moment of significant development hotspot is included in the significant hotspot duration count. After traversing all time steps, the cumulative value of hotspot intensity is divided by the sum of the significant hotspot duration count and the minimum constant to obtain the average hotspot intensity value of the grid cell. The average hotspot intensity value is added to a constant and the natural logarithm is taken. Then, it is multiplied by the ratio of the duration of significant hotspots to obtain the significant hotspot duration intensity value of the grid cell.

7. The coastal zone planning method based on MaXent and hotspot analysis according to claim 1, characterized in that, The specific process of constructing a comprehensive characterization of ecological development relationships by combining the time-series assessment results, the persistence of significant hotspots, and the intensity characteristics of development hotspots is as follows: The time series of the comprehensive development intensity value of each grid cell is received. A univariate linear regression is performed in the coordinate system of quarter number and comprehensive development intensity value. The regression slope that represents the trend of the comprehensive development intensity value over time is recorded as the comprehensive development intensity time trend slope. Obtain temporal habitat stability evolution values, habitat suitability trend slopes, and significant hotspot persistence intensity values; Multiply the slope of habitat suitability trend by the slope of comprehensive development intensity over time, substitute the result into the hyperbolic tangent function, calculate the trend coupling value, and then perform nonnegation processing on the trend coupling value. The overall ecological development conflict value of the grid cell is obtained by multiplying the trend coupling value with the non-negative temporal habitat stability evolution value and the non-negative significant hotspot persistence intensity value.

8. The coastal zone planning method based on MaXent and hotspot analysis according to claim 1, characterized in that, The specific process for identifying types of ecological development conflicts based on the comprehensive characterization of ecological development relationships is as follows: Calculate the comprehensive ecological development conflict value of each grid cell , and compare it with the multi-level conflict thresholds C1 and C2 to identify the types of ecological development conflicts; when < C1, it is determined as an area available for conventional development; when C1 ≤ < C2, it is determined as an ecological sensitive regulation area; when ≥ C2, it is determined as an ecological conflict control area; The comprehensive ecological development conflict value and the corresponding ecological development conflict type are entered into the coastal zone planning database.

9. A coastal zone planning system based on MaXent and hotspot analysis, employing the coastal zone planning method based on MaXent and hotspot analysis as described in any one of claims 1-8, characterized in that, include: The multi-source time-series data acquisition and preprocessing module is used to collect environmental variable data, species distribution data, and development intensity data, construct a unified spatiotemporal grid system, and generate structured time-series data through coordinate correction, anomaly removal, interpolation completion, trajectory denoising, and intensity integration. The temporal habitat suitability evolution analysis module is used to construct a quarterly temporal maximum entropy model based on environmental variable data and species distribution data, and to conduct a temporal assessment of the habitat status of each grid based on the quarterly temporal maximum entropy model, and to form the habitat evolution type based on the temporal assessment results. The temporal hotspot identification and ecological conflict analysis module is used to identify significant temporal development hotspots based on the development intensity data, and to calculate the persistence and intensity characteristics of significant hotspots in each grid. Combining the time-series assessment results, the persistence of the significant hotspots, and the intensity characteristics of the development hotspots, a comprehensive characterization of ecological development relationships is constructed, and the types of ecological development conflicts are identified based on the comprehensive characterization of ecological development relationships. The coastal zone planning generation and update module is used to spatially overlay habitat evolution types and ecological development conflict types to form adaptive access rules. Based on the adaptive access rules, a planning layer is generated, and differentiated planning suggestions are generated according to different partitions of the planning layer.

Images