Prediction Method for the Impact of Climate Change on Biodiversity Based on Deep Learning

Abstract

This application belongs to the field of environmental prediction technology and involves a method for predicting the impact of climate change on biodiversity based on deep learning. In this invention, climate data is converted into a 4D tensor, and depthwise separable convolution is used to extract multi-scale spatio-temporal features and capture spatio-temporal dependence relationships; then, the kernel density estimation method is used to convert biodiversity data into a density field, and a dynamic heterogeneous graph is obtained through a heterogeneous graph attention network. Then, by constructing a transmembrane state hypergraph and performing spatio-temporal projection, effective alignment of climate data and biodiversity data is achieved. A multi-modal uncertainty propagation mechanism is introduced, and a three-dimensional heatmap of joint uncertainty is output based on the multi-modal uncertainty propagation mechanism. Finally, based on the spatio-temporally aligned climate and biological data and the heatmap of joint uncertainty, a prediction result of species distribution is generated, which not only improves the prediction accuracy of the impact of climate change on biodiversity but also enhances the interpretability of the prediction result.

Description

Technical Field

[0001] This application belongs to the technical field of environmental prediction, and more specifically, it relates to a method for predicting the impact of climate change on biodiversity based on deep learning. Background Art

[0002] The impact of climate change is one of the key topics in global environmental research. Biodiversity is an important indicator of ecosystem health, and its change is closely related to climate conditions. Therefore, accurately predicting the impact of climate change on biodiversity is of great significance for environmental protection, ecological restoration, and species conservation.

[0003] However, the impact of climate change on biodiversity is a multi-factor and dynamically changing process; climate data is usually time-series and has a complex distribution pattern in space; while biodiversity data is usually observed at specific spatial location points, with strong locality and irregularity; this difference makes it a major challenge to effectively align the two and conduct joint analysis. Summary of the Invention

[0004] The present invention provides a method for predicting the impact of climate change on biodiversity based on deep learning, aiming to solve the technical problem of the difficulty in effectively aligning climate data and biodiversity data.

[0005] The method for predicting the impact of climate change on biodiversity based on deep learning includes the following steps:

[0006] Climate data collection and processing: Collect climate data, convert the collected climate data into a 4D tensor, and use depthwise separable convolution to extract multi-scale spatio-temporal features to obtain a climate data tensor;

[0007] Biodiversity data collection and processing: Obtain biodiversity data based on species observation points, convert it into a dynamic spatial density field using kernel density estimation, and then build a heterogeneous graph attention network for modeling to obtain a dynamic heterogeneous graph of species;

[0008] Spatio-temporal alignment: Based on the climate data tensor and the dynamic heterogeneous graph, build a cross-modal hypergraph and perform spatio-temporal projection to obtain spatio-temporally aligned climate and biological data;

[0009] Multi-modal uncertainty propagation: Model the climate prediction error based on a neural network differentiator, and at the same time separate the noise of biological observation data through a double-branch feature extractor to obtain observation noise from different sources, and then fuse the covariance of the climate prediction error and the biological observation noise to output a three-dimensional heat map of joint uncertainty;

[0010] Dynamic feedback prediction architecture: Based on spatio-temporal aligned climate and biological data and three-dimensional heat maps, a multi-layer spatio-temporal graph structure is constructed, processed by a spatio-temporal graph attention network, and species distribution prediction results are generated.

[0011] In the present invention, climate data is converted into a 4D tensor, and depthwise separable convolution is used to extract multi-scale spatio-temporal features and capture spatio-temporal dependencies; then, the kernel density estimation method is used to convert biodiversity data into a density field, and a dynamic heterogeneous graph is obtained through a heterogeneous graph attention network. Then, by constructing a transmembrane state hypergraph and performing spatio-temporal projection, effective alignment of climate data and biodiversity data is achieved, enabling the two to be jointly modeled and predicted within the same spatio-temporal framework. In addition, the present invention also introduces a multi-modal uncertainty propagation mechanism, and based on the multi-modal uncertainty propagation mechanism, a three-dimensional heat map of joint uncertainty is output, further improving the credibility and accuracy of the prediction. Finally, based on the spatio-temporal aligned climate and biological data and the heat map of joint uncertainty, a multi-layer spatio-temporal graph structure is constructed, and a spatio-temporal graph attention network is used for processing to generate prediction results of species distribution, which not only improves the prediction accuracy of the impact of climate change on biodiversity but also enhances the interpretability of the prediction results.

[0012] Preferably, the steps for obtaining the climate data tensor include the following:

[0013] Depth convolution: 3D convolution is used based on the 4D tensor to extract local spatio-temporal features in the climate data, where each channel is convolved separately;

[0014] Point convolution: 1D convolution is used to fuse the channels of the features extracted by the depth convolution to obtain a spatio-temporal feature map;

[0015] Multi-scale convolution: The first convolution kernel is used to extract the first feature; the second convolution kernel is used to extract the second feature, where the scale of the second convolution kernel is larger than that of the first convolution kernel; then, a concatenation operation is performed on the first feature and the second feature to obtain a multi-scale feature map;

[0016] Vertical correlation modeling: Based on the multi-scale feature map, the attention weights between each pair of height layers are calculated, and the features of different height layers are weighted and fused based on the calculated attention weights to obtain a new height-correlated feature map;

[0017] Spatio-temporal feature encoding: The multi-scale feature map and the height-correlated feature map are input into a GRU, and the gating mechanism is used to control the information flow to capture the dynamic patterns in time, and finally, the climate data tensor is output.

[0018] Preferably, the steps for processing the biodiversity data include:

[0019] Dynamic density conversion and gradient calculation: According to the species occurrence point data, through the kernel density estimation algorithm, the discrete observation points are converted into a continuous species density field. In the kernel density estimation, the spatial and temporal effects are considered during the calculation, and the Gaussian kernel function and the time decay function are used to obtain a dynamically changing species density value, that is, the species density field. Based on the calculated species density field, the spatial gradient of the species density field is extracted using the spatial gradient calculation method.

[0020] Spatial graph structure construction: Define each spatial grid point as a node of the graph. The features of the node include the species density value, the magnitude and direction of the gradient, the terrain features, the accuracy and coverage radius of the observation equipment. If the gradient directions of the species density fields of two nodes are similar, an edge connection is established between the two nodes. For each adjacent node, the weight of the edge is determined based on the gradient similarity, the geographical distance, and the equipment accuracy, based on the obtained constructed spatial graph structure.

[0021] Heterogeneous graph attention network: Map the features of each node through a trainable linear layer, and then use the multi-head heterogeneous graph attention mechanism to calculate the attention coefficients for each node and its adjacent nodes. Then, based on the adjacent nodes of each node and the calculated attention coefficients, information is weighted and fused to update the node features.

[0022] Preferably, the spatio-temporal alignment includes the following steps:

[0023] Construct a climate hypergraph: For each climate grid, extract all climate grid points within its spatio-temporal neighborhood to form a hyperedge, and then aggregate the neighborhood climate features through 3D convolution to obtain the hyperedge feature representation of the climate.

[0024] Construct a biological hypergraph: For each biological node, select its Top-k neighbor nodes according to the diffusion probability to form a hyperedge, and aggregate the features of the node and its neighbor nodes through the multi-head graph attention mechanism to obtain the hyperedge feature representation of the biology.

[0025] Transmembrane state hypergraph connection: Connect the climate grid points and the biological nodes. When both the spatial distance and the temporal difference meet the predetermined conditions, the climate grid points and the biological nodes share a transmembrane state hyperedge.

[0026] Projection of climate data to biological data: For each biological node, find the climate grid points within its spatial neighborhood, calculate the projection weight of each climate grid point on the biological node, where the projection weight considers the similarity of climate features and biological features and introduces geographical distance constraints for calculation. Based on the calculated projection weight combined with the climate grid point features, the climate feature representation on the biological node is obtained.

[0027] Projection of biological data onto climate data: For each climate grid point, identify the biological nodes within its spatial neighborhood, and perform weighted aggregation of the biological node features using diffusion probabilities to obtain the biological feature representation at the climate grid point.

[0028] Preferably, modeling climate prediction errors based on a neural network differentiator includes:

[0029] 3D convolution: Using 3D convolution on the climate data tensor to extract local spatio-temporal features, and adding an activation function after the convolutional layer to increase non-linearity;

[0030] Fully connected layer: Based on the features obtained after the 3D convolution operation, using a fully connected layer to process the features and output the error prediction result;

[0031] Error covariance calculation: Based on the output error prediction result, calculate the covariance matrix of the errors to describe the error correlation between different spatial points.

[0032] Preferably, the dual-branch feature extractor includes two branches;

[0033] The first branch is used to extract the true biological information of the observed data; the second branch is used to extract features related to noise;

[0034] Calculate the covariance matrix of the biological observation noise based on the extracted noise-related features to represent the noise correlation.

[0035] Preferably, the dynamic feedback prediction architecture is as follows:

[0036] Construct a multi-layer spatio-temporal graph structure: Construct it based on the spatio-temporally aligned climate and biological data, where the spatio-temporal graph structure includes a spatial graph and a temporal graph. Each spatial point in the spatial graph is regarded as a node in the graph, and the edges between the nodes represent the spatial adjacency relationship; each time point in the temporal graph is regarded as a node in the graph, and the edges between the nodes represent the evolutionary relationship in time; Obtain spatio-temporal graph data based on the constructed multi-layer spatio-temporal graph structure;

[0037] Spatio-temporal graph attention network processing: Using the spatio-temporal graph data and the three-dimensional heat map of uncertainty as the input to the spatio-temporal attention network, assign different importance weights to each spatio-temporal node through the spatio-temporal attention mechanism, and update the features of each spatio-temporal node based on the calculated importance weights;

[0038] Species distribution prediction: Based on the features of the updated out-of-control nodes, use a species distribution prediction network to predict the distribution of species.

[0039] The beneficial effects of the present invention include:

[0040] The present invention converts climate data into a 4D tensor, and uses depthwise separable convolution to extract multi-scale spatio-temporal features and capture spatio-temporal dependencies; then uses the kernel density estimation method to convert biodiversity data into a density field, and then obtains a dynamic heterogeneous graph through a heterogeneous graph attention network. Then, by constructing a transmembrane state hypergraph and performing spatio-temporal projection, the effective alignment of climate data and biodiversity data is achieved, enabling the two to be jointly modeled and predicted within the same spatio-temporal framework. In addition, the present invention also introduces a multi-modal uncertainty propagation mechanism, and outputs a three-dimensional heat map of joint uncertainty based on the multi-modal uncertainty propagation mechanism, further improving the credibility and accuracy of the prediction. Finally, based on the spatio-temporally aligned climate and biological data and the heat map of joint uncertainty, a multi-layer spatio-temporal graph structure is constructed, and a spatio-temporal graph attention network is used for processing to generate the prediction results of species distribution, which not only improves the prediction accuracy of the impact of climate change on biodiversity, but also enhances the interpretability of the prediction results. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0042] Figure 1 It is a block diagram of the overall steps provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present application more clearly understood, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0044] See Figure 1 As shown, a method for optimizing the energy efficiency of a diving device based on environmental data prediction includes the following steps:

[0045] A method for predicting the impact of climate change on biodiversity based on deep learning includes the following steps:

[0046] Climate data collection and processing: Collect climate data, convert the collected climate data into a 4D tensor, and use depthwise separable convolution to extract multi-scale spatio-temporal features to obtain a climate data tensor;

[0047] Exemplarily, climate data with spatial and temporal dimensions, including meteorological variables such as temperature, humidity, precipitation, wind speed, and air pressure, are obtained through means such as climate sensors, satellite remote sensing, or meteorological stations. Assume the obtained climate data matrix is C, with dimensions [T, H, W, F], where T represents the time step (e.g., monthly or annually), H represents the number of spatial height layers; W represents the spatial width (longitude and latitude dimensions); F represents the types of climate variables (e.g., temperature, humidity, precipitation, etc.);

[0048] Depth convolution: Based on the obtained climate data matrix, 3D convolution is used for local spatio-temporal feature extraction, where the 3D convolution operation is expressed as:

[0049] ;

[0050] In the formula: represents the convolution kernel, and the convolution operation extracts spatio-temporal features by applying the convolution kernel. Each channel is convolved separately, that is, the convolution operation is applied to each climate variable (such as temperature, humidity, etc.) separately to obtain multiple output feature maps;

[0051] Point convolution: 1D convolution is used to fuse the features obtained from depth convolution. Let the output of each channel be , and the 1D convolution operation is expressed as:

[0052] ;

[0053] In the formula: represents the 1D convolution kernel, and the convolution operation fuses along the time or space dimension to generate a spatio-temporal feature map;

[0054] Multi-scale convolution: To further capture features at different scales, two different-scale convolution kernels are used to extract features respectively:

[0055] Based on the output result of point convolution, the first convolution kernel and the second convolution kernel are used to extract the first feature and the second feature respectively, where the scale of the second convolution kernel is larger than that of the first convolution kernel, and then the first feature and the second feature are concatenated to obtain a multi-scale feature map ;

[0056] Vertical correlation modeling: Calculate the attention weights between each pair of height layers on the multi-scale feature map , and based on these weights, weightedly fuse the features of different height layers. Assume the attention weight matrix between height layers is A, and by multiplying the weight matrix by the multi-scale feature map, a height-correlated feature map is obtained;

[0057] The steps to calculate the attention weight matrix are as follows:

[0058] For multi-scale feature maps Perform a linear transformation to obtain query, key, and value representations for each height level. Use a learnable weight matrix to project each height level in the multi-scale feature map respectively to obtain the query, key, and value;

[0059] Calculate the similarity between the query vector of each pair of height levels and all key vectors. The dot product can be used to measure the correlation between the query and the key; then normalize the calculated correlation values to obtain the attention weights corresponding to each query; based on this, obtain the attention weight matrix;

[0060] Spatio-temporal feature encoding: Input the multi-scale feature map and the height-correlated feature map into a GRU (Gated Recurrent Unit) for spatio-temporal feature encoding to obtain a 4D climate data tensor.

[0061] Biodiversity data collection and processing: Obtain biodiversity data based on species observation points, convert it into a dynamic spatial density field using kernel density estimation, and then model it through constructing a heterogeneous graph attention network to obtain a dynamic heterogeneous graph of species;

[0062] In this embodiment, the collection of biodiversity data mainly relies on the observation point data of species. The observation point data mainly includes the occurrence location of the species and the corresponding observation time. The observation points of each species include: spatial coordinates (longitude, latitude), timestamp (observation time), and species identifier (species information corresponding to each observation point);

[0063] Since the collected biodiversity data is discrete data and cannot be directly used for modeling, further processing is required. In this embodiment, the collected discrete data is converted into a species density field, and the specific steps are as follows:

[0064] In this embodiment, the discrete data is converted into a species density field through the kernel density estimation method. By applying a kernel function (such as a Gaussian kernel) around each observation point, the species density at each location is smoothly estimated. Exemplary:

[0065] Given a set of species observation point data , where represents the spatial coordinates of observation point i, that is, longitude and latitude; the observation time of observation point i; represents the species identifier corresponding to observation point i, and the kernel density estimation function is as follows:

[0066] ;

[0067] ;

[0068] In the formula: represents the species density at position x and time t; N represents the total number of observation points; h represents the bandwidth parameter, controlling the size of the kernel; represents the Gaussian kernel function; represents the time decay function; represents the decay factor; where x and t are the specific spatial and temporal points of the species density to be estimated;

[0069] The gradient operator is used to take the derivative of the species density field to obtain the spatial gradient of the species density field:

[0070] ;

[0071] In the formula: represents the species density field The gradients in the x and y directions of the spatial coordinates, using the gradient to describe the distribution change of the species in space, and further revealing the response pattern of the species to climate change;

[0072] Spatial graph structure construction: Each grid point is used as a node of the graph, and the features of the node include:

[0073] The species density value, the species density calculated by KDE ; The modulus and direction extracted in space, the modulus and direction calculated by the spatial gradient ; Terrain data, as an additional feature of the node; the accuracy and coverage radius of the observation equipment;

[0074] The edge connection rule of the graph is based on the following criteria:

[0075] If the gradient directions of two nodes are similar, an edge connection is established;

[0076] For each adjacent node, the weight of the edge is calculated based on the following factors:

[0077] First, calculate the similarity of the gradient directions of the two nodes :

[0078] ;

[0079] In the formula: represents the species density field of node i The gradients in the x and y directions of the spatial coordinates; represents the species density field of node j The gradients in the x and y directions of the spatial coordinates;

[0080] Geographical distance calculation: The Haversine formula is used to calculate the distance between two points on the earth's surface;

[0081] Based on the calculated similarity and geographical distance, the comprehensive weight is calculated by means of weighted summation.

[0082] Heterogeneous graph attention network: Based on the constructed spatial graph structure, the heterogeneous graph attention network is used to process the node features. The specific steps are as follows:

[0083] In order to enable the neural network to better process the feature information of the nodes, the original features of each node are mapped to a new feature space through linear transformation by means of node feature mapping. The specific steps are as follows:

[0084] Given the feature vector of each node i , it is mapped to a new feature space through a linear transformation (such as a trainable weight matrix W) to obtain a transformed node feature :

[0085] ;

[0086] In the formula: W represents the trainable weight matrix; represents the original feature of node i; b represents the bias term;

[0087] Based on this, the multi-head attention mechanism is further applied for processing. The interaction between each node and its adjacent nodes is realized by calculating the attention coefficient . The attention coefficient represents the correlation between node i and adjacent node j, and based on this, the features of adjacent nodes are weighted and fused. How to calculate the attention coefficient belongs to the conventional technical means in the field, so it will not be elaborated in this embodiment;

[0088] When the attention coefficient is calculated, the attention coefficient will be used to weight and fuse the information of each adjacent node. By summing the features of adjacent nodes with weights, the feature of the current node i is updated.

[0089] Based on the updated node features, we will obtain the final updated output dynamic heterogeneous graph of the species.

[0090] Spatiotemporal alignment: Based on the climate data tensor and the dynamic heterogeneous graph, a transmembrane state hypergraph is constructed and spatiotemporal projection is performed to obtain spatiotemporally aligned climate and biological data;

[0091] As a possible implementation manner of this embodiment, the spatiotemporal alignment includes the following steps:

[0092] In this step, in order to convert the climate data and biodiversity data into a unified spatiotemporal framework so that the climate data (4D tensor) and biological data (dynamic heterogeneous graph) can correspond to each other and be analyzed under the same spatiotemporal pattern, the specific steps are as follows:

[0093] Constructing a climate hypergraph: For each climate grid point , extract all climate grid points in the spatio-temporal neighborhood around the climate grid point . Assuming the time interval of the grid is , and the spatial interval is , select a domain range R, where R is the spatio-temporal radius; all grid points within this neighborhood form a hyperedge, connecting the grid points within the domain range R;

[0094] 3D Convolution Aggregation: Aggregate the climate features within the spatio-temporal neighborhood of the climate grid points to obtain the feature representation of the climate hyperedge. Specifically, a 3D convolution operation is used to aggregate the climate features in the spatio-temporal neighborhood, and the expression form is as follows:

[0095] ;

[0096] In the formula: represents the climate data in the spatio-temporal neighborhood around the climate grid point ; represents the 3D convolution operation, which performs convolution in the three dimensions of time, latitude, and longitude to aggregate the climate features in the neighborhood; represents the feature representation of the climate hyperedge;

[0097] Construction of the climate hypergraph: Each climate grid point is connected to other climate grid points within its domain by hyperedges to obtain the climate hypergraph.

[0098] Constructing a biological hypergraph: The biological hypergraph is used to model biodiversity data, focusing on the relationships between biological nodes, and aggregating the features of neighbors through an attention mechanism. The specific steps are as follows:

[0099] For each biological node , select the Top-k neighbor nodes according to its diffusion probability to form a hyperedge, where the diffusion probability is expressed as follows:

[0100] ;

[0101] In the formula: represents the diffusion probability between the climate grid and the biological node ; represents the geographical distance between the climate grid point and the biological node ; represents a hyperparameter that controls the diffusion range; represents the maximum number of neighbors; represents the climate grid point The geographical distance between biological nodes ;

[0102] Feature aggregation of biological nodes and their neighbor nodes through the multi-head attention mechanism:

[0103] ;

[0104] Where: and respectively represent the feature representations of climate grid points and biological neighbor nodes ; represents the multi-head attention mechanism, which is used to aggregate multiple neighbor features;

[0105] Based on this, the biological node and its Top-k neighbors are connected by hyperedges to obtain a biological hypergraph.

[0106] In spatio-temporal alignment, the climate grid and biological nodes need to be connected under the same spatio-temporal framework. The transmembrane state hypergraph connection realizes the alignment of climate grid points and biological nodes. The specific steps are as follows:

[0107] Spatially, if the distance between two nodes is less than the set threshold , and temporally, the time difference between two nodes is less than the set time threshold , then a transmembrane state hyperedge is shared between the climate grid point and the biological node, where the spatial distance is calculated by the Haversine formula, and the time difference can be directly calculated as the difference between two time points; if both the spatial distance and the time difference meet the above conditions, then a transmembrane state hyperedge will be formed between the climate grid and the biological node .

[0108] Projection of climate data onto biological data: By calculating the projection weights, the climate data is mapped into the biological data, as follows:

[0109] For each biological node , calculate the projection weight of the climate grid points in its spatial neighborhood on this biological node. The formula for calculating the weight is as follows:

[0110] ;

[0111] Where: represents the geographical distance between the climate grid point and the biological node ; represents the geographical distance between the climate grid point and the biological node ;

[0112] Based on the calculated projection weights and combined with the characteristics of climate grid points , the climate feature representation on the biological nodes is obtained:

[0113] ;

[0114] In the formula: represents the climate feature representation on the biological nodes.

[0115] For each climate grid point , find the biological nodes in its spatial neighborhood, and weighted aggregate the features of the biological nodes through the diffusion probability:

[0116] ;

[0117] In the formula: represents the biological feature representation on the climate grid point.

[0118] In this embodiment, by constructing a climate hypergraph and a biological hypergraph, as well as the connection and projection of the transmembrane state hypergraph, the alignment of climate data and biodiversity data is achieved. This not only helps to integrate two different modal data into the same spatio-temporal framework, but also improves the alignment accuracy through weighted aggregation and the attention mechanism, providing a reliable data basis for subsequent predictions.

[0119] Multi-modal uncertainty propagation: Model the climate prediction error based on the neural network differentiator, and at the same time separate the noise of biological observation data through a two-branch feature extractor to obtain observation noise from different sources. Then, fuse the covariance of the climate prediction error and the biological observation noise to output a three-dimensional heat map of joint uncertainty;

[0120] As a possible implementation manner of this embodiment, the multi-modal uncertainty propagation includes the following steps:

[0121] Use a 3D convolutional layer to extract local spatio-temporal features. Assume that the neural network uses standard convolution operations, and the convolution kernel size is ( ), and the output feature map is , where , , represents the spatial and temporal dimensions after the convolution operation, , represents the high dimension of space and the wide dimension of space, such as longitude and latitude; represents the number of output channels;

[0122] Further process the convolution output through a fully connected layer to predict the error, where the calculation formula of the fully connected layer is as follows:

[0123] ;

[0124] Wherein: and represent the weights and biases of the fully connected layer; represents the operation of flattening the convolutional output; represents the predicted error;

[0125] Based on the output error calculate the covariance matrix of the error , which is used to describe the error correlation of different spatial points, and the calculation formula is as follows:

[0126] ;

[0127] Wherein: represents the number of samples; represents the mean value of the error; represents the i-th error value;

[0128] To process the noise of biodiversity data, a dual-branch feature extractor is adopted to extract real biological information and noise-related features respectively;

[0129] Among them, the first branch extracts the real biological information in the observed data through a convolutional layer or other neural network layers to obtain the features representing the real information of biodiversity , for example, the processing using a convolutional layer is as follows:

[0130] ;

[0131] Wherein: represents the activation function (such as ReLU, Sigmoid activation function); represents the weight matrix in the first branch; represents the biological observation data; represents the bias term of the first branch;

[0132] The second branch extracts the noise-related features , and calculates the covariance matrix of the biological observation noise. Similarly, the noise feature extraction process can also be performed using a convolutional layer to obtain the noise features ; and then calculate the covariance matrix of the noise based on this:

[0133] ;

[0134] Wherein: represents the covariance matrix of the noise features; M represents the number of samples; represents the noise feature of the i-th sample; Represents the mean of the noise characteristics; Represents the transpose operation.

[0135] Covariance fusion: The covariance matrix of the climate error and the covariance matrix of the biological observation noise are fused by weighted summation to obtain the joint covariance matrix ; Based on the joint covariance matrix generate a three-dimensional heat map as the output of uncertainty ; This three-dimensional heat map represents the joint uncertainty between climate change and biodiversity in the spatial and temporal dimensions.

[0136] Dynamic feedback prediction architecture: Based on the spatio-temporal aligned climate and biological data and the three-dimensional heat map, construct a multi-layer spatio-temporal graph structure, and generate species distribution prediction results based on the spatio-temporal graph attention network processing;

[0137] As a possible implementation of this embodiment, the dynamic feedback prediction architecture specifically includes the following steps:

[0138] Construct a multi-layer spatio-temporal graph structure based on the spatio-temporal aligned data, specifically including a spatial graph and a temporal graph. Each spatial point in the spatial graph is used as a node, and the edges between the nodes represent the adjacency relationship in space; for example, adjacent spatial points (such as adjacent pixels) can be connected by edges; Each time point in the temporal graph is used as a node, and the edges between the nodes represent the evolutionary relationship in time. For example, the edges between consecutive time points are connected;

[0139] The spatio-temporal graph structure combines the spatial graph and the temporal graph into a multi-layer spatio-temporal graph. The features of each node include both the spatial information and the temporal information of the node. The spatio-temporal data is represented by the following tensor:

[0140] ;

[0141] Where: N represents the number of spatial points (i.e., the number of nodes in the graph); T represents the number of time dimensions; D represents the feature dimension of each spatio-temporal node;

[0142] Spatio-temporal graph attention network processing: The input of the spatio-temporal graph attention network includes the spatio-temporal graph data , the three-dimensional heat map of the joint uncertainty ;

[0143] Among them, the goal of the spatio-temporal graph attention mechanism is to assign different importance weights to each spatio-temporal node according to the context information of each spatio-temporal node, use the attention mechanism to calculate the weights of each spatio-temporal node, and update the features of the nodes based on the calculated weights; Assume that the input matrix of the spatio-temporal graph network is , the output representation of the spatio-temporal attention network is as follows:

[0144] ;

[0145] Where: represents the spatio-temporal attention operation, which calculates weights and updates node features based on the context information (spatial neighborhood and temporal neighborhood) of each node and the uncertainty heat map. The specific calculation formula is as follows:

[0146] ;

[0147] In the formula: and are the feature vectors of spatio-temporal nodes i and j respectively; represents the feature transformation matrix; represents the feature concatenation operation; represents the transpose of the weight vector of the attention mechanism; represents the attention weight of node i to node j;

[0148] The species distribution prediction network predicts the species distribution through the temporal node features processed by the temporal attention network and the uncertainty three-dimensional heat map; among them, the species distribution prediction network can adopt a hybrid model of CNN and LSTM, which can process spatial and temporal information simultaneously; based on the given model and input data, how to use this model to achieve species distribution prediction belongs to the conventional technical means in the technical solution of this field, so it will not be described in detail here.

[0149] In the present invention, climate data is converted into a 4D tensor, and depthwise separable convolution is used to extract multi-scale spatio-temporal features and capture spatio-temporal dependence relationships; then, the kernel density estimation method is used to convert biodiversity data into a density field, and a dynamic heterogeneous graph is obtained through a heterogeneous graph attention network. Then, an effective alignment of climate data and biodiversity data is achieved by constructing a transmembrane state hypergraph and performing spatio-temporal projection, enabling the two to be jointly modeled and predicted in the same spatio-temporal framework. In addition, the present invention also introduces a multi-modal uncertainty propagation mechanism, and outputs a three-dimensional heat map of joint uncertainty based on the multi-modal uncertainty propagation mechanism, further improving the credibility and accuracy of the prediction. Finally, based on the spatio-temporally aligned climate and biological data and the heat map of joint uncertainty, a multi-layer spatio-temporal graph structure is constructed, and the spatio-temporal graph attention network is used for processing to generate the prediction result of the species distribution, which not only improves the prediction accuracy of the impact of climate change on biodiversity, but also enhances the interpretability of the prediction result.

[0150] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. A method for predicting the impact of climate change on biodiversity based on deep learning, characterized in that It includes the following steps: Climate data collection and processing: Collect climate data, convert the collected climate data into a 4D tensor, and use depthwise separable convolution to extract multi-scale spatio-temporal features to obtain a climate data tensor; Biodiversity data collection and processing: Obtain biodiversity data based on species observation points, convert it into a dynamic spatial density field using kernel density estimation, and then model it through constructing a heterogeneous graph attention network to obtain a dynamic heterogeneous graph of species; Spatio-temporal alignment: Based on the climate data tensor and the dynamic heterogeneous graph, construct a cross-modal hypergraph and perform spatio-temporal projection to obtain spatio-temporally aligned climate and biological data; The spatio-temporal alignment includes the following steps: Construct a climate hypergraph: For each climate grid, extract all climate grid points within its spatio-temporal neighborhood to form a hyperedge, and then aggregate the neighborhood climate features through 3D convolution to obtain a hyperedge feature representation of the climate; Construct a biological hypergraph: For each biological node, select its Top-k neighbor nodes according to the diffusion probability to form a hyperedge, and aggregate the features of the node and its neighbor nodes through a multi-head graph attention mechanism to obtain a hyperedge feature representation of the biology; Cross-modal hypergraph connection: Connect the climate grid points and biological nodes. When both the spatial distance and the temporal difference meet the predetermined conditions, the climate grid points and biological nodes share a cross-modal hyperedge; Projection of climate data to biological data: For each biological node, find the climate grid points within its spatial neighborhood, calculate the projection weight of each climate grid point on the biological node, where the projection weight considers the similarity of climate features and biological features and introduces a geographical distance constraint for calculation; Combine the calculated projection weight with the climate grid point features to obtain a climate feature representation on the biological node; Projection of biological data to climate data: For each climate grid point, find the biological nodes within its spatial neighborhood, and perform weighted aggregation of the biological node features through the diffusion probability to obtain a biological feature representation on the climate grid point; Multi-modal uncertainty propagation: Model the climate prediction error based on a neural network differentiator, and at the same time separate the noise of biological observation data through a double-branch feature extractor to obtain observation noises from different sources, and then fuse the covariance of the climate prediction error and the biological observation noise to output a three-dimensional heat map of joint uncertainty; Dynamic feedback prediction architecture: Based on the spatio-temporally aligned climate and biological data and the three-dimensional heat map, construct a multi-layer spatio-temporal graph structure, and process it based on the spatio-temporal graph attention network to generate a species distribution prediction result.

2. The method for predicting the impact of climate change on biodiversity based on deep learning according to claim 1, wherein The steps for obtaining the climate data tensor include the following: Depth convolution: Use 3D convolution based on the 4D tensor to extract local spatio-temporal features in the climate data, where each channel is convolved separately; Point convolution: Perform channel fusion on the features extracted by the depth convolution through 1D convolution to obtain a spatio-temporal feature map; Multi-scale convolution: Use a first convolution kernel to extract a first feature; use a second convolution kernel to extract a second feature, where the scale of the second convolution kernel is larger than that of the first convolution kernel; then perform a concatenation operation on the first feature and the second feature to obtain a multi-scale feature map; Vertical correlation modeling: Based on multi-scale feature maps, calculate the attention weights between each pair of height layers, and based on the calculated attention weights, weight and fuse the features of different height layers to obtain a new height-correlated feature map; Spatio-temporal feature encoding: Input the multi-scale feature maps and the height-correlated feature maps into a GRU, and use the gating mechanism to control the information flow to capture the dynamic patterns in time, and finally output the climate data tensor.

3. The method for predicting the impact of climate change on biodiversity based on deep learning according to claim 1, wherein The steps for processing the biodiversity data include: Dynamic density conversion and gradient calculation: According to the species occurrence point data, through the kernel density estimation algorithm, convert the discrete observation points into a continuous species density field. In the calculation of kernel density estimation, the influence of space and time is considered, and the Gaussian kernel function and the time decay function are used to obtain a dynamically changing species density value, that is, the species density field; Based on the calculated species density field, use the spatial gradient calculation method to extract the spatial gradient of the species density field; Spatial graph structure construction: Define each spatial grid point as a node of the graph. The features of the node include the species density value, the magnitude and direction of the gradient, the terrain features, the accuracy and coverage radius of the observation equipment; If the gradient directions of the species density fields of two nodes are similar, establish an edge connection between the two nodes. For each adjacent node, the weight of the edge is determined based on the gradient similarity, geographical distance and equipment accuracy, and based on the obtained constructed spatial graph structure; Heterogeneous graph attention network: Map the features of each node through a trainable linear layer, and then use the multi-head heterogeneous graph attention mechanism to calculate the attention coefficients for each node and its adjacent nodes, and then perform weighted fusion of information based on the adjacent nodes of each node and the calculated attention coefficients to update the node features.

4. The method for predicting the impact of climate change on biodiversity based on deep learning according to claim 1, wherein Modeling climate prediction error based on a neural network differentiator includes: 3D convolution: Use 3D convolution based on the climate data tensor to extract local spatio-temporal features, and add an activation function after the convolution layer to increase non-linearity; Fully connected layer: Based on the features obtained after the 3D convolution operation, use a fully connected layer to process the features and output the error prediction result; Error covariance calculation: Based on the output error prediction result, calculate the covariance matrix of the error to describe the error correlation between different spatial points.

5. The method for predicting the impact of climate change on biodiversity based on deep learning according to claim 1, wherein The dual-branch feature extractor includes two branches; The first branch is used to extract the true biological information of the observation data; the second branch is used to extract the features related to noise; Calculate the covariance matrix of the biological observation noise based on the extracted noise-related features to represent the correlation of the noise.

6. The method for predicting the impact of climate change on biodiversity based on deep learning according to claim 1, wherein, The dynamic feedback prediction architecture is as follows: Construct a multi-layer spatio-temporal graph structure: Construct it based on the spatio-temporally aligned climate and biological data. The spatio-temporal graph structure includes a spatial graph and a time graph. Each spatial point in the spatial graph is regarded as a node in the graph, and the edge between the nodes represents the spatial adjacency relationship; Each time point in the time graph is regarded as a node in the graph, and the edge between the nodes represents the evolutionary relationship in time; Based on the constructed multi-layer spatio-temporal graph structure, obtain the spatio-temporal graph data; Spatiotemporal graph attention network processing: Using a 3D heat map based on spatiotemporal graph data and uncertainty as the input to the spatiotemporal attention network, different importance weights are assigned to each spatiotemporal node through the spatiotemporal attention mechanism, and the features of each spatiotemporal node are updated based on the calculated importance weights; Species distribution prediction: Based on the features of the updated runaway nodes, a species distribution prediction network is used to predict the distribution of species.

Images