Method for quantifying influence of climate change and human activity on vegetation and related device

Abstract

The invention discloses a method and a related device for quantifying the influence of climate change and human activity on vegetation, and the method comprises the steps: determining a subregion / time period in which the climate change and human activity jointly influence the growth of vegetation through a residual trend method, inputting the corresponding climate data and detrended climate data into a deep learning model, and carrying out the quantification of the influence of climate change and human activity on vegetation. According to the method, the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2 are obtained, so that the quantitative result of the influence of the climate change and the human activity on the vegetation is obtained, and compared with a traditional statistical regression model, the quantitative accuracy is improved.

Description

Technical Field

[0001] The invention relates to a method and a related device for quantifying the impact of climate change and human activities on vegetation, and belongs to the field of data statistical quantification. Background Art

[0002] With the intensification of climate change and the widespread development of human activities, the dynamic changes of vegetation are affected by many factors. In general, the factors affecting vegetation dynamics are divided into climate factors and human activity factors. For example, in arid areas, the increase in precipitation promotes vegetation growth, while the increase in temperature may inhibit regional vegetation growth; excessive reclamation may lead to vegetation degradation. Therefore, accurately quantifying the impact of climate change and human activities on vegetation growth can deepen the understanding of the impact of climate change and human activities on regional vegetation ecology, and is the premise of vegetation protection plans in the region. Existing quantitative methods mainly rely on traditional statistical regression models. Although they can quantify the impact of climate and human activities to a certain extent, the quantitative accuracy is poor. Summary of the invention

[0003] The present invention provides a method and a related device for quantifying the impact of climate change and human activities on vegetation, which solves the problems disclosed in the background technology.

[0004] According to one aspect of the present application, a method for quantifying the impact of climate change and human activities on vegetation is provided, comprising: Obtain remote sensing vegetation index and climate data for the area to be quantified; According to the remote sensing vegetation index and climate data of the area to be quantified, the residual trend method is used to determine the first sub-area / time period A1; wherein the first sub-area / time period A1 is a sub-area or time period in which climate change and human activities jointly affect vegetation growth in the area to be quantified; Input the climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre1 corresponding to the first sub-region / time period A1; input the detrended climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre2 corresponding to the first sub-region / time period A1; wherein the detrended climate data is the climate data after removing the climate change trend data; Calculate the quantitative results of the impact of climate change on vegetation based on the remote sensing vegetation index Pre1 and remote sensing vegetation index Pre2; According to the remote sensing vegetation index Pre1 and the remote sensing vegetation index corresponding to the first sub-region / time period A1, the quantitative result of the impact of human activities on vegetation is calculated.

[0005] Furthermore, the deep learning model is pre-trained, and the pre-training process includes: Obtain remote sensing vegetation index and climate data of a training area; wherein the climate of the training area is similar to that of the area to be quantified; According to the remote sensing vegetation index and climate data of the training area, the second sub-area / time period A2 is determined by using the residual trend method; wherein the second sub-area / time period A2 is the sub-area or time period in the training area where only climate change affects vegetation growth; The remote sensing vegetation index and climate data corresponding to the second sub-region / time period A2 are used to train the deep learning model.

[0006] Furthermore, based on remote sensing vegetation index and climate data, the residual trend method is used to determine the sub-region / time period, including: According to the climate data, a pre-built regression model is used to predict the remote sensing vegetation index; wherein the remote sensing vegetation index is used as the dependent variable and the climate data is used as the independent variable in the regression model; Calculate the residual between the acquired remote sensing vegetation index and the predicted remote sensing vegetation index; If the trend of the residual is not significant, the sub-region / time period corresponding to the residual is considered to be the sub-region or time period where only climate change affects vegetation growth; If the trend of the residual is significant, the sub-region / time period corresponding to the residual is considered to be a sub-region or time period where climate change and human activities jointly affect vegetation growth.

[0007] Furthermore, when training the deep learning model, in addition to obtaining the remote sensing vegetation index and climate data of the training area, the underlying surface factor and / or human activity data of the training area are also obtained, and the remote sensing vegetation index, climate data, and underlying surface factor and / or human activity data corresponding to the second sub-area / time period A2 are used for training; During quantification, in addition to obtaining the remote sensing vegetation index and climate data of the area to be quantified, the underlying surface factors and / or human activity data of the area to be quantified are also obtained, and the climate data corresponding to the first sub-area / time period A1, as well as the underlying surface factors and / or human activity data are input into the deep learning model.

[0008] Furthermore, based on the remote sensing vegetation index Pre1 and remote sensing vegetation index Pre2, the quantitative results of the impact of climate change on vegetation are calculated, including: The difference between the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2 was calculated and used as the quantitative result of the impact of climate change on vegetation.

[0009] Furthermore, according to the remote sensing vegetation index Pre1 and the remote sensing vegetation index corresponding to the first sub-region / time period A1, the quantitative results of the impact of human activities on vegetation are calculated, including: Calculate the difference between the remote sensing vegetation index corresponding to the first sub-region / time period A1 and the remote sensing vegetation index Pre1, and use the difference as the quantification result of the impact of human activities on vegetation.

[0010] Furthermore, the deep learning model is a ConvLSTM model. A convolution operation is introduced into the gating mechanism of the ConvLSTM model, and the weights and biases of the gating are calculated through convolutional kernels.

[0011] According to another aspect of the present application, there is provided a device for quantifying the impact of climate change and human activities on vegetation, including: An acquisition module for acquiring the remote sensing vegetation index and climate data of the area to be quantified; A residual trend module for determining the first sub-region / time period A1 according to the climate data of the area to be quantified by using the residual trend method; wherein, the first sub-region / time period A1 is the sub-region or time period where climate change and human activities jointly affect vegetation growth, and the first sub-region is a sub-region of the area to be quantified; A prediction module for inputting the climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre1 corresponding to the first sub-region / time period A1, and inputting the detrended climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre2 corresponding to the first sub-region / time period A1; wherein, the detrended climate data is the climate data after removing the climate change trend data; A first quantification module for calculating the quantification result of the impact of climate change on vegetation according to the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2; A second quantification module for calculating the quantification result of the impact of human activities on vegetation according to the remote sensing vegetation index Pre1 and the remote sensing vegetation index corresponding to the first sub-region / time period A1.

[0012] According to another aspect of the present application, there is provided a computer-readable storage medium storing one or more programs, and the one or more programs include instructions that, when executed by a computing device, cause the computing device to execute the method for quantifying the impact of climate change and human activities on vegetation.

[0013] According to another aspect of the present application, there is provided a computer device including one or more processors and one or more memories. One or more programs are stored in the one or more memories and are configured to be executed by the one or more processors. The one or more programs include instructions for executing the method for quantifying the impact of climate change and human activities on vegetation.

[0014] The beneficial effects achieved by the present invention are as follows: the present invention adopts the residual trend method to determine the sub-regions / time periods where climate change and human activities jointly affect vegetation growth, inputs the corresponding climate data and detrended climate data into the deep learning model, obtains the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2, and thereby obtains the quantitative results of the impact of climate change and human activities on vegetation, which improves the accuracy of quantification compared to traditional statistical regression models. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A flowchart of the method for quantifying the impacts of climate change and human activities on vegetation; Figure 2 This is a schematic diagram of the structure of the ConvLSTM model; Figure 3 Block diagram of a device for quantifying the impact of climate change and human activities on vegetation. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. It is obvious that the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present application and its application or use. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0017] Unless specifically stated otherwise, the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present application.

[0018] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0019] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0020] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0021] It should be noted that similar symbols and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it does not require further discussion in subsequent figures.

[0022] Deep learning is a machine learning method that imitates the structure and function of the human brain neural network. It automatically extracts complex features from large-scale data by building a multi-layer neural network model and realizes end-to-end learning. Its core lies in hierarchical feature extraction. Through multi-layer nonlinear transformation, high-level semantic features are gradually abstracted from the original data. Common deep learning models include LSTM (Long Short-Term Memory), CNN (Convolutional Neural Networks), GAN (Generative Adversarial Network), ConvLSTM (Convolutional Long Short-Term Memory, a deep learning model that combines convolutional neural networks and long short-term memory networks), etc. Deep learning models are widely used in image recognition, natural language processing, weather forecasting, etc.

[0023] The residual trend method is a statistical method that analyzes the relationship between variables by separating the trend component and the residual component in the data. Its core steps include building a basic model, calculating residuals and trend analysis. This method is mainly used in ecological and environmental sciences (such as quantifying the contribution of climate change and human activities to vegetation cover), economic forecasting (such as analyzing time series data of economic indicators and detecting potential influencing factors not captured by the model), medicine and engineering (testing whether the assumptions of the time series model are valid and identifying the source of abnormal fluctuations), etc.

[0024] The embodiment of the present application provides a method for quantifying the impact of climate change and human activities on vegetation based on deep learning technology and residual trend method. The quantification method can be performed by a quantification device, which can be a terminal device or a server. Among them, the terminal device can include but is not limited to mobile phones, computers, smart wearable devices, etc., which are not limited by the embodiment of the present application; the server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, big data and artificial intelligence platforms, etc., which are not limited by the embodiment of the present application. Optionally, the quantification method can also be collaboratively executed by multiple electronic devices with computing power. For the sake of ease of explanation, the subsequent embodiments are described as being executed by a quantification device.

[0025] See also Figure 1 , Figure 1 This is a flow chart of a method for quantifying the impact of climate change and human activities on vegetation provided by an embodiment of the present application. The quantification method can be executed by a quantification device, and the quantification method can at least include the following steps: Step 1: Obtain the remote sensing vegetation index (Normalized Difference Vegetation Index, NDVI) and climate data of the area to be quantified.

[0026] It should be noted that remote sensing vegetation index: such as GIMMS NDVI3g (Global Inventory Modeling and Mapping Studies Normalized Difference Vegetation Index version 1 of the third generation), MODIS (Moderate Resolution Imaging Spectroradiometer) and other vegetation index data with different time series and resolutions; climate data: including temperature (maximum, minimum, average), precipitation, radiation, wind speed, air pressure, etc., or expanded to other factors (such as soil moisture, evapotranspiration, etc.) according to actual needs. These acquired data are time series data of spatial data, and the spatial resolution and time range of each factor should be consistent in the end.

[0027] It should be noted that in order to improve the recognition ability of subsequent models, in some embodiments, underlying surface factors and / or human activity data are further obtained; among them, underlying surface factors: such as land use, topography (elevation, slope, etc.) and other static or semi-static information that may affect vegetation growth; human activity data: such as land use change intensity, socioeconomic indicators, etc.

[0028] To facilitate subsequent processing, the acquired data will be further preprocessed, such as interpolation, cropping, alignment, and unification of temporal and spatial resolutions.

[0029] Step 2: Determine the first sub-region / time period A1 using the residual trend method based on the remote sensing vegetation index and climate data of the area to be quantified; wherein the first sub-region / time period A1 is the sub-region or time period in the area to be quantified where climate change and human activities jointly affect vegetation growth.

[0030] In some embodiments, the specific process of step 2 may be as follows: 21) Based on the climate data, a pre-built regression model is used to predict the remote sensing vegetation index; in the regression model, the remote sensing vegetation index is used as the dependent variable and the climate data is used as the independent variable.

[0031] It should be noted that multiple linear or other suitable regression forms (linear regression, nonlinear regression, etc.) can be used to construct a regression model with remote sensing vegetation index as the dependent variable and climate data as the independent variable, so as to obtain the theoretical trend of the impact of climate change on NDVI, which can be expressed as follows: NDVI 预测 =f(temperature, precipitation, radiation, ...); Among them, NDVI 预测 represents the predicted remote sensing vegetation index, and f is the regression function.

[0032] 22) Calculate the residual between the acquired remote sensing vegetation index and the predicted remote sensing vegetation index.

[0033] The residual can be expressed as ΔNDVI 预测 =NDVI 获取 -NDVI 预测 ; Among them, NDVI 获取 The remote sensing vegetation index obtained is the actual remote sensing vegetation index.

[0034] 23) Perform a significance test on the trend of the residuals. If the trend of the residuals is not significant, it is considered that the sub-region / time period corresponding to the residuals is a sub-region or time period where only climate change affects vegetation growth. If the trend of the residuals is significant, it is considered that the sub-region / time period corresponding to the residuals is a sub-region or time period where climate change and human activities jointly affect vegetation growth.

[0035] It should be noted that the significance test can be performed using the Mann-Kendall test or t-test to screen out regions / time periods where human activities have a significant impact at a significance level of α=0.05. If the residual trend is not significant (i.e. not less than α) or close to zero, it is believed that climate change can explain most of the vegetation changes, that is, it is believed that only climate change affects vegetation growth; if it is significant (i.e. below α), it means that human activities have a greater impact in this region / time period, that is, climate change and human activities jointly affect vegetation growth. Specifically, if the residual shows a significant positive trend in time or space, it usually indicates that human activities in this region / time period are promoting vegetation growth. If it shows a significant negative trend, it indicates that human activities may have an inhibitory or destructive effect on vegetation growth.

[0036] Step 3: Input the climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre1 corresponding to the first sub-region / time period A1; input the detrended climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre2 corresponding to the first sub-region / time period A1; wherein the detrended climate data is the climate data after removing the climate change trend data.

[0037] It should be noted that the deep learning model requires pre-selection and training. First, the remote sensing vegetation index and climate data of the training area are obtained. The climate of the training area is similar to that of the area to be quantified. According to the remote sensing vegetation index and climate data of the training area, the residual trend method is used to determine the second sub-area / time period A2. The second sub-area / time period A2 is the sub-area or time period in the training area where only climate change affects vegetation growth. Then, the remote sensing vegetation index and climate data corresponding to the second sub-area / time period A2 are used to train the deep learning model.

[0038] It should be noted that the deep learning model here can use LSTM or CNN, but in some embodiments, the ConvLSTM model is used. The ConvLSTM model is a deep learning model that combines the spatial feature extraction capability of a convolutional neural network (CNN) with the time series learning capability of LSTM. Here, adaptive changes are made based on the ConvLSTM model, as follows: See also Figure 2 In order to make full use of the spatial feature extraction capability of the convolutional network, the regional grid data (i.e., the climate data and remote sensing vegetation index of the input data) are input into the ConvLSTM network (specifically, a two-layer network) in the form of "multi-channel images" and "time series image sequences", that is, each time step (e.g. Figure 2 A month or a quarter in the map corresponds to a "two-dimensional map", in which the pixel values ​​can correspond to climate data such as NDVI, temperature, precipitation, radiation, etc. (multi-channel), which are then stacked over time to form a four-dimensional data structure (height, width, channel, time).

[0039] In the ConvLSTM network, convolution operations are introduced into the gating mechanisms of the traditional LSTM, such as the input gate, forget gate, and output gate, so that the gated weights and biases are calculated through the convolution kernel instead of the fully connected layer, thereby retaining spatial information.

[0040] For the multi-dimensional time, space, and feature dimension variables output by the ConvLSTM network, a pooling layer is first added to reduce the spatial dimension to one dimension, and then the feature dimension is reduced to one dimension through a fully connected layer, and finally spliced ​​to obtain the predicted vegetation index for the target time period.

[0041] The output layer of the ConvLSTM model can be set to regression mode as needed to predict the NDVI value. For each time step, ConvLSTM outputs the predicted spatial distribution of NDVI. By performing convolutional gating operations on input sequences of multiple time periods, the network can capture the complex spatiotemporal correlation between vegetation and climate data and potential time lag effects.

[0042] Comparing the prediction results of the above ConvLSTM model with the random forest output, it is superior in capturing spatiotemporal dependencies and nonlinear relationships.

[0043] In order to improve the model's ability to identify the impact of human activities, during training, underlying surface factors and / or human activity data can be added to the input channel of the ConvLSTM model. These data are entered into the model together with climate data for model training.

[0044] Correspondingly, during quantification, in addition to obtaining the remote sensing vegetation index and climate data, the underlying surface factors and / or human activity data are also obtained, and the climate data corresponding to the first sub-area / time period A1, as well as the underlying surface factors and / or human activity data are input into the ConvLSTM model to obtain the remote sensing vegetation index.

[0045] It should be noted that the climate change trend data is the straight line that best represents the time series trend obtained by fitting the linear regression model. The detrended climate data can be obtained by subtracting the climate change trend data from the actual climate data. For example, the detrended climate data at time k can be obtained by subtracting the climate change trend data at time k from the climate data at time k.

[0046] Step 4: Calculate the quantitative results of the impact of climate change on vegetation based on the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2.

[0047] Step 4 specifically calculates the difference between the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2, and uses the difference as the quantitative result of the impact of climate change on vegetation. The formula can be expressed as C = Pre1- Pre2, and the corresponding relative contribution value can be expressed as R C =C / Pre1.

[0048] Step 5: Calculate the quantitative result of the impact of human activities on vegetation based on the remote sensing vegetation index Pre1 and the remote sensing vegetation index corresponding to the first sub-region / time period A1.

[0049] Step 5 specifically calculates the difference between the remote sensing vegetation index corresponding to the first sub-region / time period A1 and the remote sensing vegetation index Pre1, and uses the difference as the quantitative result of the impact of human activities on vegetation. The formula can be expressed as H = Obs-Pre1, Obs is the remote sensing vegetation index corresponding to the first sub-region / time period A1, and the corresponding relative contribution value can be expressed as R H =H / Pre1.

[0050] The results obtained by the above method can be visualized in combination with the spatiotemporal distribution map of land use change to help decision makers or researchers understand the regional situation.

[0051] In the above method, the ConvLSTM model can comprehensively consider the spatiotemporal characteristics, learn the spatiotemporal characteristics through the convolution process of spatial range data in different time periods, and at the same time use the time-lag correlation of vegetation to different climate data to determine the climate data with the optimal memory time limit as the input of the deep learning model, thereby improving the accuracy of quantification.

[0052] The above method uses the residual trend method to determine the sub-regions / time periods where climate change and human activities jointly affect vegetation growth, and inputs the corresponding climate data and detrended climate data into the deep learning model to obtain the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2, thereby obtaining the quantitative results of the impact of climate change and human activities on vegetation. Compared with the traditional statistical regression model, the accuracy of quantification is improved.

[0053] See also Figure 3 , Figure 3 A block diagram of a device for quantifying the impact of climate change and human activities on vegetation provided in an embodiment of the present application, Figure 3 The embodiment is a virtual device that can be loaded and executed by a computer device, and the computer device may include the above-mentioned quantization device, Figure 3 The device may include an acquisition module, a residual trend module, a prediction module, a first quantization module and a second quantization module, which, when used to execute the above quantization method, may: The acquisition module obtains the remote sensing vegetation index and climate data of the area to be quantified.

[0054] The residual trend module uses the residual trend method to determine the first sub-region / time period A1 based on the remote sensing vegetation index and climate data of the area to be quantified; wherein the first sub-region / time period A1 is the sub-region or time period in the area to be quantified where climate change and human activities jointly affect vegetation growth.

[0055] The prediction module inputs the climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre1 corresponding to the first sub-region / time period A1, and inputs the detrended climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre2 corresponding to the first sub-region / time period A1; wherein the detrended climate data is the climate data after removing the climate change trend data.

[0056] The first quantification module calculates the quantitative results of the impact of climate change on vegetation based on the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2.

[0057] The second quantification module calculates the quantitative result of the impact of human activities on vegetation according to the remote sensing vegetation index Pre1 and the remote sensing vegetation index corresponding to the first sub-region / time period A1.

[0058] The above device uses the residual trend method to determine the sub-regions / time periods where climate change and human activities jointly affect vegetation growth, inputs the corresponding climate data and detrended climate data into a deep learning model to obtain the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2, thereby obtaining the quantitative results of the impact of climate change and human activities on vegetation. Compared with the traditional statistical regression model, the accuracy of quantification is improved.

[0059] This application also relates to a computer-readable storage medium storing one or more programs, the one or more programs including instructions that, when executed by a computing device, cause the computing device to execute the method for quantifying the impact of climate change and human activities on vegetation.

[0060] This application also relates to a computer device including one or more processors and one or more memories, the one or more programs being stored in the one or more memories and configured to be executed by the one or more processors, the one or more programs including instructions for executing the method for quantifying the impact of climate change and human activities on vegetation.

[0061] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0062] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0063] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device that implements the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 A function specified in one or more boxes.

[0064] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0065] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are included in the scope of the claims of the present invention to be approved.

Claims

1. A method for quantifying the impact of climate change and human activities on vegetation, characterized by: include: Obtain remote sensing vegetation index and climate data for the area to be quantified; According to the remote sensing vegetation index and climate data of the area to be quantified, the residual trend method is used to determine the first sub-area / time period A1; wherein the first sub-area / time period A1 is a sub-area or time period in which climate change and human activities jointly affect vegetation growth in the area to be quantified; Input the climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre1 corresponding to the first sub-region / time period A1; input the detrended climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre2 corresponding to the first sub-region / time period A1; wherein the detrended climate data is the climate data after removing the climate change trend data; Calculate the quantitative results of the impact of climate change on vegetation based on the remote sensing vegetation index Pre1 and remote sensing vegetation index Pre2; According to the remote sensing vegetation index Pre1 and the remote sensing vegetation index corresponding to the first sub-region / time period A1, the quantitative result of the impact of human activities on vegetation is calculated.

2. The method according to claim 1, characterized in that: Deep learning model pre-training, the pre-training process includes: Obtain remote sensing vegetation index and climate data of a training area; wherein the climate of the training area is similar to that of the area to be quantified; According to the remote sensing vegetation index and climate data of the training area, the second sub-area / time period A2 is determined by using the residual trend method; wherein the second sub-area / time period A2 is the sub-area or time period in the training area where only climate change affects vegetation growth; The remote sensing vegetation index and climate data corresponding to the second sub-region / time period A2 are used to train the deep learning model.

3. The method according to claim 1 or 2, characterized in that: Based on remote sensing vegetation indices and climate data, the process of determining sub-regions / time periods using the residual trend method includes: According to the climate data, a pre-built regression model is used to predict the remote sensing vegetation index; wherein the remote sensing vegetation index is used as the dependent variable and the climate data is used as the independent variable in the regression model; Calculate the residual between the acquired remote sensing vegetation index and the predicted remote sensing vegetation index; If the trend of the residual is not significant, the sub-region / time period corresponding to the residual is considered to be the sub-region or time period where only climate change affects vegetation growth; If the trend of the residual is significant, the sub-region / time period corresponding to the residual is considered to be a sub-region or time period where climate change and human activities jointly affect vegetation growth.

4. The method according to claim 2, characterized in that: When training the deep learning model, in addition to obtaining the remote sensing vegetation index and climate data of the training area, the underlying surface factor and / or human activity data of the training area are also obtained, and the remote sensing vegetation index, climate data, and underlying surface factor and / or human activity data corresponding to the second sub-area / time period A2 are used for training; During quantification, in addition to obtaining the remote sensing vegetation index and climate data of the area to be quantified, the underlying surface factors and / or human activity data of the area to be quantified are also obtained, and the climate data corresponding to the first sub-area / time period A1, as well as the underlying surface factors and / or human activity data are input into the deep learning model.

5. The method according to claim 1, characterized in that Based on the remote sensing vegetation index Pre1 and remote sensing vegetation index Pre2, the quantitative results of the impact of climate change on vegetation are calculated, including: The difference between the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2 was calculated and used as the quantitative result of the impact of climate change on vegetation.

6. The method according to claim 1, characterized in that According to the remote sensing vegetation index Pre1 and the remote sensing vegetation index corresponding to the first sub-region / time period A1, the quantitative results of the impact of human activities on vegetation are calculated, including: The difference between the remote sensing vegetation index corresponding to the first sub-region / time period A1 and the remote sensing vegetation index Pre1 is calculated, and the difference is used as the quantitative result of the impact of human activities on vegetation.

7. The method according to claim 1, characterized in that The deep learning model is the ConvLSTM model. The convolution operation is introduced into the gating mechanism of the ConvLSTM model, and the gating weights and biases are calculated through the convolution kernel.

8. A device for quantifying the impact of climate change and human activities on vegetation, characterized in that: include: An acquisition module is used to obtain remote sensing vegetation index and climate data of the area to be quantified; The residual trend module determines the first sub-region / time period A1 by using the residual trend method according to the remote sensing vegetation index and climate data of the area to be quantified; wherein the first sub-region / time period A1 is the sub-region or time period in which climate change and human activities jointly affect vegetation growth in the area to be quantified; The prediction module inputs the climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre1 corresponding to the first sub-region / time period A1, and inputs the detrended climate data corresponding to the first sub-region / time period A1 into the deep learning model to obtain the remote sensing vegetation index Pre2 corresponding to the first sub-region / time period A1; wherein the detrended climate data is the climate data after removing the climate change trend data; The first quantification module calculates the quantitative results of the impact of climate change on vegetation based on the remote sensing vegetation index Pre1 and the remote sensing vegetation index Pre2; The second quantification module calculates the quantitative result of the impact of human activities on vegetation according to the remote sensing vegetation index Pre1 and the remote sensing vegetation index corresponding to the first sub-region / time period A1.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores one or more programs, and the one or more programs include instructions. When the instructions are executed by a computing device, the computing device executes any one of the methods of claims 1 to 7.

10. A computer device, characterized in that include: One or more processors and one or more memories, one or more programs are stored in the one or more memories and are configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the methods described in claims 1 to 7.

Images