A method for establishing afforestation area for exotic tree species

By identifying the negative ecological environment impact points and predicting potential risk areas in the introduction areas of alien tree species, and determining the location of afforestation land based on the data of the origin, the problems of low afforestation efficiency and ecological risks in traditional methods are solved, and the success rate and ecological system of afforestation are achieved Improved sustainability.

CN118761631BActive Publication Date: 2024-11-22INST OF FORESTRY CHINESE ACAD OF FORESTRY +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411218508.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-02
Publication Date
2024-11-22
Estimated Expiration
2044-09-02

AI Technical Summary

Technical Problem

The traditional afforestation method for establishing afforestation area ignores the multifunctionality and negative ecological environment impact of the ecosystem, resulting in low afforestation efficiency, high economic costs and increased ecological risks, making it difficult to achieve sustainable development.

Method used

By extracting the distribution data of the origin and introduction areas of alien tree species, identifying negative ecological environment impact points, using environmental threshold species distribution model to predict potential risk areas, deleting risk area data, obtaining normal growth point data, and combining the origin area data Predict the location of potential afforestation land and establish the ultimate suitable afforestation area.

Benefits of technology

Significantly reduce the probability of afforestation failure and economic costs, improve the success rate of afforestation, ensure the stability of the ecosystem and the performance of multi-ecological service functions, and adapt to future climate change.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118761631B_ABST
    Figure CN118761631B_ABST
Patent Text Reader

Abstract

The present invention provides a method for establishing an afforestation area for alien tree species, and relates to the field of scientific afforestation technology for afforestation, including: extracting first distribution data and second distribution data respectively; based on the second distribution data, identifying distribution point data and environmental data corresponding to the distribution point data of the alien tree species in the introduction afforestation area that have negative ecological environmental impacts; inputting the distribution point data and the environmental data corresponding to the distribution point data into an environmental threshold species distribution model to obtain potential risk area data for negative ecological impacts of afforestation, and deleting the data falling into the potential risk area data in the second distribution data to obtain normal growth point data of the introduction afforestation area after deletion; and predicting the potential afforestation location of the alien tree species target afforestation area using the normal growth point data of the introduction afforestation area and the first distribution data of the alien tree species origin area. The present invention can improve the success rate and ecological benefits of afforestation and ensure the sustainable development of the forest ecosystem.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of scientific afforestation and greening, and in particular to a method for establishing an afforestation area for exotic tree species. Background Art

[0002] With the needs of ecological environment construction and forestry development, the introduction of alien tree species for afforestation has become a common practical demand. This is because alien tree species usually have advantageous characteristics in a certain aspect. For example, some alien tree species have a faster growth rate and can provide timber resources or play an ecological protection function in a shorter period of time. The productivity of the most widely cultivated alien tree species plantations in Europe is on average 20-30% higher than that of European species. Some alien tree species may be more adaptable to specific difficult site conditions, such as drought, saline-alkali land, etc., and can be successfully afforested in these places. Some alien tree species may have higher economic value, such as good wood quality and can be used in specific industries. Some alien tree species can make up for the deficiencies of local tree species in certain ecological functions and better cope with ecological problems.

[0003] The afforestation area of ​​exotic tree species is constantly expanding. Due to improper selection of afforestation sites for exotic tree species, a series of negative ecological and environmental problems caused by them have gradually emerged, including low afforestation survival rate, slow forest growth, premature aging or death of forest stands, soil moisture depletion, formation of dry soil layer, decreased soil fertility, and reduced biodiversity. This has greatly reduced the effectiveness of afforestation, increased the economic cost and failure probability of afforestation, and is not conducive to the sustainable development of artificial forest ecosystems. Therefore, the precise establishment of afforestation areas for exotic tree species is a problem and hot topic that is widely concerned at home and abroad.

[0004] However, the traditional method of establishing afforestation areas for exotic tree species only compares the ecological environment of the original area of ​​the exotic tree species with the ecological environment of the target area, and selects the areas that initially meet the conditions as afforestation areas for afforestation. This traditional method only focuses on the dominant characteristics of a certain aspect of exotic tree species after afforestation, and believes that as long as this dominant characteristic is evaluated well, the afforestation is successful; however, it ignores the system attributes of the ecosystem itself with multiple functions and benefits, and does not consider the fact that the current exotic tree species have a negative impact on the ecological environment. Summary of the invention

[0005] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a method for establishing an afforestation area for exotic tree species, so as to improve the success rate and ecological benefits of afforestation and ensure the sustainable development of the forest ecosystem.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] A method for establishing an exotic tree species afforestation area, comprising:

[0008] Extracting first distribution data of the original production area of ​​the alien tree species and second distribution data of the alien tree species in the current afforestation area respectively;

[0009] Based on the second distribution data, identifying the distribution point data of the alien tree species that have a negative ecological environmental impact in the afforestation area and the environmental data corresponding to the distribution point data;

[0010] Input the distribution point data and the environmental data corresponding to the distribution point data into an environmental threshold species distribution model to obtain data of potential risk areas where afforestation may have a negative ecological impact, and delete the data in the second distribution data that fall into the potential risk area data to obtain normal growth point data of the afforestation area after deletion;

[0011] The normal growth point data of the introduced afforestation area and the first distribution data of the original production area of ​​the alien tree species are used to predict the potential afforestation site location in the target afforestation area of ​​the alien tree species.

[0012] Preferably, the first distribution data is obtained by obtaining the intersection of the distribution range of the original production area of ​​the alien tree species and the distribution point data of the original production area of ​​the alien tree species.

[0013] Preferably, the second distribution data is obtained by extracting published documents, vegetation maps and online resources.

[0014] Preferably, the method for acquiring the distribution point data of the negative ecological environment impact caused by the alien tree species in the afforestation area includes:

[0015] Identify the geographical locations of afforestation failure points and growth retardation points through field surveys;

[0016] The geographical coordinates of the points where afforestation caused negative ecological environment were extracted from published literature;

[0017] The univariate linear regression trend analysis method was used to analyze the changing trend of the long-term remote sensing vegetation index representing the growth status of alien tree species to identify the geographical location of the growth decline point;

[0018] The geographical location coordinates of afforestation failure points, slow growth points, points where afforestation produces negative ecological environment, and the geographical location coordinates of growth decline points are determined as the distribution point data that produces negative ecological environment impacts.

[0019] Preferably, the environmental data includes climate, soil and topographic data.

[0020] Preferably, the environmental threshold species distribution model includes any one of the maximum entropy model MAXENT, DOMAIN, BioClim or interface threshold model.

[0021] Preferably, predicting the potential afforestation location of the target afforestation area of ​​the alien tree species by using the normal growth point data of the afforestation area of ​​the introduced tree species and the first distribution data of the origin area of ​​the alien tree species includes:

[0022] constructing a first species distribution model using the first distribution data and corresponding environmental data;

[0023] Using the normal growth point data of the introduced afforestation area and its corresponding environmental data, a second species distribution model is constructed;

[0024] Using the first species distribution model and the second species distribution model to predict the target afforestation area, respectively, to obtain a first prediction result and a second prediction result;

[0025] The combination of the first prediction result and the second prediction result is determined as the final suitable afforestation location of the target afforestation area under future climate change conditions.

[0026] Preferably, the construction method of the first species distribution model and the second species distribution model includes: any one of random forest, classification regression tree, artificial neural network, generalized boosting method, support vector machine, generalized linear model, generalized additive model, multivariate adaptive spline smoothing function or flexible discriminant analysis method.

[0027] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0028] The present invention provides a method for establishing an afforestation area for alien tree species, comprising: extracting first distribution data of the original production area of ​​the alien tree species and second distribution data of the alien tree species in the current introduced afforestation area respectively; based on the second distribution data, identifying the distribution point data of the alien tree species in the introduced afforestation area that have negative ecological environmental impacts and the environmental data corresponding to the distribution point data; inputting the distribution point data and the environmental data corresponding to the distribution point data into an environmental threshold species distribution model to obtain the potential risk area data of the negative ecological impacts of afforestation, and deleting the data falling into the potential risk area data in the second distribution data to obtain the normal growth point data of the introduced afforestation area after deletion; using the normal growth point data of the introduced afforestation area and the first distribution data of the original production area of ​​the alien tree species to predict the potential afforestation location of the target afforestation area of ​​the alien tree species. The present invention establishes the final suitable afforestation area on the basis of considering both the positive and negative aspects of afforestation success and afforestation negative ecological environmental impacts, which can greatly reduce the probability and economic cost of afforestation failure and the unexpected potential ecological risks caused, and is conducive to the stability and sustainability of the ecosystem after afforestation and the performance of multiple ecological services and multiple ecological functions. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0030] Figure 1 A flow chart of a method provided by an embodiment of the present invention;

[0031] Figure 2 A schematic diagram of steps provided for an embodiment of the present invention; DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] The purpose of the present invention is to provide a method for establishing an afforestation area for exotic tree species, which can improve the success rate and ecological benefits of afforestation and ensure the sustainable development of the forest ecosystem.

[0034] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0035] Figure 1 A flow chart of a method provided by an embodiment of the present invention, such as Figure 1 As shown, the present invention provides a method for establishing an exotic tree species afforestation area, comprising:

[0036] Step 100: extracting first distribution data of the origin area of ​​the alien tree species and second distribution data of the alien tree species in the current afforestation area respectively;

[0037] Step 200: Based on the second distribution data, identifying distribution point data of the alien tree species that have a negative ecological environmental impact in the afforestation area and environmental data corresponding to the distribution point data;

[0038] Step 300: input the distribution point data and the environmental data corresponding to the distribution point data into an environmental threshold species distribution model to obtain data of potential risk areas where afforestation may have a negative ecological impact, and delete the data in the second distribution data that fall into the potential risk area data to obtain normal growth point data of the afforestation area after deletion;

[0039] Step 400: predict the potential afforestation location of the target afforestation area of ​​the alien tree species by using the normal growth point data of the introduced afforestation area and the first distribution data of the origin area of ​​the alien tree species.

[0040] Preferably, the first distribution data is obtained by obtaining the intersection of the distribution range of the original production area of ​​the alien tree species and the distribution point data of the original production area of ​​the alien tree species.

[0041] Preferably, the second distribution data is obtained by extracting published documents, vegetation maps and online resources.

[0042] Specifically, Figure 2 As shown, the technical route of this embodiment is as follows:

[0043] Step S1: extracting distribution data (first distribution data) of the origin area (natural distribution area) of the alien tree species;

[0044] The data of the origin distribution of alien tree species is obtained by finding the intersection of the distribution range of alien tree species and the distribution point data of alien tree species. The sources of the distribution range of alien tree species include: (1) the description of the administrative boundaries of their origin distribution range in relevant flora books; (2) the existing expert maps of the distribution range of alien tree species.

[0045] The sources of data on the distribution of origin of alien tree species include: (1) literature databases such as journal articles, dissertations and monographs, such as Web of Science and CNKI; (2) vegetation maps or remote sensing interpretation data; and (3) online platforms, such as the Global Biodiversity Platform (GBIF) and the Chinese Digital Herbarium (CVH).

[0046] Step S2: extracting the distribution data (second distribution data) of the foreign tree species in the current afforestation area;

[0047] The sources of data on the distribution of introduced afforestation areas include: (1) literature databases such as journal articles, dissertations and monographs, such as Web of Science and CNKI; (2) vegetation maps or remote sensing interpretation results; and (3) online platforms, such as the Global Biodiversity Platform (GBIF) and the Chinese Digital Herbarium (CVH).

[0048] Preferably, the method for acquiring the distribution point data of the negative ecological environment impact caused by the alien tree species in the afforestation area includes:

[0049] Identify the geographical locations of afforestation failure points and growth retardation points through field surveys;

[0050] The geographical coordinates of the points where afforestation caused negative ecological environment were extracted from published literature;

[0051] The univariate linear regression trend analysis method was used to analyze the changing trend of the long-term remote sensing vegetation index representing the growth status of alien tree species to identify the geographical location of the growth decline point;

[0052] The geographical location coordinates of afforestation failure points, slow growth points, points where afforestation produces negative ecological environment, and the geographical location coordinates of growth decline points are determined as the distribution point data that produces negative ecological environment impacts.

[0053] Furthermore, step S3 of this embodiment is: identifying distribution points where foreign tree species have negative ecological environmental impacts in the afforestation area;

[0054] Methods for screening data on distribution points of negative environmental impacts of alien tree species in introduced afforestation areas include:

[0055] (1) Identify the geographical locations of afforestation failure points and growth retardation points through field surveys;

[0056] (2) Extracting the geographical coordinates of points where afforestation has a negative impact on the ecological environment from published literature;

[0057] (3) Identify the geographical locations of growth decline points by analyzing the changing trends of long-term remote sensing vegetation indices that characterize the growth status of alien tree species.

[0058] Specifically, in the literature extraction of the distribution point data of the negative ecological and environmental impacts of afforestation with alien species, the geographical location data of the deterioration of the ecological environment caused by afforestation with alien tree species compared with the local native vegetation are used. The indicators of ecological and environmental deterioration include reduced soil moisture content, decreased soil fertility indicators such as carbon and nitrogen, reduced biodiversity, low afforestation survival rate, and slow growth and development. The databases for extracting literature include: Web of Science, China National Knowledge Infrastructure (CNKI), etc.

[0059] In the analysis of identifying growth decline points based on vegetation index, the distribution position of alien tree species grid units is first established based on vegetation maps or remote sensing interpretation data, and then the univariate linear regression trend analysis method is used to analyze the evolution trend of alien tree vegetation index in the past 20 years, and statistical analysis software (such as R software) is used to complete the temporal trend analysis and significance test of vegetation index. Remote sensing data sources include: MODIS data and Landsat data. The long-term series vegetation plants that can be used include: Normalized Difference Vegetation Plant NDVI, Enhanced Vegetation Index EVI, Net Primary Productivity NPP, etc.; database sources include MODIS remote sensing data (such as EVI product MOD13Q1, NDVI product MOD13Q1, NPP product MOD17A3HGF, etc.) and Landsat remote sensing data.

[0060] Preferably, the environmental data includes climate, soil and topographic data.

[0061] Preferably, the environmental threshold species distribution model includes any one of the maximum entropy model MAXENT, DOMAIN, BioClim or interface threshold model.

[0062] Specifically, step S4 of this embodiment is specifically: screening the data of normal growth distribution points of foreign tree species in the introduction and afforestation area;

[0063] Screening of data on normal growth distribution points of alien tree species in introduced afforestation areas: First, the afforestation data of negative ecological environment points collected in step S3 and the corresponding environmental data are associated through an environmental threshold species distribution model that only relies on species existence data. The environmental data include climate, soil and terrain data. Then, the model is used to predict the entire introduced afforestation area to identify areas with afforestation risks. Finally, the data falling into the afforestation risk identification area from the afforestation area distribution data collected in step S2 is deducted, and the remaining data is used as the normal growth point data of the introduced afforestation area.

[0064] Global climate data sources include: (1) climate data from the WorldClim platform (https: / / www.worldclim.org / ); (2) climate data from the CHELSA platform (https: / / chelsa-climate.org / ), etc.

[0065] Sources of global soil data include: SoilGRID250m (https: / / soilgrids.org / ), (2) the World Soil Database (HWSD) (https: / / www.fao.org / soils-portal / data-hub / soil-maps-and-databases / harmonized-world-soil-database-v20 / en / ), etc.

[0066] The sources of global terrain data include elevation, slope and aspect. Terrain data can be generated from terrain maps or digital elevation models (DEMs) using the Geographic Information System (GIS). Digital elevation models include:

[0067] (1) European Space Agency Copernicus Digital Elevation Model Cop-DEM (https: / / spacedata.copernicus.eu / collections / copernicus-digital-elevation-model#!);

[0068] (2) Japan Aerospace Exploration Agency elevation data AW3D30 (https: / / www.eorc.jaxa.jp / ALOS / en / dataset / aw3d30 / aw3d30_e.htm);

[0069] (3) German Space Agency TanDEM-X digital elevation model ( https: / / geoservice.dlr.de / web / dataguide / tdm90 / );

[0070] (4) NASA DEM ( https: / / search.earthdata.nasa.gov / search );

[0071] (5) ASTER DEM (data jointly developed by NASA and the Ministry of Economy, Trade and Industry (METI) of Japan) https: / / asterweb.jpl.nasa.gov / gdem.asp )wait.

[0072] Environmental threshold species distribution models that rely only on species existence data include: maximum entropy model MAXENT, DOMAIN, BioClim, surface range envelope model, etc.

[0073] Preferably, predicting the potential afforestation location of the target afforestation area of ​​the alien tree species by using the normal growth point data of the afforestation area of ​​the introduced tree species and the first distribution data of the origin area of ​​the alien tree species includes:

[0074] constructing a first species distribution model using the first distribution data and corresponding environmental data;

[0075] Using the normal growth point data of the introduced afforestation area and its corresponding environmental data, a second species distribution model is constructed;

[0076] Using the first species distribution model and the second species distribution model to predict the target afforestation area, respectively, to obtain a first prediction result and a second prediction result;

[0077] The combination of the first prediction result and the second prediction result is determined as the final suitable afforestation location of the target afforestation area under future climate change conditions.

[0078] Specifically, the target afforestation area and the introduced afforestation area have different meanings. The former refers to the area where it is necessary to predict whether it is suitable for afforestation of this tree species, and the latter refers to the area where this tree species has already been used for afforestation.

[0079] Preferably, the construction method of the first species distribution model and the second species distribution model includes: any one of random forest, classification regression tree, artificial neural network, generalized boosting method, support vector machine, generalized linear model, generalized additive model, multivariate adaptive spline smoothing function or flexible discriminant analysis method.

[0080] Furthermore, step S5 of this embodiment is specifically as follows: the location of potential afforestation sites in the target afforestation area for exotic tree species is established.

[0081] The potential afforestation location of the target afforestation area for alien tree species is established: first, the data in step S1 and the corresponding environmental data are used to construct a species distribution model, and then the data in step S4 and its corresponding environmental data are used to construct a species distribution model. The environmental data include climate, soil and terrain data; finally, the two models are used to predict the target afforestation area respectively, and the two prediction results are combined as the final suitable afforestation location of the target afforestation area. The latest future climate change scenario data is used in the two models to predict the target afforestation area, and the two prediction results are combined as the final suitable afforestation location of the target afforestation area under future climate change conditions.

[0082] Species distribution models include:

[0083] (1) Machine learning techniques that use both species presence and absence data, such as random forest, classification tree analysis, artificial neural network, generalized boosted method, support vector machine, etc.

[0084] (2) Traditional statistical models that use both presence and absence data, such as generalized linear models, generalized additive models, multivariate adaptive regression splines, flexible discriminant analysis, etc.

[0085] (3) Environmental threshold species distribution models that rely only on species existence data, such as the maximum entropy model MAXENT, DOMAIN, BioClim, and the surface range envelope model.

[0086] Future climate change data include: (1) future climate change scenario data selected from the International Coupled Model Intercomparison Program initiated by the World Climate Research Program (WCRP). The latest data comes from the new round of International Coupled Model Intercomparison Program (CMIP6) initiated by WCRP; (2) customized future climate change scenarios.

[0087] Optionally, the model building method in this embodiment is as follows:

[0088] (1) Collect data. The collected data include: environmental data and distribution data. Environmental data refers to the collection of environmental factors related to tree species growth, such as temperature, precipitation, soil type, sunshine, altitude, etc. Distribution data refers to the distribution locations of tree species in the origin and introduction areas and the corresponding good growth points.

[0089] (2) Data preprocessing: The preprocessing steps include cleaning data and normalizing or standardizing environmental data. Cleaning data is used to remove outliers and missing values; normalizing or standardizing environmental data can facilitate model processing.

[0090] (3) Feature selection: This embodiment analyzes which environmental factors have the greatest impact on the distribution of tree species, and selects these factors as the input features of the model. Exemplarily, the climate factors after feature selection in this embodiment include average temperature, extreme temperature, precipitation and its seasonal variation, and annual vapor pressure (affecting humidity); geographical factors include: altitude, terrain slope and slope aspect; soil factors include: soil type and its fertility, soil moisture and pH value; biological factors include: plant community structure, density, and diffusion pattern of competing tree species; human activity factors include: changes in land use patterns, such as urbanization and agricultural expansion, historical and current deforestation patterns; ecological factors include: biodiversity index, potential invasive species. Through a scientific feature selection process, this embodiment can ensure that the input data of the model is optimized and can maximize the accuracy and reliability of the model's prediction of tree species distribution.

[0091] (4) Model selection and construction. In this embodiment, a suitable model algorithm is selected and, based on known distribution data and environmental factors, the model is trained to predict the survival probability of species under specific environmental conditions.

[0092] (5) Model training and verification: This embodiment uses a portion of the data for model training and uses the remaining data for model verification to evaluate the prediction accuracy of the model. This embodiment also adjusts the model parameters to improve the model performance.

[0093] (6) Prediction and analysis: Use the trained model to make predictions for the target area. Combined with the assessment of future climate change scenarios, apply the model to different climate prediction data to determine suitable and unsuitable areas.

[0094] Through the above steps, this embodiment constructs and applies a species distribution model to predict the location of suitable afforestation areas for alien tree species.

[0095] Furthermore, the specific method of feature selection in this embodiment includes: assuming that we have n candidate factors, and the characteristic value of each factor is x i , the characteristic value of the target variable (tree species distribution) is y The core goal of feature selection is to find the most influential factor set S.

[0096] The factor set formula in this embodiment is:

[0097]

[0098] in, is a feature set score. are weight parameters, set according to specific project requirements and meet . Relevance score It is used to measure the strength of the linear relationship between the feature and the target variable. In this example, the Pearson correlation coefficient is used:

[0099]

[0100] in, This refers to the variable x i With the target variable The correlation coefficient between the two variables is usually calculated using the Pearson correlation coefficient. This coefficient measures the strength and direction of the linear relationship between two variables, ranging from -1 to 1. 1 indicates a perfect positive correlation: x i When increasing, -1 indicates a perfect negative correlation: when x i When increasing, On the contrary, it decreases. 0 indicates no linear relationship. This embodiment can help identify which feature may have a significant impact on the target variable by calculating the correlation coefficient between each feature variable and the target variable.

[0101] Importance score It is to use machine learning models to evaluate feature importance, usually using random forests or decision trees. The following is the public statement:

[0102]

[0103]

[0104] in, It is an indicator used to evaluate the amount of information provided by a feature in a model such as a decision tree, usually calculated by methods such as information gain or Gini impurity. The average amount by which the uncertainty or impurity of a model segment can be reduced. The more important a feature is, the greater its contribution to improving model performance. The information gain method is to measure the extent to which a feature can reduce the uncertainty of the target classification in a classification task, usually taking the difference between its previous uncertainty and the uncertainty after using the feature. Gini impurity is used to measure the impurity of a node. The higher the node impurity, the more categories it contains. This embodiment helps select the most discriminative features for the target variable by calculating the implementation effect of the feature at each node.

[0105] Diversity score It is used to measure the redundancy and synergy between selected features and encourage the selection of diverse feature sets. This embodiment uses forward-backward feature selection to evaluate the similarity between features.

[0106]

[0107] in, Features x i With features The correlation coefficient between can be calculated using Pearson or other correlation measurement methods. This indicator is used to measure the characteristics x i and The linear relationship or degree of interdependence between them.

[0108] For example:

[0109] High correlation: When When it is close to 1 or -1, it means that the two features have a strong linear correlation and may be repeated in information.

[0110] Low correlation: When When it is close to 0, it means that the two features are independent in information.

[0111] By comparing the correlations between features, we can ensure that the selected feature set is diverse and avoid redundancy, thereby improving the learning ability of the model.

[0112] In the implementation process, the above formula of this embodiment needs to be cross-validated and adjusted to find the best , to meet the goals of the application scenario. Through these multi-dimensional indicators, it is ensured that the selected features are not only highly correlated with the target variable, but also diverse and non-repetitive, thereby enhancing the predictive ability of the model.

[0113] Furthermore, when constructing a species distribution model, this embodiment uses artificial intelligence neural networks (especially deep learning methods) to provide a powerful tool for processing complex nonlinear relationships. When combined with the latest future climate change scenario data for prediction, the model construction method of this embodiment includes:

[0114] 1. Data preparation: Input variables include a wide range of environmental factors (temperature, precipitation, soil properties, sunshine, altitude, wind speed, etc.). Output variables include binary classification of species presence (1) or absence (0).

[0115] 2. Neural network model selection: This embodiment selects a recurrent neural network (RNN). The model architecture is designed as follows:

[0116] ① Input layer: processes input features.

[0117] ②Hidden layer: Several hidden layers, using activation functions such as ReLU to capture nonlinear relationships.

[0118] ③Output layer: Apply Sigmoid activation function for binary classification output.

[0119] 3. Training and optimization: This embodiment uses the data set to train the network through back propagation and optimization algorithms (such as Adam or RMSprop), and uses the cross entropy loss function to optimize the classification task.

[0120] 4. Prediction and evaluation: Use trained models to predict species distribution under future climate scenarios.

[0121] As an optional implementation, taking Robinia pseudoacacia as an example, the experimental process and results of this embodiment are as follows:

[0122] Through step S1, the geographical distribution range and geographical distribution point data of Robinia pseudoacacia in the place of origin are obtained.

[0123] Through steps S2, S3 and S4, the data of Robinia pseudoacacia distribution points are extracted from literature data (web of Science and CNKI) and vegetation map data to obtain the distribution point data of Robinia pseudoacacia in the introduced afforestation area and the distribution data of Robinia pseudoacacia afforestation failure points; by using the data of points on negative ecological and environmental impacts caused by Robinia pseudoacacia afforestation, the species distribution model MAXENT is used to simulate the scope of the potential risk area for afforestation failure in the Robinia pseudoacacia introduced afforestation area.

[0124] Through step S5, the MAXENT model predicts the location of the afforestation land in the target introduction area of ​​Robinia pseudoacacia using the data within the natural distribution range of the Robinia pseudoacacia original area; the MAXENT model predicts the location of the afforestation land in the target introduction area of ​​Robinia pseudoacacia using the data of normal growth distribution points in the introduction afforestation area of ​​Robinia pseudoacacia; the above two prediction results are combined to obtain the final afforestation land location of Robinia pseudoacacia under the current climatic conditions of the target afforestation area.

[0125] In addition, this embodiment also provides a method for establishing an alien tree species afforestation area, which is the same as Embodiment 1 except that step S4, "using the DOMAIN model to predict the risk area where alien species may have negative ecological environmental impacts in the afforestation area", is different from Embodiment 1.

[0126] In addition, this embodiment also provides a method for establishing an afforestation area for alien tree species. Except for step S5, "using machine learning technology random forest to predict the location of alien species in the target afforestation area" is different from Example 1, the rest is the same.

[0127] The beneficial effects of the present invention are as follows:

[0128] (1) The present invention establishes the final suitable afforestation area based on the consideration of both the positive and negative aspects of afforestation success and the negative ecological and environmental impacts of afforestation, which can greatly reduce the probability of afforestation failure and economic costs as well as the unexpected potential ecological risks caused, and is beneficial to the stability and sustainability of the ecosystem after afforestation and the performance of multiple ecological services and multiple ecological functions.

[0129] (2) The present invention provides a method for adaptive afforestation planning of alien tree species in response to future climate change, and a method for predicting suitable areas for afforestation of alien tree species in the future under future climate change scenarios. This method has a high reference value and guidance for the forestry department to plan afforestation of alien species in response to future climate change, so that response measures can be taken in advance. For example, afforestation activities can be arranged immediately in areas that are suitable for afforestation under current and future climate conditions, and afforestation plans can be arranged in advance for areas that are not currently suitable but will be suitable in the future.

[0130] (3) The present invention can establish specific locations suitable for afforestation at the regional scale, and can also establish suitable locations for afforestation of alien tree species at the global scale. Compared with relying solely on traditional methods to establish suitable locations for afforestation at the local scale, the present invention has obvious advantages at the geographical scale.

[0131] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0132] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for establishing an afforestation area for exotic tree species, characterized in that: include: Extracting first distribution data of the original production area of ​​the alien tree species and second distribution data of the alien tree species in the current afforestation area respectively; Based on the second distribution data, identifying the distribution point data of the alien tree species that have a negative ecological environmental impact in the afforestation area and the environmental data corresponding to the distribution point data; Input the distribution point data and the environmental data corresponding to the distribution point data into an environmental threshold species distribution model to obtain data of potential risk areas where afforestation may have a negative ecological impact, and delete the data in the second distribution data that fall into the potential risk area data to obtain normal growth point data of the afforestation area after deletion; The potential afforestation sites in the target afforestation areas of alien tree species are predicted using the normal growth point data in the afforestation areas of introduced tree species and the first distribution data in the origin areas of alien tree species; The method for acquiring the distribution point data of the negative ecological environment impact caused by the introduced tree species in the afforestation area comprises: Identify the geographical locations of afforestation failure points and growth retardation points through field surveys; The geographical coordinates of the points where afforestation caused negative ecological environment were extracted from published literature; The univariate linear regression trend analysis method was used to analyze the changing trend of the long-term remote sensing vegetation index representing the growth status of alien tree species to identify the geographical location of the growth decline point; The geographical location coordinates of afforestation failure points, slow growth points, points where afforestation produces negative ecological environment, and the geographical location coordinates of growth decline points are determined as the distribution point data that produces negative ecological environment impacts.

2. The method for establishing an exotic tree species afforestation area according to claim 1, characterized in that: The first distribution data is obtained by finding the intersection of the distribution range of the original production area of ​​the alien tree species and the distribution point data of the original production area of ​​the alien tree species.

3. The method for establishing an exotic tree species afforestation area according to claim 1, characterized in that: The second distribution data is obtained by extracting published documents, vegetation maps and online resources.

4. The method for establishing an exotic tree species afforestation area according to claim 1, characterized in that: The environmental data includes climate, soil and terrain data.

5. The method for establishing an exotic tree species afforestation area according to claim 1, characterized in that: The environmental threshold species distribution model includes any one of the maximum entropy model MAXENT, DOMAIN, BioClim or interface threshold model.

6. The method for establishing an exotic tree species afforestation area according to claim 1, characterized in that: The potential afforestation locations of target afforestation areas for alien tree species are predicted using the normal growth point data of the afforestation areas for introduced species and the first distribution data of the origin areas of alien tree species, including: constructing a first species distribution model using the first distribution data and corresponding environmental data; Using the normal growth point data of the introduced afforestation area and its corresponding environmental data, a second species distribution model is constructed; Using the first species distribution model and the second species distribution model to predict the target afforestation area, respectively, to obtain a first prediction result and a second prediction result; The combination of the first prediction result and the second prediction result is determined as the final suitable afforestation location of the target afforestation area under future climate change conditions.

7. The method for establishing an exotic tree species afforestation area according to claim 6, characterized in that: The construction method of the first species distribution model and the second species distribution model includes: any one of random forest, classification regression tree, artificial neural network, generalized boosting method, support vector machine, generalized linear model, generalized additive model, multivariate adaptive spline smoothing function or flexible discriminant analysis method.