Method and system for monitoring and predicting degradation of alpine grassland

Abstract

The invention discloses an alpine grassland degradation monitoring and predicting method and system, and relates to the technical field of data processing. According to the alpine grassland degradation monitoring and predicting method and system, a grazing monitoring area and a mouse hole monitoring area of the alpine grassland are determined based on analysis of grazing monitoring data and mouse hole monitoring data of each monitoring area of the alpine grassland and historical area ground data of the alpine grassland; monitoring the grazing monitoring area and the mouse hole monitoring area of the alpine grassland to obtain ground data of the selected monitoring area; based on meteorological altitude factors of the alpine region, data acquisition conditions of the plurality of remote sensing devices are analyzed, first syn-position remote sensing representative data of the alpine grassland are selected, and grassland vegetation conditions of the monitoring region are evaluated in combination with ground data of the monitoring region, so that alpine grassland degradation prediction is realized. The grassland degradation condition is comprehensively analyzed and predicted, the accuracy of grassland degradation monitoring in the high-cold region is expected to be improved, and a scientific basis is provided for grassland resource management and protection.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for monitoring and predicting alpine grassland degradation. Background Art

[0002] Grassland degradation refers to phenomena such as decreased grassland productivity, changes in vegetation structure, and decreased soil quality, which lead to weakened grassland functions, reduced biodiversity, and even the evolution of grassland into desertification or bare land.

[0003] Currently, remote sensing technology has broad application prospects in grassland degradation monitoring and can provide large-scale, full-time data acquisition capabilities. However, when using remote sensing equipment to acquire data in high-altitude areas, meteorological altitude factors in high-altitude areas, such as snow cover, atmospheric interference, and complex terrain, may cause data errors, thereby causing inaccurate grassland degradation predictions.

[0004] On the other hand, grassland degradation in alpine areas is closely related to grazing and ecological balance. Ground data is expected to provide important information such as grassland utilization and grassland ecological status, while combined with remote sensing data, it can more comprehensively analyze and predict the trend of grassland degradation. Therefore, combining multiple data sources, comprehensively analyzing and predicting grassland degradation is expected to improve the accuracy and reliability of grassland degradation monitoring in alpine areas and provide a scientific basis for grassland resource management and protection.

[0005] Therefore, in response to the above problems, a method and system for monitoring and predicting alpine grassland degradation is urgently needed. Summary of the invention

[0006] In view of the shortcomings of the existing technology, the present invention provides a method and system for monitoring and predicting alpine grassland degradation, which solves the problem of inaccurate grassland monitoring caused by detection errors of remote sensing equipment in alpine areas and the lack of multiple data sources for predicting alpine grassland degradation.

[0007] To achieve the above objectives, the present invention is implemented by the following technical scheme: based on the similarity analysis of the regional ground data of the alpine grassland in each monitoring area of ​​the alpine grassland and the overall ground data of the historical area of ​​the alpine grassland, the monitoring areas of the alpine grassland are numbered in descending order according to the similarity between the regional ground data of the alpine grassland in each monitoring area and the overall ground data of the historical area of ​​the alpine grassland;

[0008] The optimal number of monitoring areas in each monitoring area is determined by Monte Carlo algorithm for multiple monitoring areas of alpine grasslands, and the monitoring areas are selected in order of size according to the similarity between each monitoring area and the overall ground data of the historical area of ​​alpine grasslands, and the ground data of the selected monitoring areas are obtained by monitoring;

[0009] Based on the meteorological altitude factors in the alpine areas, the data collection of multiple remote sensing equipment was analyzed, the first-priority remote sensing representative data of the alpine grassland was selected, and the grassland degradation status of the monitoring area was evaluated in combination with the ground data of the selected monitoring area, thereby realizing the prediction of alpine grassland degradation;

[0010] The monitoring area includes grazing monitoring area and rat hole monitoring area.

[0011] Furthermore, the specific analysis for determining the selected monitoring areas for grazing and the selected monitoring areas for rat holes in the alpine regions is as follows: the alpine grasslands are divided into regions and numbered according to the geographical location coordinates: AGH1, AGH2, AGH3, ..., AGHn; based on the database of alpine grasslands, the grazing monitoring data and rat hole monitoring data of each historical monitoring area are extracted, and the grazing monitoring data and rat hole monitoring data are preprocessed and feature extracted to obtain the grazing characteristic data set and the rat hole characteristic data set of each monitoring area respectively, and further obtain the mean of grazing characteristic data and the mean of rat hole characteristic data of each monitoring area; based on the database of alpine grasslands, the historical regional ground data of the alpine grasslands are extracted, and the historical ground data include the historical overall grazing data and the historical overall rat hole data of the alpine grasslands, and the historical overall grazing data and the historical overall rat hole data are preprocessed and feature extracted. The data were preprocessed and feature extracted to obtain the historical grazing overall characteristic data set and the historical rat hole overall characteristic data set of the alpine grassland, and further to obtain the historical grazing overall characteristic data mean and the historical rat hole overall characteristic data mean of the alpine grassland; based on the Euclidean distance, the mean grazing characteristic data and the rat hole characteristic data of each monitoring area were clustered with the historical grazing overall characteristic data mean and the historical rat hole overall characteristic data mean of the alpine grassland, and each monitoring area was renumbered according to the similarity value: AGH1', AGH2', AGH3',..., AGHn'; based on the Monte Carlo sampling algorithm, the optimal regional sampling number in the alpine area was determined, and then the selected monitoring areas of the alpine grassland were marked according to the number sorting, and the selected monitoring areas included the grazing selected monitoring areas and the rat hole selected monitoring areas.

[0012] Furthermore, the grazing characteristic data set of each selected monitoring area is specifically a grazing growth rate set of each selected monitoring area; the rat hole characteristic data set of each selected monitoring area is specifically a rat hole number reduction rate set of each selected monitoring area.

[0013] Furthermore, the sampled ground data of the selected monitoring area specifically include grazing monitoring data of the selected monitoring area, rat hole monitoring data of the selected monitoring area; the grazing monitoring data specifically refers to the number of grazing livestock.

[0014] Furthermore, the specific analysis of selecting the first-priority remote sensing representative data of alpine grassland is as follows: the meteorological altitude factors in the alpine area specifically include the highest altitude, highest wind speed, and lowest temperature in the alpine area; the highest altitude, highest wind speed, and lowest temperature in the alpine area are preprocessed to determine the minimum support threshold, the remote sensing equipment in the alpine area is numbered, and the data collection information of each remote sensing equipment is obtained based on the historical data collection of the remote sensing equipment stored in the database, the data collection information of each remote sensing equipment specifically includes the optimal adaptation altitude, optimal adaptation wind speed, and optimal adaptation temperature of each remote sensing equipment; confidence and support analysis is performed based on the optimal adaptation altitude, optimal adaptation wind speed, and optimal adaptation temperature of each remote sensing equipment and the highest altitude, highest wind speed, and lowest temperature in the alpine area, the remote sensing equipment with the best comprehensive ranking based on confidence and support is used as the first-priority remote sensing equipment for the alpine grassland, and the monitoring data of the first-priority remote sensing equipment for the alpine grassland is used as the first-priority remote sensing representative data.

[0015] Furthermore, the specific analysis process for obtaining the data collection information of each remote sensing device is as follows: based on the database, the historical data of each remote sensing device is collected, including the working environment parameters of the device and the collected data; according to the parameter targets of the altitude, temperature and wind speed of the remote sensing device, the historical data is screened, and the working environment parameters and the corresponding collected data under the parameter target requirements are retained; the working environment parameters and the corresponding collected data under the retained parameter target requirements are counted and analyzed, and the working environment parameters with the smallest data error collected by each remote sensing device are selected as the data collection information of each remote sensing device.

[0016] Furthermore, the specific analysis for realizing the prediction of alpine grassland degradation is as follows: obtaining the first-order remote sensing representative data and sampling the ground data of the selected monitoring area within each monitoring time period of the first-order remote sensing equipment and the selected monitoring area; using the time series model to perform predictive analysis on the first-order remote sensing representative data and the ground data of the selected monitoring area, predicting the future vegetation coverage of the selected monitoring area, and comprehensively averaging the predicted future vegetation coverage of each sampling area to obtain the degradation prediction coefficient of the alpine grassland, which is used to quantify the degradation prediction status of the alpine grassland.

[0017] Furthermore, the time series model adopts the ARIMAX model, and the specific steps of predicting and analyzing the first-order remote sensing representative data and the ground data of the sampled selected monitoring area through the time series model are analyzed as follows: obtaining the first-order remote sensing representative data and the ground data of the sampled selected monitoring area as prediction data, and cleaning and preprocessing the first-order remote sensing representative data and the ground data of the sampled selected monitoring area; establishing an ARIMAX model, and using the historical remote sensing data and historical ground data of alpine grassland to perform maximum likelihood estimation fitting on the ARIMAX model to estimate the model parameters; using the residual analysis method to diagnose the fitted ARIMAX model to check whether the residuals of the model meet the white noise assumption; using the fitted ARIMAX model to predict future alpine grassland monitoring data.

[0018] Furthermore, the first-order remote sensing representative data specifically include vegetation coverage, and the exogenous variables in the ARIMAX model include the number of grazing livestock and the number of rat holes. The ARIMAX model formula is:

[0019] y t =β0+β1y t-1 +α1Δy t-1 +γ1ε t +γ2ε t-1 +γ3θ t +γ4θ t-1 +ε t ; In the formula, y t represents the vegetation coverage at time point t, y t-1 represents a lag term of vegetation coverage, Δy t-1 represents a difference term of vegetation coverage, ε t represents the number of grazing livestock at time point t, ε t-1 represents a difference term for the number of grazing livestock, θ t represents the number of rat holes at time point t, θ t-1 represents a differential term of the number of rat holes, β0 and β1 represent the coefficients of the autoregressive part, representing the autocorrelation of vegetation coverage, α1 represents the coefficient of the differential part, representing the stability of vegetation coverage, γ1, γ2, γ3, and γ4 represent the coefficients of exogenous variables, representing the impact of grazing quantity and rat hole quantity on vegetation coverage, and ε t It represents the error term of the ARIMAX model, which indicates the random fluctuations that cannot be explained by the ARIMAX model.

[0020] A system for monitoring and predicting degradation of alpine grasslands, which applies the above-mentioned method for monitoring and predicting degradation of alpine grasslands, comprises: a selected monitoring area module, which is used to determine the selected monitoring areas for grazing and the selected monitoring areas for rat holes of the alpine grasslands based on the grazing monitoring data and rat hole monitoring data of each selected monitoring area of ​​the alpine grassland and the ground data analysis of the historical area of ​​the alpine grasslands; a ground data acquisition module, which is used to monitor the selected monitoring areas for grazing and the selected monitoring areas for rat holes of the alpine grasslands to obtain the ground data of the sampled selected monitoring areas; a degradation prediction module, which is used to analyze the data collection conditions of multiple remote sensing devices based on the meteorological altitude factors of the alpine regions, select the first-order remote sensing representative data of the alpine grasslands, and evaluate whether the grassland in the sampled selected monitoring areas is degraded and the degree of degradation in combination with the ground data of the sampled selected monitoring areas, thereby realizing the prediction of degradation of the alpine grasslands.

[0021] The present invention has the following beneficial effects:

[0022] (1) The alpine grassland degradation monitoring and prediction method and system can provide more accurate and reliable monitoring results by combining grazing monitoring data, rat hole monitoring data and historical ground data of alpine grassland, as well as analysis of meteorological and altitude factors. The comprehensive use of data can improve the monitoring accuracy of alpine grassland degradation, which helps to find problems early and take corresponding management measures; considering multiple factors, including grazing conditions, rat hole conditions, and environmental factors such as meteorology and altitude, and comprehensively considering various factors affecting alpine grassland degradation, it can more comprehensively evaluate the health status of grassland and potential degradation risks; by adopting a method combining remote sensing data and ground monitoring data, as well as analysis of meteorological and altitude factors, the method is scientific and can improve the credibility of monitoring and prediction, making it an important tool for scientific management of alpine grassland; by combining historical data and environmental factors, the degradation of alpine grassland is predicted. This predictability can help relevant departments and managers take measures early to prevent further degradation of alpine grassland and protect the stability of the ecological environment and ecosystem.

[0023] (2) The alpine grassland degradation monitoring and prediction method and system can predict future vegetation coverage, which can help ecological and environmental managers better understand and predict changes in the alpine grassland ecosystem, help to detect signs of vegetation degradation or ecological imbalance early, and take corresponding management measures to protect and maintain the health of the grassland ecosystem. By monitoring and predicting vegetation coverage, grazing activities can be better planned. If the forecast shows that the vegetation coverage is declining, managers can adjust grazing strategies in time to avoid grassland degradation caused by overgrazing, which will help achieve the sustainability of grazing activities and protect the integrity of the grassland ecosystem. Based on the prediction of vegetation coverage, resources can be better allocated, policies can be formulated, and projects can be planned to promote the effective implementation of grassland management and protection. The use of time series analysis and vegetation coverage prediction can provide important data and cases for scientific research and education, which will help to enhance the understanding of the dynamic changes of alpine grassland ecosystems and promote the development of related disciplines and the enrichment of educational resources.

[0024] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 The present invention is a flow chart of a method for monitoring and predicting alpine grassland degradation.

[0026] Figure 2 This is a structural diagram of an alpine grassland degradation monitoring and prediction system of the present invention. DETAILED DESCRIPTION

[0027] The embodiments of the present application provide a method and system for monitoring and predicting alpine grassland degradation, which solves the problem of inaccurate grassland monitoring caused by detection errors of remote sensing equipment in alpine areas and the lack of multiple data sources for predicting alpine grassland degradation.

[0028] The overall idea of ​​the problem in the embodiment of this application is as follows:

[0029] Grazing monitoring data, rat hole monitoring data and historical ground data were collected from each monitoring area of ​​alpine grassland. These data contain descriptions of different aspects of grassland ecosystems: grazing intensity, vegetation coverage and composition, and soil conditions. These data were analyzed to understand the grassland conditions and change trends in different regions. Based on the analysis, grazing monitoring areas and rat hole monitoring areas were determined. Then, ground monitoring was carried out in these monitoring areas to obtain real-time ground data, including the number of grazing livestock and the number of rat holes. Combined with the meteorological and altitude factors in the alpine region, the data collection of multiple remote sensing equipment was analyzed. Representative remote sensing data suitable for alpine grasslands were selected. These data can provide a wider coverage and more frequent monitoring cycles. Finally, the ground data of the monitoring area was combined with the remote sensing data to conduct a comprehensive assessment of the grassland monitoring status in the monitoring area. Through the analysis of grassland conditions and consideration of environmental factors, the degradation of alpine grasslands was predicted so that timely measures could be taken for intervention and protection.

[0030] See also Figure 1 , an embodiment of the present invention provides a technical solution: a method for monitoring and predicting alpine grassland degradation, comprising the following steps: performing similarity analysis based on regional ground data of alpine grassland in each monitoring area of ​​the alpine grassland and overall ground data of the historical area of ​​the alpine grassland, and numbering each monitoring area of ​​the alpine grassland in descending order according to the similarity between the regional ground data of the alpine grassland in each monitoring area and the overall ground data of the historical area of ​​the alpine grassland;

[0031] The optimal number of monitoring areas in the alpine grassland is determined by Monte Carlo algorithm, and the monitoring areas are selected as selected monitoring areas in order of size according to the similarity between each monitoring area and the overall ground data of the historical area of ​​the alpine grassland, and the ground data of the selected monitoring areas are obtained by monitoring;

[0032] Based on the meteorological altitude factors in the alpine areas, the data collection of multiple remote sensing equipment was analyzed, the first-priority remote sensing representative data of the alpine grassland was selected, and the grassland degradation status of the monitoring area was evaluated in combination with the ground data of the selected monitoring area, thereby realizing the prediction of alpine grassland degradation;

[0033] The monitoring area includes grazing monitoring area and rat hole monitoring area.

[0034] Specifically, the specific analysis of determining the selected monitoring areas for grazing sampling and rathole sampling in the alpine areas is as follows: the alpine grassland is divided into regions and numbered according to the geographical location coordinates: AGH1, AGH2, AGH3, ..., AGHn; the historical grazing monitoring data and rathole monitoring data of each monitoring area are extracted based on the database of alpine grassland, and the grazing monitoring data and rathole monitoring data are preprocessed and feature extracted to obtain the grazing characteristic data set and rathole characteristic data set of each monitoring area, and further obtain the mean grazing characteristic data and rathole characteristic data of each monitoring area; the historical regional ground data of the alpine grassland is extracted based on the database of the alpine grassland, and the historical ground data includes the historical grazing overall data and the historical rathole overall data of the alpine grassland, and the historical grazing overall data and the historical rathole overall data are preprocessed and feature extracted. The data were preprocessed and feature extracted to obtain the historical grazing overall characteristic data set and the historical rat hole overall characteristic data set of the alpine grassland, and further to obtain the historical grazing overall characteristic data mean and the historical rat hole overall characteristic data mean of the alpine grassland; based on the Euclidean distance, the mean grazing characteristic data and the rat hole characteristic data of each monitoring area were clustered with the historical grazing overall characteristic data mean and the historical rat hole overall characteristic data mean of the alpine grassland, and each monitoring area was renumbered according to the similarity value: AGH1', AGH2', AGH3',..., AGHn'; based on the Monte Carlo sampling algorithm, the optimal regional sampling number in the alpine area was determined, and then the selected monitoring areas of the alpine grassland were marked according to the number sorting, and the selected monitoring areas included the grazing selected monitoring areas and the rat hole selected monitoring areas.

[0035] In this implementation scheme, the grazing characteristic data set of each monitoring area is specifically the grazing growth rate set of each monitoring area; the rat hole characteristic data set of each monitoring area is specifically the rat hole number reduction rate set of each monitoring area, and the grazing growth rate is calculated as follows: Where Δε represents the grazing growth rate, ε t represents the number of grazing livestock at the current monitoring time point, ε t-1 Indicates the number of grazing livestock at the last monitoring time point; the calculation formula for the reduction rate of rat holes is: Where Δθ represents the rathole reduction rate, θ t represents the number of ratholes at the current monitoring time point, θ t-1 The number of rat holes at the last monitoring time point.

[0036] Based on the Euclidean distance, the specific analysis of clustering the mean grazing characteristic data, the mean rat hole characteristic data of each monitoring area, the mean historical grazing characteristic data of the alpine grassland, and the mean historical rat hole characteristic data is as follows: The clustering algorithm formula is:

[0037] Where A is the number of region A, B is the number of region B, C is the number of alpine grassland, and d(A,C) FM G represents the Euclidean distance between the mean of grazing characteristic data in region A and the mean of overall historical grazing characteristic data in alpine grassland. A represents the mean of grazing characteristic data in area A, G C represents the mean of the overall characteristic data of historical grazing in alpine grassland, d(B,C) ST represents the Euclidean distance between the mean rathole characteristic data of region B and the mean of the overall characteristic data of historical ratholes in alpine grasslands, H B represents the mean value of rathole characteristic data in area B, H C It represents the average value of the overall characteristic data of historical rat holes in alpine grasslands; and the specific algorithm formula for the similarity value is: sim(A,C) FM Represents the similarity value between the mean of grazing characteristic data in region A and the mean of overall historical grazing characteristic data of alpine grassland, sim(B,C) ST Represents the similarity value between the mean of rat hole characteristic data in area B and the mean of overall characteristic data of historical rat holes in alpine grassland, sim(A,C) FM , sim(B,C) ST The value range of is between 0 and 1. The larger the similarity value, the higher the similarity. The similarity between the regional ground data of the alpine grassland in all monitoring areas and the overall ground data of the historical alpine grassland area is compared with the similarity threshold. The monitoring areas of the alpine grassland are numbered in descending order according to the similarity between the regional ground data of the alpine grassland in each monitoring area and the overall ground data of the historical alpine grassland area.

[0038] The optimal number of regional sampling in alpine areas is determined based on the Monte Carlo sampling algorithm. The specific method is to simulate the situations under different sampling numbers through random sampling, evaluate the performance indicators of each situation, and finally determine the optimal number of sampling. Specifically, the Monte Carlo method is used to randomly generate different numbers of sampling areas and perform sampling. For each number of sampling areas, the sampling points are used to estimate the target indicators. The target indicators include the mean and variance of the resource distribution. According to the estimated target indicators, the performance corresponding to each number of sampling areas is evaluated to determine the optimal number of regional sampling.

[0039] Specifically, the ground data of the sampled selected monitoring area specifically includes grazing monitoring data of the grazing sampled selected monitoring area, and rat hole monitoring data of the rat hole sampled selected monitoring area; the grazing monitoring data specifically refers to the number of grazing livestock.

[0040] In this implementation scheme, the number of grazing livestock refers to the number of animals grazing in the selected monitoring area through remote sensing equipment including high-resolution satellite images or drone aerial photography technology, including cattle, sheep, horses and other livestock, and is obtained by regularly monitoring the selected monitoring area and counting the number and distribution of grazing livestock; too many rat holes usually indicate that the productivity of grassland vegetation is reduced. This is because excessive rodent activity may lead to the destruction of vegetation, which in turn affects the productivity of grassland. In alpine grasslands, remote sensing equipment including high-resolution satellite images or drone aerial photography technology is used to remotely monitor grassland. Through image analysis, rat holes on the grassland can be identified and their number and distribution can be calculated; by monitoring the number of grazing livestock, the intensity and frequency of grazing on the grassland can be understood, and the carrying capacity and degree of damage of the grassland can be evaluated; monitoring the number of rat holes can help evaluate the complexity and stability of the ecosystem, so as to better understand the functional state and ecological balance of the ecosystem; the monitoring data of the number of grazing livestock and the number of rat holes can be used as input variables of the grassland degradation prediction model, reflecting the health status of the grassland ecosystem and the changes in ecological processes, and helping to predict the trend of grassland degradation and possible influencing factors.

[0041] Specifically, the specific analysis of selecting the first-priority remote sensing representative data of alpine grassland is as follows: the meteorological altitude factors in the alpine area specifically include the highest altitude, the highest wind speed, and the lowest temperature in the alpine area; the highest altitude, the highest wind speed, and the lowest temperature in the alpine area are preprocessed, and the minimum support threshold is determined based on the distribution characteristics and statistical indicators, the remote sensing equipment in the alpine area is numbered, and the data collection information of each remote sensing equipment is obtained based on the historical data collection of the remote sensing equipment stored in the database, and the data collection information of each remote sensing equipment specifically includes the optimal adaptation altitude, the optimal adaptation wind speed, and the optimal adaptation temperature of each remote sensing equipment; confidence and support analysis is performed based on the optimal adaptation altitude, the optimal adaptation wind speed, the optimal adaptation temperature of each remote sensing equipment and the highest altitude, the highest wind speed, and the lowest temperature in the alpine area, the remote sensing equipment with the best ranking in terms of the mean of confidence and support is taken as the first-priority remote sensing equipment for the alpine grassland, and the monitoring data of the first-priority remote sensing equipment for the alpine grassland is taken as the first-priority remote sensing representative data.

[0042] The database of this solution is built based on historical data collection, and there are many types of remote sensing equipment for multiple monitoring of data. In this embodiment, the remote sensing equipment can be a drone equipped with a high-resolution optical camera or a multi-spectral camera, which can carry out fine monitoring of a small area, thereby accurately monitoring the number and location of livestock in a smaller area. The drone can fly at low altitudes to capture real-time images, which is suitable for small-scale, short-term monitoring. At the same time, the remote sensing equipment can also be a high-resolution satellite such as WorldView-3 or Pleiades, which can provide images with a resolution of 0.5 meters to 1 meter. Through image analysis technology (such as target recognition algorithms), the activity trajectory and distribution of livestock can be identified on a large range of grasslands, and the fine structure of rat holes on the surface can be captured, which will not be elaborated here.

[0043] In this implementation plan, the specific analysis process for obtaining the data collection information of each remote sensing device is as follows: based on the database, the historical data of each remote sensing device is collected, including the working environment parameters of the device and the collected data; according to the parameter targets of the altitude, temperature and wind speed of the remote sensing device, the historical data is screened, and the working environment parameters and the corresponding collected data under the parameter target requirements are retained; the working environment parameters and the corresponding collected data under the retained parameter target requirements are counted and analyzed, and the working environment parameters with the smallest data error collected by each remote sensing device are selected as the data collection information of each remote sensing device.

[0044] The selection of the first-priority remote sensing representative data of alpine grassland is based on the Apriori algorithm, which is as follows: altitude, temperature and wind speed can be used as sub-criteria to construct a judgment matrix, compare the relative importance of each factor, and then assign weights of altitude, temperature and wind speed according to the judgment matrix, perform weighted analysis on altitude, temperature and wind speed, and obtain environmental characteristic values. The specific formula is:

[0045] Where HJ represents the environmental characteristic value, μ1 represents the altitude after de-dimensionalization, μ2 represents the temperature after de-dimensionalization, μ3 represents the wind speed after de-dimensionalization, ω1 represents the weight coefficient of altitude, ω2 represents the weight coefficient of temperature, and ω3 represents the weight coefficient of wind speed. Based on the calculation logic of the environmental characteristic value, the weight coefficient is continuously adjusted to obtain the environmental characteristic value set of each remote sensing device and the environmental characteristic value set of the alpine region. Its support Where M represents the number of remote sensing equipment, N represents the number of high-cold regions, Num(MN) represents the number of identical environmental characteristic values ​​between remote sensing equipment M and high-cold region N, Num(As) represents the total number of samples of environmental characteristic value set, and confidence Where Support(M∩N) represents the probability that the environmental characteristic values ​​of remote sensing device M and alpine region N are the same, and Support(M) represents the probability of the environmental characteristic value of remote sensing device M appearing in the total number of all data.

[0046] Specifically, the specific analysis to achieve the prediction of alpine grassland degradation is as follows: obtain the first-order remote sensing representative data and the ground data of the sampled selected monitoring area in each monitoring time period of the first-order remote sensing equipment and the sampled selected monitoring area; use the time series model to predict and analyze the first-order remote sensing representative data and the ground data of the sampled selected monitoring area, predict the future vegetation coverage of the selected monitoring area, and comprehensively average the predicted vegetation coverage of each sampling area in the future to obtain the degradation prediction coefficient of the alpine grassland: Where ξ represents the degradation prediction coefficient of alpine grassland, which is used to quantify the degradation prediction status of alpine grassland, j represents the number of the selected monitoring area, j = 1, 2, 3, ..., j′, and j′ represents the total number of selected monitoring areas.

[0047] In this implementation scheme, the prediction of alpine grassland degradation also includes real-time monitoring of the operating status of the first-priority remote sensing equipment. For the fault condition of the first-priority remote sensing equipment monitored in real time, based on the above-mentioned confidence and support analysis, the second-ranked remote sensing equipment is selected as the second-priority remote sensing equipment for the alpine grassland. The monitoring data of the second-priority remote sensing equipment for the alpine grassland is used as the second-priority remote sensing representative data. The second-priority remote sensing representative data and the ground data of the sampled selected monitoring area are used for prediction and analysis to predict the future vegetation coverage of the sampled selected monitoring area. When the second-priority remote sensing equipment detects a fault condition, the third-priority remote sensing equipment is switched as above.

[0048] Specifically, the time series model adopts the ARIMAX model. The specific steps of predicting and analyzing the first-order remote sensing representative data and the ground data of the sampled selected monitoring area through the time series model are analyzed as follows: obtaining the first-order remote sensing representative data and the ground data of the sampled selected monitoring area as prediction data, and cleaning and preprocessing the first-order remote sensing representative data and the ground data of the sampled selected monitoring area; establishing the ARIMAX model, and using the historical remote sensing data and historical ground data of the alpine grassland to perform maximum likelihood estimation fitting on the ARIMAX model to estimate the model parameters; using the residual analysis method to diagnose the fitted ARIMAX model to check whether the residuals of the model meet the white noise assumption; using the fitted ARIMAX model to predict future alpine grassland monitoring data.

[0049] In this implementation plan, the first-order remote sensing representative data specifically includes vegetation coverage, and the exogenous variables in the ARIMAX model include the number of grazing livestock and the number of rat holes. The ARIMAX model formula is:t =β0+β1y t-1 +α1Δy t-1 +γ1ε t +γ2ε t-1 +γ3θ t +γ4θ t-1 +∈ t ; In the formula, y t represents the vegetation coverage at time point t, y t-1 represents a lag term of vegetation coverage, Δy t-1 represents a difference term of vegetation coverage, ε t represents the number of grazing livestock at time point t, ε t-1 represents a difference term for the number of grazing livestock, θ t represents the number of rat holes at time point t, θ t-1 represents a difference term of the number of rat holes, β0 and β1 represent the coefficients of the autoregressive part, representing the autocorrelation of vegetation coverage, α1 represents the coefficient of the difference part, representing the stability of vegetation coverage, γ1, γ2, γ3, and γ4 represent the coefficients of exogenous variables, representing the impact of grazing quantity and rat hole quantity on vegetation coverage, ∈ t It represents the error term of the ARIMAX model, which indicates the random fluctuations that cannot be explained by the ARIMAX model.

[0050] The specific analysis of the model parameters (β0, β1, α1, γ1, γ2, γ3, γ4) obtained by maximum likelihood estimation fitting is: the error term ∈ t The mean is 0 and the variance is σ 2 The specific formula of the likelihood function is: Where T represents the number of samples, ψ represents the model parameters (β0, β1, α1, γ1, γ2, γ3, γ4), and y t Indicates the vegetation coverage monitored by remote sensing equipment, x t represents the exogenous variables (the number of grazing livestock and the number of rat holes); by taking the logarithm of the likelihood function, we get the log-likelihood function:

[0051] Maximizing the log-likelihood function is equivalent to minimizing the sum of squares of errors. Therefore, by maximizing the log-likelihood function, we obtain β0, β1, α1, γ1, γ2, γ3, and γ4.

[0052] The specific analysis of the diagnosis of the fitted ARI MAX model using the residual analysis method is as follows: the historical remote sensing data and historical ground data of alpine grassland are used to fit the ARI MAX model to obtain parameter estimates, and then the residual sequence of the model is calculated, that is, the difference between the observed value and the model predicted value. The autocorrelation function and partial autocorrelation function graph of the residual sequence are used to observe whether there is significant autocorrelation or partial autocorrelation. If there is autocorrelation or partial autocorrelation in the residual sequence, it means that there is a problem with the ARIMAX model and further adjustment or improvement is needed. By performing residual analysis on the ARI MAX model, it can help determine the applicability and reliability of the model, thereby improving the accuracy and credibility of the prediction of alpine grassland degradation.

[0053] See also Figure 2 A system for monitoring and predicting degradation of alpine grasslands includes: a sampling and selecting monitoring area module, which is used to determine the grazing sampling and selecting monitoring areas and the rat hole sampling and selecting monitoring areas of the alpine grasslands based on the grazing monitoring data and rat hole monitoring data of each monitoring area of ​​the alpine grassland and the ground data of the historical area of ​​the alpine grassland; a ground data acquisition module, which is used to monitor the grazing sampling and selecting monitoring areas and the rat hole sampling and selecting monitoring areas of the alpine grasslands to obtain the ground data of the sampling and selecting monitoring areas; a degradation prediction module, which is used to analyze the data collection of multiple remote sensing devices based on the meteorological altitude factors of the alpine areas, select the first-order remote sensing representative data of the alpine grassland, and evaluate the grassland monitoring status of the sampling and selecting monitoring areas in combination with the ground data of the sampling and selecting monitoring areas, so as to realize the prediction of degradation of the alpine grasslands.

[0054] In summary, this application has at least the following effects:

[0055] By comprehensively analyzing the grazing monitoring data, rathole monitoring data and historical ground data of alpine grasslands, as well as the data collection of multiple remote sensing equipment, it is possible to provide comprehensive and accurate alpine grassland monitoring information; using the ground data of the monitoring areas selected by grazing sampling and rathole sampling, combined with remote sensing data for evaluation, it is possible to more effectively monitor the status and changes of alpine grasslands; combining meteorological and altitude factors with the selection of first-priority remote sensing representative data, it is possible to obtain high-quality monitoring data in a timely manner, providing timely support for the prediction of alpine grassland degradation; by evaluating the grassland monitoring status in the monitoring areas selected by sampling, combined with historical data and meteorological factors, it is possible to predict the degradation trend of alpine grasslands, and help to take corresponding protection and management measures.

[0056] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods and systems. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may 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 codes.

[0057] The present invention is described with reference to the flowcharts and structure diagrams of the methods and systems according to the embodiments of the present invention. It should be understood that each process and combination of modules in the flowcharts and structure diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts. Figure 1 process or multiple processes and structures Figure 1 A device that specifies functionality in a module or modules.

[0058] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 process or multiple processes and structures Figure 1 A function specified in a module or multiple modules.

[0059] 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 process or multiple processes and structures Figure 1 Steps to specify functionality in a module or multiple modules.

[0060] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0061] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for monitoring and predicting alpine grassland degradation, characterized in that: The following steps are involved: Based on the similarity analysis of the regional ground data of the alpine grassland in all the monitoring areas of the alpine grassland and the overall ground data of the historical areas of the alpine grassland, the monitoring areas of the alpine grassland are numbered in descending order according to the similarity between the regional ground data of the alpine grassland in each monitoring area and the overall ground data of the historical areas of the alpine grassland; The optimal number of monitoring areas among multiple monitoring areas is determined by Monte Carlo algorithm for each monitoring area of ​​the alpine grassland, and the monitoring areas are selected in order of size according to the similarity between each monitoring area and the overall ground data of the historical area of ​​the alpine grassland, and the ground data of the selected monitoring areas are obtained by monitoring; Based on the meteorological altitude factors in the alpine areas, the data collection of multiple remote sensing equipment was analyzed, the first-priority remote sensing representative data of the alpine grassland was selected, and the grassland degradation status of the monitoring area was evaluated in combination with the ground data of the selected monitoring area, thereby realizing the prediction of alpine grassland degradation; The monitoring area includes grazing monitoring area and rat hole monitoring area.

2. The alpine grassland degradation monitoring and prediction method according to claim 1 is characterized in that: The regional ground data of the alpine grassland includes grazing monitoring data and rat hole monitoring data. The specific analysis of determining the selected grazing monitoring area and rat hole monitoring area in the alpine region is as follows: The alpine grasslands are divided and numbered according to their geographical coordinates: AGH1, AGH2, AGH3, ..., AGHn; Based on the database of alpine grassland, the historical grazing monitoring data and rat hole monitoring data of each monitoring area are extracted, and the grazing monitoring data and rat hole monitoring data are preprocessed and feature extracted to obtain the grazing characteristic data set and rat hole characteristic data set of each monitoring area, and further obtain the mean of grazing characteristic data and rat hole characteristic data of each monitoring area; Extracting historical regional ground data of alpine grassland based on the database of alpine grassland, the historical ground data includes overall grazing data and overall rat hole data of alpine grassland, preprocessing and feature extraction of the overall historical grazing data and the overall historical rat hole data, respectively obtaining the overall historical grazing feature data set and the overall historical rat hole feature data set of the alpine grassland, and further obtaining the average value of the overall historical grazing feature data and the average value of the overall historical rat hole feature data of the alpine grassland; Based on the Euclidean distance, the mean grazing characteristic data and rat hole characteristic data of each monitoring area were clustered with the mean historical grazing characteristic data and the mean historical rat hole characteristic data of the alpine grassland, and each monitoring area was renumbered according to the similarity value: AGH1',AGH2',AGH3',...,AGHn'; The optimal number of monitoring areas in alpine regions is determined based on the Monte Carlo sampling algorithm, and then the selected monitoring areas of alpine grasslands are marked according to the numbering order. The selected monitoring areas include grazing selected monitoring areas and rat hole selected monitoring areas.

3. The alpine grassland degradation monitoring and prediction method according to claim 2 is characterized in that: The grazing characteristic data set of each selected monitoring area is specifically a grazing growth rate set of each selected monitoring area; the rat hole characteristic data set of each selected monitoring area is specifically a rat hole number reduction rate set of each selected monitoring area.

4. The alpine grassland degradation monitoring and prediction method according to claim 1 is characterized in that: The grazing monitoring data specifically refers to the number of grazing livestock.

5. The alpine grassland degradation monitoring and prediction method according to claim 1 is characterized in that: The specific analysis of the first-tier remote sensing representative data of alpine grassland is as follows: The meteorological altitude factors in high-cold areas specifically include the highest altitude, highest wind speed, and lowest temperature in high-cold areas; The highest altitude, highest wind speed and lowest temperature in the alpine areas are preprocessed to determine the minimum support threshold, and the remote sensing equipment in the alpine areas is numbered. Based on the historical data collection of the remote sensing equipment stored in the database, the data collection information of each remote sensing equipment is obtained. The data collection information of each remote sensing equipment specifically includes the optimal adaptation altitude, optimal adaptation wind speed and optimal adaptation temperature of each remote sensing equipment. Based on the confidence and support analysis of the optimal adaptation altitude, optimal adaptation wind speed, optimal adaptation temperature of each remote sensing equipment and the highest altitude, highest wind speed, and lowest temperature in the alpine areas, the remote sensing equipment with the best comprehensive ranking based on confidence and support is used as the first-priority remote sensing equipment for alpine grasslands, and the monitoring data of the first-priority remote sensing equipment for alpine grasslands is used as the first-priority remote sensing representative data.

6. The alpine grassland degradation monitoring and prediction method according to claim 5, characterized in that: The specific analysis process of obtaining the data collection information of each remote sensing device is as follows: Collect historical data of each remote sensing device based on the database, including equipment working environment parameters and collected data; filter historical data according to the parameter targets of altitude, temperature and wind speed of the remote sensing device, and retain the working environment parameters and corresponding collected data under the parameter target requirements; The working environment parameters under the retained parameter target requirements and the corresponding collected data are counted and analyzed, and the working environment parameters with the smallest data error collected by each remote sensing device are selected as the data collection information of each remote sensing device.

7. The alpine grassland degradation monitoring and prediction method according to claim 1, characterized in that: The specific analysis for predicting the degradation of alpine grassland is as follows: Obtaining the first-priority remote sensing representative data and the ground data of the selected monitoring area during each monitoring time period of the first-priority remote sensing equipment and the selected monitoring area; The time series model is used to predict and analyze the first-order remote sensing representative data and the ground data of the selected monitoring area, and the future vegetation coverage of the selected monitoring area is predicted. The predicted future vegetation coverage of each selected monitoring area is comprehensively averaged to obtain the degradation prediction coefficient of the alpine grassland. The degradation prediction coefficient of the alpine grassland is used to quantify the degradation prediction status of the alpine grassland.

8. The alpine grassland degradation monitoring and prediction method according to claim 7, characterized in that: The time series model adopts the ARIMAX model. The specific steps of predicting and analyzing the ground data of the monitoring area selected by the time series model for the first-order remote sensing representative data are as follows: Acquire the first-priority remote sensing representative data and the ground data of the selected monitoring area as prediction data, and clean and pre-process the first-priority remote sensing representative data and the ground data of the selected monitoring area; Establish an ARIMAX model, use historical remote sensing data and historical ground data of alpine grassland to perform maximum likelihood estimation fitting on the ARIMAX model, and estimate model parameters; The residual analysis method is used to diagnose the fitted ARIMAX model to check whether the residual of the model conforms to the white noise assumption; The fitted ARIMAX model is used to predict future alpine grassland monitoring data.

9. The alpine grassland degradation monitoring and prediction method according to claim 8, characterized in that: The first-order remote sensing representative data specifically include vegetation coverage. The exogenous variables in the ARIMAX model include the number of grazing livestock and the number of rat holes. The ARIMAX model formula is: y t =β0+β1y t-1 +α1Δy t-1 +γ1ε t +γ2ε t-1 +γ2θ t +γ4θ t-1 +e t ; In the formula, y t represents the vegetation coverage at time point t, y t-1 represents a lag term of vegetation coverage, Δy t-1 represents a difference term of vegetation coverage, ε t represents the number of grazing livestock at time point t, ε t-1 represents a difference term for the number of grazing livestock, θ t represents the number of rat holes at time point t, θ t-1 represents a differential term of the number of rat holes, β0 and β1 represent the coefficients of the autoregressive part, representing the autocorrelation of vegetation coverage, α1 represents the coefficient of the differential part, representing the stability of vegetation coverage, γ1, γ2, γ3, and γ4 represent the coefficients of exogenous variables, representing the impact of grazing quantity and rat hole quantity on vegetation coverage, and ε t It represents the error term of the ARIMAX model, which indicates the random fluctuations that cannot be explained by the ARIMAX model.

10. A system for monitoring and predicting alpine grassland degradation, using the method for monitoring and predicting alpine grassland degradation according to any one of claims 1 to 9, characterized in that: include: The selected monitoring area module is used to determine the selected monitoring areas for grazing and the selected monitoring areas for rat holes in the alpine grassland based on the grazing monitoring data and rat hole monitoring data of each selected monitoring area in the alpine grassland and the ground data of the historical area of ​​the alpine grassland; A ground data acquisition module is used to monitor the selected monitoring area for grazing in alpine grassland and the selected monitoring area for rat holes to obtain ground data of the selected monitoring area; The degradation prediction module is used to analyze the data collection of multiple remote sensing devices based on the meteorological altitude factors in the alpine areas, select the first-priority remote sensing representative data of the alpine grassland, and combine the ground data of the selected monitoring area to evaluate whether the grassland in the selected monitoring area is degraded and the degree of degradation.

Images