Driving factor analysis method and system for vegetation coverage spatio-temporal variation

Abstract

The invention discloses a drive factor analysis method and system for vegetation coverage spatio-temporal variation, and relates to the field of remote sensing information and environmental sciences, and the method comprises the steps: carrying out the time variation analysis of the annual average vegetation coverage of a river basin for many years; the change trend of the vegetation coverage is analyzed by adopting a Theil-Sen Median method, and the significance of the vegetation coverage is checked and evaluated by adopting Mann-Kendall. The variation coefficient is used for analyzing the fluctuation degree. And finally, predicting the development trend of the vegetation coverage by using a Hurst index, and analyzing spatial change. 13 natural factors and human factors are selected for discussing driving factors of temporal and spatial variation of vegetation coverage of a certain river basin; a geographic detector method is utilized to analyze dominant factors influencing spatial change of vegetation coverage, and a model of contribution rate of land utilization transfer to vegetation coverage is constructed to quantify the contribution rate of land utilization transfer to regional vegetation coverage. According to the method, various driving factors can be comprehensively analyzed, so that the driving factors influencing the temporal and spatial change of the vegetation coverage are accurately determined.

Description

Technical Field

[0001] The invention relates to the fields of remote sensing information and environmental science, and in particular to a method and system for analyzing driving factors of spatiotemporal changes in vegetation coverage. Background Art

[0002] The study of spatial and temporal changes in vegetation cover is crucial for understanding ecosystem stability, monitoring environmental changes, and formulating corresponding protection and management strategies. With the development of remote sensing technology, scientists are able to monitor and analyze changes in vegetation cover on a large scale.

[0003] In the prior art, the geographic detector Geodetector of method 1 is a statistical method used to analyze the impact of different factors on the spatial distribution of vegetation coverage. Existing studies have used the Geodetector method to analyze the driving factors of the spatiotemporal changes in the normalized vegetation index NDVI of vegetation in a certain area, and found that climate, topography and human activities are the main influencing factors. However, the interpretation ability of the Geodetector method may be limited by factors such as data quality, sample size and model assumptions. Method 2 uses the pixel fusion model Dimidiatepixel, which can effectively deal with the problem of mixed pixels in remote sensing images. Through the dimidiate pixel model, vegetation coverage information can be more accurately extracted, and then the spatiotemporal variation characteristics of vegetation coverage can be analyzed. However, the dimidiate pixel model requires high computing power and complex data processing procedures when processing remote sensing images, which is a challenge for the processing and analysis of large-scale data.

[0004] In summary, how to comprehensively and accurately determine the driving factors that affect the spatiotemporal changes in vegetation cover is an important issue that needs to be solved urgently. Summary of the invention

[0005] The embodiment of the present invention provides a method and system for analyzing driving factors of spatiotemporal changes in vegetation coverage, which can solve the problem in the prior art of how to comprehensively and accurately determine the driving factors that affect the spatiotemporal changes in vegetation coverage.

[0006] The embodiment of the present invention provides a method for analyzing driving factors of spatiotemporal changes in vegetation coverage, comprising the following steps: Obtain the average annual vegetation coverage of each river basin section for many years; According to the annual average vegetation coverage of each river basin section over many years, the dynamic change trend of different annual average vegetation coverage over time in a long time series is obtained, and the median slope estimation method is used to obtain the spatial change trend of the annual average vegetation coverage of each river basin section over many years; According to multiple driving factors covering natural and human factors, factor detection is carried out on the multiple driving factors using Geodetector, and the explanatory degree of each driving factor for the temporal and spatial change trends of the annual average vegetation coverage of each river basin segment in the river basin is quantified; among the explanatory degrees corresponding to the multiple driving factors, the driving factor with the largest explanatory degree is used as the driving factor affecting the spatio-temporal change of the vegetation coverage in the river basin.

[0007] Furthermore, the method for analyzing the driving factors of the spatio-temporal change of the vegetation coverage further includes: Determine that the land use type transfer factor is the driving factor affecting the spatio-temporal change of the vegetation coverage in the river basin, where the land use type transfer factor is the driving factor for the conversion of land use from one type to another; Construct a land use transfer vegetation coverage contribution rate model for evaluating the impact of land use type transfer on the vegetation coverage degree in the affected area. The land use transfer vegetation coverage contribution rate model has the formula: FVCR =( FVC 1 - FVC 0 )x( LA / TA ); Wherein, FVCR is the vegetation coverage contribution rate; FVC 1 and FVC 0 are respectively the average vegetation coverages at the end and the beginning when the land use change type changes; LA is the land area where the land use type in the river basin changes; TA is the total area of the river basin; Evaluate the impact of land use type transfer on the vegetation coverage according to the land use transfer vegetation coverage contribution rate model; According to the land use type data of each river basin segment in the river basin over the years, obtain the area of each land use type converted into another land use type; according to the annual average vegetation coverage of each river basin segment in the river basin over the years, obtain the average vegetation coverages at the end and the beginning when the land use change type changes during the conversion of each land use type into another land use type; Input the proportion of the area of each land use type converted into another land use type in the total area of the river basin and the difference between the vegetation coverage at the end of the land use type transfer change and the vegetation coverage at the beginning into the land use transfer vegetation coverage contribution rate model to obtain the impact degree of each land use type converted into another land use type on the vegetation coverage.

[0008] Further, the step of obtaining the dynamic change trend of different annual average vegetation coverage over time in a long time series specifically includes: Dividing the annual average vegetation coverage of each river basin segment in each year into grades according to a set threshold; Sorting the classification results of multiple years according to time, and obtaining the dynamic change of the grade of the vegetation coverage of each river basin segment in the river basin over time in a long time series.

[0009] Further, the step of dividing the annual average vegetation coverage of each river basin segment in each year into grades according to a set threshold specifically includes: Classifying the annual average vegetation coverage greater than 0 and less than or equal to 0.05 as a low level; classifying the annual average vegetation coverage greater than 0.05 and less than or equal to 0.1 as a low level; classifying the annual average vegetation coverage greater than 0.1 and less than or equal to 0.15 as a medium level; classifying the annual average vegetation coverage greater than 0.15 and less than or equal to 0.2 as a high level; classifying the annual average vegetation coverage greater than or equal to 0.2 as a high level.

[0010] Further, the spatial change trend of the annual average vegetation coverage of each river basin segment in multiple years needs to use the non-parametric test method Mann-Kendall to verify the significance of the spatial change trend.

[0011] Further, the method further includes: predicting the change trend of the vegetation coverage of each river basin segment in the river basin according to the median slope estimation method and the Hurst index Hurst.

[0012] Further, the step of obtaining the annual average vegetation coverage of each river basin segment in multiple years specifically includes: Obtaining remote sensing image data of the river basin in multiple years, and taking the average value of NDVI from April to October each year as the annual average NDVI according to the remote sensing image data; Obtaining the vegetation coverage according to the improved pixel dichotomy method FVC : FVC = NDVI - NDVI soil / NDVI veg - NDVI soil ; Wherein, NDVI represents the value of a single pixel, NDVI soil represents the information of the non-vegetation-covered part in the pixel, NDVI veg represents the information of the vegetation coverage ratio in the pixel.

[0013] An embodiment of the present invention provides a system for analyzing driving factors of spatio-temporal changes in vegetation coverage, including: A vegetation coverage acquisition module for acquiring the annual average vegetation coverage of each river basin segment in a river basin over multiple years; A spatio-temporal change analysis module for the vegetation coverage, based on the annual average vegetation coverage of each river basin segment in a river basin over multiple years, to obtain the dynamic change trend of different annual average vegetation coverages over time in a long time series, and use the median slope estimation method to obtain the spatial change trend of the annual average vegetation coverage of each river basin segment in the river basin over multiple years; A driving factor analysis module for using the Geodetector to detect multiple driving factors covering natural and human factors, quantifying the explanatory degree of each driving factor for the time change trend and spatial change trend of the annual average vegetation coverage of each river basin segment in the river basin; among the explanatory degrees corresponding to the multiple driving factors, taking the driving factor with the largest explanatory degree as the driving factor affecting the spatio-temporal changes in the vegetation coverage of the river basin.

[0014] An embodiment of the present invention provides a method and system for analyzing driving factors of spatio-temporal changes in vegetation coverage. Compared with the prior art, the beneficial effects are as follows: Based on the annual average vegetation coverage of each river basin segment in a river basin over multiple years, obtain the dynamic change trend of different annual average vegetation coverages over time in a long time series, and use the median slope estimation method to obtain the spatial change trend of the annual average vegetation coverage of each river basin segment in the river basin over multiple years; according to multiple driving factors covering natural and human factors, use the Geodetector to detect the multiple driving factors, quantify the explanatory degree of each driving factor for the time change trend and spatial change trend of the annual average vegetation coverage of each river basin segment in the river basin; among the explanatory degrees corresponding to the multiple driving factors, take the driving factor with the largest explanatory degree as the driving factor affecting the spatio-temporal changes in the vegetation coverage of the river basin.

[0015] Among them, by using the Geodetector to detect multiple driving factors, it is possible to quantify the explanatory degree of each driving factor for the time change trend and spatial change trend of the annual average vegetation coverage of each river basin segment in the river basin, and identify the main driving factors affecting the spatio-temporal changes in the annual average vegetation coverage of the river basin, comprehensively analyze a variety of driving factors, so as to accurately determine the driving factors affecting the spatio-temporal changes in vegetation coverage. Description of the Drawings

[0016] Figure 1 It is a flowchart provided by an embodiment of the present invention; Figure 2Schematic diagram of the process for analyzing vegetation cover change and its driving forces from the perspective of river basins provided by the embodiments of the present invention; Figure 3 Results of vegetation cover under the influence of different impact factors in the river basins provided by the embodiments of the present invention; Figure 4 Vegetation cover contribution rate of land use transfer provided by the embodiments of the present invention; Figure 5 Trend of the overall vegetation cover level in the river basin changing over time provided by the embodiments of the present invention. Detailed implementation manners

[0017] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following describes the detailed implementation manners of the present invention with reference to the accompanying drawings. Many specific details are set forth in the following description to fully understand the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0018] Refer to Figure 1 , the embodiments of the present invention provide a method for analyzing the driving factors of the spatio-temporal change of vegetation coverage, including the following steps: Step 1: Obtain the average annual vegetation coverage of each river basin segment in the river basin over the years.

[0019] Step 2: According to the average annual vegetation coverage of each river basin segment in the river basin over the years, obtain the dynamic change trend of different average annual vegetation coverages over time in the long time series, and use the median slope estimation method to obtain the spatial change trend of the average annual vegetation coverage of each river basin segment in the river basin over the years.

[0020] Step 3: According to multiple driving factors covering natural factors and human factors, use Geodetector to detect the factors of multiple driving factors, and quantify the explanatory degree of each driving factor on the time change trend and spatial change trend of the average annual vegetation coverage of each river basin segment in the river basin; among the explanatory degrees corresponding to multiple driving factors, take the driving factor with the largest explanatory degree as the driving factor affecting the spatio-temporal change of the vegetation coverage in the river basin.

[0021] Step 4: According to the explanatory degree, finally determine that the land use type transfer factor is the driving factor affecting the spatio-temporal change of the vegetation coverage in the river basin. Construct a land use transfer vegetation cover contribution rate model for evaluating the impact of land use type transfer on the vegetation coverage degree in the region.

[0022] The land use transfer vegetation cover contribution rate model, the formula is: FVCR =(FVC 1 - FVC 0 )x( LA / TA ) wherein, FVCR is the contribution rate of vegetation coverage; FVC 1 and FVC 0 are the average vegetation coverages at the end and the beginning of the period when the land use change type changes, respectively; LA is the land area where the land use type in the river basin has changed; TA is the total area of the river basin.

[0023] According to the land use transfer vegetation coverage contribution rate model, evaluate the impact of land use type transfer on vegetation coverage.

[0024] According to the land use type data of each river basin section in the river basin over the years, obtain the area of each land use type changing to another land use type; according to the average annual vegetation coverage of each river basin section in the river basin over the years, obtain the average vegetation coverages at the end and the beginning of the period when the land use change type changes during the process of each land use type changing to another land use type.

[0025] Input the proportion of the area of each land use type changing to another land use type in the total area of the river basin, and the difference between the vegetation coverage at the end of the land use type transfer change and the vegetation coverage at the beginning of the change into the land use transfer vegetation coverage contribution rate model to obtain the impact degree of each land use type changing to another land use type on vegetation coverage.

[0026] The main methods and steps used in the present invention are as Figure 2As shown in the figure. To study the specific spatio-temporal changes in vegetation cover in a certain river basin, the study area is divided into upper, middle, and lower reaches according to the river source and administrative division. Based on Landsat remote sensing satellite images, the average value from April to October is selected as the annual NDVI, and the maximum value composite method and the medium pixel model are used to calculate the annual vegetation coverage. At the same time, the Theil-Sen Median method, a median slope estimation method, is used to analyze the change trend of vegetation coverage, and the Mann-Kendall test, a non-parametric test method, is used to evaluate its significance. The coefficient of variation is used to analyze the degree of fluctuation. Finally, the Hurst index is used to predict the development trend of vegetation coverage. To explore the driving factors of the spatio-temporal changes in vegetation coverage in a certain river basin, 13 natural and anthropogenic factors are selected, including precipitation, land use type, sunshine hours, drought index, soil type, surface temperature, population density, relative humidity, altitude, temperature, soil erosion, slope, and GDP.

[0027] Then, the geodetector method is used to analyze the dominant factors affecting the spatial change of vegetation coverage, and to quantify the influence degree and interaction of different factors on vegetation coverage. To explore the impact of anthropogenic factors, especially land use conversion, on vegetation cover, this paper innovatively introduces the formula for the contribution rate of land use transfer to vegetation cover. This formula is used to quantify the contribution rate of land use transfer to regional vegetation cover.

[0028] The specific optimization steps are as follows: (1) Determination of the basin area: It is carried out according to methods such as the topographic map method, remote sensing image method, and field survey method. Based on high-precision topographic maps, remote sensing images, or field measurement data, through steps such as drawing the watershed line, extracting water system information, and calculating the area, the boundary and area of the basin are finally determined. In the process, factors such as data accuracy, basin characteristics, and calculation methods need to be fully considered to ensure the accuracy and reliability of the results;

[0029] (2) Data acquisition and preprocessing: Using satellite remote sensing technology, collect image data of the study area at different time series within the study area, and calculate the vegetation coverage according to the improved pixel dichotomy method FVC : As shown in Formula 1: FVC = NDVI - NDVI soil / NDVI veg - NDVI soil (1) In the formula, NDVI represents the value of a single pixel, NDVI soil represents the information of the non-vegetation-covered part of the pixel, NDVIveg It represents the proportion of vegetation cover in the representative pixel.

[0030] (3) Multi-source data integration: Through GIS technology, obtain data on various land use types within the study area, including historical data and current data. Clean, calibrate, and perform projection conversion on the collected data to ensure data consistency and accuracy.

[0031] (4) Long-term analysis of vegetation cover: Divide the level of vegetation cover within the basin into different grades to explore the dynamic changes in basin vegetation cover over a long time series and analyze the time variation pattern.

[0032] (5) Analysis of spatial changes in vegetation cover: This invention uses the non-parametric statistical method of Theil–Sen Median slope estimation method to analyze the spatial change trend of basin vegetation cover, as shown in Equation 2: 𝛽 = 𝑀𝑒d𝑖𝑎𝑛(𝐹𝑉𝐶 𝑗 −𝐹𝑉𝐶 𝑖 / 𝑗 − 𝑖), ∀𝑖 < 𝑗 (2) Based on this method, this patent inputs the basin image data of a long time series (multiple years) that has been preprocessed and calculates the annual average vegetation cover to explore the spatial change trend of the annual average vegetation coverage of each basin segment in the river basin. If 𝛽 is greater than zero, it indicates an increasing trend in vegetation coverage; otherwise, it indicates a decreasing trend in vegetation coverage. What this method finally obtains is the spatial change of the overall vegetation coverage of the basin at the pixel scale within a certain time series.

[0033] The main advantage of this method lies in its strong error avoidance ability. Even in the case of non-normal or heteroscedastic data, it can provide reliable trend estimates. This method takes the median of these slopes as the trend of the entire time series, representing the long-term change trend of the time series data. The Mann–Kendall method is also a non-parametric statistical method, commonly used to detect monotonic trends in time series data. Similar to the Theil–Sen Median method, the Mann–Kendall method is also applicable to various types of data distributions, including non-normal distributions and data with outliers. This method constructs a cumulative sum sequence by comparing the relative magnitudes between data pairs and evaluates the significance of the trend based on this sequence.

[0034] (6) Analysis of driving force factors: This invention is to explore the factors affecting vegetation cover ( FVCThe driving forces of spatio-temporal changes, 13 driving factors covering natural and anthropogenic factors were selected for analysis. The Geodetector method was adopted, which is a set of statistical methods capable of effectively detecting the influence of driving factors without the need to meet traditional statistical assumptions. The factor detector in the Geodetector method was used to analyze the explanatory power of each factor for FVC the spatial variability.

[0035] (7) Through factor detector analysis, it was found that all factors significantly affected the FVC spatial variability (p < 0.01). The specific factor explanatory powers (q values) were different. Among them, factors such as precipitation, sunshine hours, drought index, and soil erosion had relatively large explanatory powers, reaching 34.5%, 13.3%, 12.3%, and 12.1% respectively.

[0036] (8) Interaction detection: The interaction detector in the Geodetector method was used to analyze the interaction between factors, and it was found that there were non-linear enhancement effects and two-factor enhancement effects between factors. This indicates that when certain factors are combined, their influence on FVC is not simply additive, but there are more complex interaction relationships.

[0037] (9) Ecological detector analysis: The ecological detector was used to compare and analyze the significant differences in the spatial distribution of each pair of factors to further understand the relative importance between factors. The average values within different factor intervals were analyzed to determine the optimal conditions for vegetation growth. FVC The average values within different factor intervals were analyzed to determine the optimal conditions for vegetation growth. FVC of vegetation growth were determined.

[0038] (10) Calculation of the contribution rate of land use transfer to vegetation cover: According to the results of the analysis of the driving forces of vegetation cover change, it was found that land use transformation had the greatest impact on vegetation cover change. To explore which specific land transfer had the greatest impact on the vegetation cover of the basin, the present invention proposed a formula for the contribution rate of land use transfer to vegetation cover, as shown in Equation 3.

[0039] FVCR =( FVC 1 - FVC 0 )x( LA / TA ) (3) In the formula, FVCR is the contribution rate of vegetation cover, FVC 1 and FVC 0 are the average vegetation coverages at the end and beginning of land use change respectively. LA is the land area of the change type; TAis the total area of the study area. This formula can directly quantify the contribution of land use change to vegetation cover. This helps analyze which type of land use change has the greatest impact on regional vegetation cover and provides valuable suggestions for regional vegetation restoration and ecological protection. As Figure 4 shown is the contribution rate of land use transfer to vegetation cover.

[0040] The summary content of the present invention is as follows: Integrated data processing and analysis platform: The present invention constructs an integrated data processing and analysis platform, which integrates multiple functional modules such as remote sensing data processing, vegetation coverage extraction, driving factor analysis, and future trend prediction. This integrated design enables users to complete the entire process of studying the spatio-temporal changes of vegetation cover on one platform, improving the efficiency and accuracy of the research.

[0041] Integration of Geodetector method and multi-factor analysis: The present invention combines the Geodetector method with the multi-factor analysis method. By comprehensively considering the impacts of various natural and human factors on the spatio-temporal changes of vegetation cover, the accuracy and comprehensiveness of driving factor analysis are improved. This method can not only identify the main driving factors but also quantify the specific contributions of each factor to the vegetation cover change.

[0042] Future trend prediction model: Based on historical data and the results of driving factor analysis, the present invention constructs a future trend prediction model. This model can predict the spatio-temporal change trends of vegetation cover in the future for a period of time, providing a scientific basis and decision-making support for ecosystem protection and management.

[0043] The beneficial effects of the present invention also include: (1) From the perspective of the basin, a comprehensive analysis is carried out on the regional vegetation cover change and driving force analysis, involving the integration and fusion of multi-source data, including remote sensing data, meteorological data, socio-economic data, etc. The comprehensive application of these data can provide more comprehensive and accurate information support.

[0044] (2) The driving forces leading to the change of basin vegetation cover are multi-faceted, including natural factors and non-natural factors. The present invention comprehensively considers multiple factors and selects representative driving force factors for analysis in order to more accurately reveal the laws and mechanisms of vegetation cover change.

[0045] (3) The formula for the contribution rate of land transfer to vegetation cover proposed by the present invention aims to measure the impact of specific land use transformation on basin vegetation cover. This formula can quantitatively analyze the contribution degree of the transformation between different land use types to the vegetation cover change, thus revealing the internal connection between land use transformation and vegetation cover change.

[0046] An embodiment of the present invention provides a system for analyzing the driving factors of the spatio-temporal variation of vegetation coverage, including: A vegetation coverage acquisition module for acquiring the average annual vegetation coverage of each river basin segment in a river basin over the years.

[0047] A spatio-temporal variation analysis module of vegetation coverage for obtaining the dynamic change trend of different average annual vegetation coverages over time in a long time series according to the average annual vegetation coverage of each river basin segment in a river basin over the years, and using the median slope estimation method to obtain the spatial change trend of the average annual vegetation coverage of each river basin segment in a river basin over the years.

[0048] A driving factor analysis module for using Geodetector to detect multiple driving factors covering natural factors and human factors, quantifying the degree of explanation of each driving factor for the temporal change trend and spatial change trend of the average annual vegetation coverage of each river basin segment in a river basin; among the degrees of explanation corresponding to the multiple driving factors, taking the driving factor with the largest degree of explanation as the driving factor affecting the spatio-temporal variation of vegetation coverage in a river basin.

[0049] A specific embodiment is as follows: This embodiment proposes a method for analyzing the driving factors of the spatio-temporal variation of vegetation coverage, including the following steps: S1. Select the river basin to be studied.

[0050] A high-precision topographic map of a certain river basin was obtained, and the boundary of the river basin was clarified by drawing the watershed line. At the same time, using the contour line information on the topographic map and combining with the geomorphic characteristics of the river basin, the area of the river basin was preliminarily calculated. Using satellite remote sensing images, the water system information of a certain river basin was extracted to further verify the river basin boundary determined by the topographic map method. At the same time, through image processing software, the area of the river basin was accurately calculated.

[0051] S2. Data collection and processing.

[0052] After determining the area of the river basin, image data of different time series within a certain river basin were collected using satellite remote sensing technology. Subsequently, the NDVI value was calculated according to the pixel dichotomy method. In order to more comprehensively analyze the vegetation cover change of a certain river basin, the present invention obtained the data of each land use type in the research area through GIS technology and integrated these data. Including land use type, vegetation type, soil type, etc. After obtaining the data, the collected data were cleaned, calibrated, and projection transformed to ensure the consistency and accuracy of the data.

[0053] S3. Temporal variation analysis based on the vegetation cover of the river basin.

[0054] Analyze the temporal variation trend of the annual average vegetation coverage in a river basin over the long time series from 2000 to 2022 through linear regression and t-tests, and conduct hierarchical classification based on the size of the average vegetation coverage. As Figure 5 shown.

[0055] Classify the annual average vegetation coverage greater than 0 and less than or equal to 0.05 as the low level; classify the annual average vegetation coverage greater than 0.05 and less than or equal to 0.1 as the low level; classify the annual average vegetation coverage greater than 0.1 and less than or equal to 0.15 as the medium level; classify the annual average vegetation coverage greater than 0.15 and less than or equal to 0.2 as the high level; classify the annual average vegetation coverage greater than or equal to 0.2 as the high level.

[0056] S4. Spatial variation analysis based on the vegetation coverage of the basin.

[0057] Use the Theil-Sen Median and the Mann-Kendall trend analysis methods to explore the spatio-temporal variation characteristics of FVC in a river basin from 2000 to 2022. The FVC of 45.9% of the area in the river basin shows an increasing trend, and the significantly increasing areas are in the southern part of the middle reaches and the downstream area of the river. The FVC of about 51.5% of the area in the river basin shows a decreasing trend, and the significantly decreasing areas are mainly located in the northwest of the study area.

[0058] Predict the future change trend by superimposing the Theil-Sen slope and the Hurst index. When the Theil-Sen is greater than 0, it indicates that the vegetation coverage is increasing; when it is less than 0, it decreases; when the Hurst is greater than 0.5, it indicates that the change trend of FVC continues (that is, if the past vegetation coverage decreased, it will still decrease in the future, and if it increased, it will continue to increase), and when it is less than 0.5, the change trend may reverse (that is, if the past vegetation coverage decreased, it may increase in the future, and if it increased, it may decrease). When the Hurst is equal to 0.5, it indicates that the change regularity of the vegetation coverage is difficult to predict. Therefore, superimposing the two can analyze the future change trend of the basin.

[0059] Therefore, superimposing the two can analyze the future change trend of the basin. It should be noted that the present invention is based on the study at the pixel scale, so the change trend of the vegetation coverage of a single pixel in the basin can be obtained. When the Theil-Sen is greater than 0 and the Hurst is greater than 0.5, the change trend is continuously increasing; when the Theil-Sen is greater than 0 and the Hurst is less than 0.5, the change trend is increasing first and then decreasing; when the Theil-Sen is less than 0 and the Hurst is greater than 0.5, the change trend is continuously decreasing; when the Theil-Sen is less than 0 and the Hurst is less than 0.5, the change trend is decreasing first and then increasing; when the Hurst is equal to 0.5, the change trend is unstable.

[0060] In the future, the FVC in a certain river basin will mainly show a downward trend, accounting for 58.0% of the region. Generally speaking, there is a trend of improvement first and then degradation in the upper reaches of a certain river, while the vegetation improvement in the middle and lower reaches has increased, but the vegetation restoration still needs to be strengthened.

[0061] S5. Analysis of the driving forces of vegetation cover change in a certain river basin.

[0062] Geodetector factor detection: The spatio-temporal analysis results of vegetation cover show that its distribution has strong spatial heterogeneity. Through factor detector analysis of the influencing factors of FVC spatial differentiation, the explanatory power from large to small is land use type, precipitation, sunshine hours, drought index, soil type, population density, surface temperature, relative humidity, altitude, temperature, soil erosion, slope and GDP. Among them, the q values of the human activity factor land use type and the meteorological factors precipitation, sunshine hours, and drought index are all greater than 12%, which are 34.5%, 13.3%, 12.3%, and 12.1% respectively. This indicates that the spatial change of vegetation cover in a certain river basin is comprehensively driven by human activities combined with meteorological factors, and the land use pattern largely determines the spatial change of vegetation cover.

[0063] Geodetector interaction detection: Using the interaction detector to detect the explanatory power of factor interaction on FVC change, the results show that there are obvious interaction effects on the influence of each factor on FVC. The influence of each factor on FVC is not independent, but occurs synergistically. The types of interaction effects are two-factor enhancement and non-linear enhancement, and the non-linear enhancement effect is greater than the two-factor enhancement. The strongest explanatory power of the interaction is between land use and precipitation, and the two-factor interaction q value reaches 0.523. The worst explanatory power of the interaction is between GDP and slope, and its q value is only 0.055. This may be because the slope and GDP themselves have poor explanatory power for FVC, so the interaction explanatory power is insufficient. At the same time, the land use type factor can produce relatively high q values when combined with most factors, but its combination with precipitation, sunshine hours, and soil type has the best effect, and the q value can reach above 0.5. The interaction explanatory power of slope and GDP with most factors is relatively low.

[0064] Risk detection: Based on the risk detector, the average value of FVC in the factor range or type of vegetation growth was determined. In terms of meteorological factors, with the increase in temperature, FVC showed a trend of first decreasing and then gradually increasing, reaching the maximum value at a temperature of 18 - 20 °C; with the increase in precipitation, FVC showed an upward trend in temperature, reaching the maximum value at a precipitation of 601 - 648 mm; with the increase in relative humidity, FVC showed a trend of first increasing, then decreasing and then increasing, reaching the maximum value at a relative humidity of 53.4% - 54.4%; with the increase in sunshine hours, FVC showed a trend of first increasing and then decreasing, reaching the maximum value at a sunshine hours of 7.86 - 8.07 hours.

[0065] As Figure 3 shown, the vegetation in a certain river basin is suitable for growing in a meteorological environment with high temperature, low sunshine, high rainfall, and medium humidity.

[0066] The above-described embodiments merely represent several implementation manners of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the appended claims.

Claims

1. A method for analyzing driving factors of spatiotemporal changes in vegetation coverage, characterized in that: The following steps are involved: Obtain the average annual vegetation coverage of each river basin section for many years; According to the annual average vegetation coverage of each river basin section over many years, the dynamic change trend of different annual average vegetation coverage over time in a long time series is obtained, and the median slope estimation method is used to obtain the spatial change trend of the annual average vegetation coverage of each river basin section over many years; Based on multiple driving factors covering natural factors and human factors, Geodetector was used to detect multiple driving factors and quantify the degree to which each driving factor explained the temporal and spatial trends of the annual average vegetation cover in each river basin segment. Among the explanation degrees corresponding to multiple driving factors, the driving factor with the greatest explanation degree is taken as the driving factor affecting the spatiotemporal changes of vegetation coverage in river basins.

2. The method for analyzing driving factors of spatiotemporal changes in vegetation coverage according to claim 1, characterized in that: The method for analyzing driving factors of spatiotemporal changes in vegetation coverage also includes: Determine the land use type transfer factor as the driving factor affecting the spatiotemporal change of vegetation coverage in the river basin, wherein the land use type transfer factor is the driving factor for the conversion of land use from one type to another; A land use conversion and coverage contribution rate model for evaluating the degree of regional vegetation coverage affected by land use type transfer is constructed. The land use conversion and coverage contribution rate model has the following formula: FVCR =( FVC 1 - FVC 0 )x( LA / TA )； in, FVCR is the vegetation coverage contribution rate; FVC 1 and FVC 0 are the average vegetation coverage at the end and beginning of the period when the land use change type changes; LA land area that has undergone a change in land use type in the river basin; TA is the total area of ​​the river basin; According to the land use conversion to vegetation coverage contribution rate model, the impact of land use type transfer on vegetation coverage was evaluated; Based on the land use type data of each river basin section over the years, the area of ​​each land use type converted to another land use type is obtained; based on the annual average vegetation coverage of each river basin section over the years, the average vegetation coverage at the end and beginning of the period when each land use type is converted to another land use type is obtained; The proportion of the area converted from each land use type to another land use type to the total area of ​​the river basin, as well as the difference between the vegetation coverage at the end of the land use type change and the vegetation coverage at the beginning of the change, are input into the land use conversion coverage contribution rate model to obtain the impact of the conversion of each land use type to another land use type on vegetation coverage.

3. The method for analyzing driving factors of spatiotemporal changes in vegetation coverage according to claim 1, characterized in that: The specific steps of obtaining the dynamic change trend of different annual average vegetation coverage over time in a long time series include: The annual average vegetation coverage of each river basin section in each year is divided into grades according to the set threshold; The division results of many years are sorted by time to obtain the dynamic changes of vegetation coverage levels in each river basin section over time series.

4. The method for analyzing driving factors of spatiotemporal changes in vegetation coverage according to claim 3, characterized in that: The specific steps of classifying the annual average vegetation coverage of each river basin section in each year according to the set threshold value include: The annual average vegetation coverage greater than 0 and less than or equal to 0.05 is classified as a low level; The annual average vegetation coverage greater than 0.05 and less than or equal to 0.1 is classified as low level; The annual average vegetation coverage greater than 0.1 and less than or equal to 0.15 is classified as medium level; The annual average vegetation coverage greater than 0.15 and less than or equal to 0.2 is classified as high level; The annual average vegetation coverage greater than or equal to 0.2 is considered as a high level.

5. The method for analyzing driving factors of spatiotemporal changes in vegetation coverage according to claim 1, characterized in that: The spatial variation trend of the annual average vegetation coverage of each river basin segment over the years needs to be verified using the non-parametric test method Mann-Kendall to verify the significance of the spatial variation trend.

6. The method for analyzing driving factors of spatiotemporal changes in vegetation coverage according to claim 1, characterized in that: The method further comprises: The changing trend of vegetation coverage in each river basin section is predicted based on the median slope estimation method and Hurst exponent.

7. The method for analyzing driving factors of spatiotemporal changes in vegetation coverage according to claim 1, characterized in that: The specific steps of obtaining the annual average vegetation coverage of each river basin section over many years include: Obtain remote sensing image data of river basins for many years, and obtain the average NDVI value from April to October each year as the annual average NDVI based on the remote sensing image data; Obtaining vegetation coverage based on improved pixel dichotomy FVC : FVC = NDVI - NDVI soil / NDVI veg - NDVI soil ； in, NDVI Represents the value of a single pixel. NDVI soil Represents the information of the non-vegetation covered part of the pixel. NDVI veg Represents the percentage of vegetation coverage in the pixel.

8. A system for analyzing driving factors of spatiotemporal changes in vegetation coverage, characterized in that: include: The vegetation coverage acquisition module is used to obtain the annual average vegetation coverage of each river basin section over many years; The spatiotemporal variation analysis module of vegetation coverage is used to obtain the dynamic variation trend of different annual average vegetation coverage over time in a long time series based on the annual average vegetation coverage of each river basin section over many years, and use the median slope estimation method to obtain the spatial variation trend of the annual average vegetation coverage of each river basin section over many years; The driving factor analysis module is used to detect multiple driving factors based on natural and human factors using Geodetector, and to quantify the degree to which each driving factor explains the temporal and spatial trends of the annual average vegetation coverage in each river basin segment; Among the explanation degrees corresponding to multiple driving factors, the driving factor with the greatest explanation degree is taken as the driving factor affecting the spatiotemporal changes of vegetation coverage in river basins.

Images